¹ Tierney, J. J., ‘The Celtic ethnography of Posidonius’, Proc. Roy. Irish Acad. 60 (1960), 189–275. Page-references cited in the text below are to this article, and unless otherwise stated, information about Poseidonios and his predecessors is also drawn from it. All references to Poseidonios's text are to the twenty-third book of his History only. I would like to thank Mr. C. E. Stevens, Miss B. M. Levick and Mr. G. E. M. de Ste Croix for their helpful criticisms of an earlier draft of this article. Naturally I take sole responsibility for the views expressed here.Google Scholar

² Athenaeus iv. 36 p. 151E–152D; iv. 40 p. 154A–C; vi. 49 p. 246C–D; iv. 37 p. 152D–F; Diadorus v, 25–32; Strabo 2. 5. 28, 4. 1–4; Caesar, BG 6. 11–28.

³ Strabo 4. 1. 9; 4. 1. 13; 4. 1. 14; 4. 4. 5; 4. 4. 6.

⁴ F. Jacoby, FGH 11A, 87, Fr. 116; 11c, p. 212 ff.

⁵ RE s.v. Diodoros col. 669; cf. Diodorus Siculus ed. C. H. Oldfather (Loeb, 1960), p. xvii.

⁶ Poseidonios: Strabo 4. 1. 7; 4. 1. 13; 4. 4. 5; 4. 4. 6. Asinius Pollio: 4. 4. 3; Aristotle: 4. 1. 7; Ephorus: 4. 4. 6; Caesar: 4. 1. 1; Polybius: 4. 1. 8; 4. 2. 1; Timaeus: 4. 1. 8; Artemidorus: 4. 1. 8; 4. 1. 1; 4. 4. 6; Pytheas: 4. 4. 1; Aeschylus: 4. 1. 7. Even in passages using material certainly drawn from Poseidonios, eg. 4. 4. 3, Strabo incorporated information from other sources as well. For instance, what he said at 4. 4. 3 about the choice of leaders and generals may be argued to be a reference to Caesar's comments on the Aedui (BG 1. 16. 5; 7. 32. 3) and Vercingetorix's acclamation (BG 7. 63. 4–6). It certainly cannot be proved to be from Poseidonios.

⁷ Jacoby, FGH 11c p. 220.

⁸ Ammianus Marcellinus xv. 9. 2–8, and xv. 12, drawing on Timagenes, has many detailed points of similarity with the minimum reconstruction of Poseidonios; cf. Jacoby FGH 11c p. 225 paragraph 5 and p. 227 Fr. 11.

⁹ Diodorus v. 31. 1; Strabo 4. 4. 2; 4. 4. 3; 4. 4. 5; 4. 2. 3; Ammianus Marcellinus xv. 12.

¹⁰ Diodorus does not mention the Aedui by name, although there is no other people he could be referring to. The fact that both use the same terms σνγγέϑ∊ια, Φιλíα (Diod.), σνγγ∊ϑῖς Φιíα (Strabo) is not proof of a common source as these are the obvious and appropriate technical words to use on this subject.

¹¹ Livy, per. 137, to which my attention was drawn by Miss B. M. Levick; Tacitus, Ann. 11. 25. The Aeduan Diviciacus visited Rome and was known to Cicero: De Div. 1. xli. 90.

¹² Strabo 4. 1. 6–14.

¹³ Strabo 4. 3. 4.

¹⁴ Strabo 4. 2. 3; 4. 3. 4. It might be argued that this, too, was due to Poseidonios and that therefore the Arverni then lived on the Loire. But this argument would be weak for the same reasons. The Allier was known as distinct from the Loire by Strabo's time (Caesar, BG 7. 34. 2; 7. 35. 2; 7. 53. 4) and Nemossus, which Strabo puts on the Loire, is surely Clermont-Ferrand (Augustonemetum) on the Allier. Tierney (p. 211) noted that Strabo put the source of the Seine in the Alps and suggested that he was therefore following an early source since the Romans would have known the truth. Surely in all these cases Strabo himself could have been responsible for the errors, all made in the same way. If the orbis pictus of Agrippa was as distorted for Gaul as is the Table of Peutinger, he could perhaps have been misled by a bad map (Desjardins, E., Géographie de la Gaule romaine (Paris, 1896) iii, 482).Google Scholar

¹⁵ A good example of the outcome of this position is to be found on p. 213: ‘In the Gallic ethnography, moreover, where he [Caesar] had a fully articulated whole from which to borrow, he has succeeded in making the result look nearly… incoherent. The geography as such is simply omitted, except for his introductory chapter (i, 1) and for the numerous small points appearing at suitable places throughout his history. Nearly all the ethnography is similarly omitted….’ It is particularly difficult to defend finding the hand of Poseidonios behind the scattered remarks in Caesar's narrative which arise out of the immediate context.

¹⁶ The fact that Caesar does not himself comment on his contrast with Poseidonios's account may either indicate his ignorance of it in any detail (unlikely, perhaps, considering its subject-matter), or suggest that Poseidonios's work was not the all-important tract we now suppose it to have been, and that therefore Caesar was under no obligation to take note of it when describing his campaigns in a completely different part of Gaul.

¹⁷ Polybius 2. 17. 12; Caesar, BG 6. 13. 2. It is difficult to justify Tierney's argument, quoting Jacoby, that Poseidonios ‘must’ have gone into more detail than we have evidence for (pp. 203, 206). This assertion is at most thinly grounded in what Poseidonios's method is thought to have been (cf. note 36 below). None of the redactions of Poseidonios's work mentions clientage as such, and it still remains fact that the only systematic treatment of Celtic society which survives is in Caesar; it is perverse to ascribe all that is best in it to another author.

¹⁸ It may be added that the Romans were in an unusually good position to recognize and describe clientage as they still had elements of such an institution in their own background. Another case where Tierney sought to find evidence of literary borrowing through the use of ‘motifs’ is at p. 217. Because both Caesar on the Germans and Ephorus on the Cretans praised hard work and commented both on the absence of private property and on the redistribution of land, he hinted at a direct connection between them. Surely this stretches credulity.

¹⁹ The Celts' greed for wealth was notorious in the ancient world and numerous stories are told about it from a very early date, e.g. Livy 5. 48; 44. 26; Polybius 1. 66 f.; 2. 17. 11; 4. 46. 3; Polyaenus 4. 17; ?Poseidonios via Strabo 4. 4. 4; Diodorus v. 27. 3–4. Caesar observed bribery in operation at a public and private level and can have needed no textbook to notice it (BG 7. 63. 2; 7. 37. 1; 1. 9. 3; 5. 55. 1; 6. 2. 1–2; 7. 64. 8). Wealth played an indispensable part in the operation of clientage and in mercenary service (e.g. Polybius 2. 22. 2), and can scarcely be called a ‘motif’.

²⁰ No real argument is offered for this statement, only that there is ‘every reason’ to think that it was from Poseidonios. Diodorus v. 25. 5 leaves open the question as to how much his source (Poseidonios) said about rivers, and which rivers. Clearly he left out a sizeable chunk, but it begs the question to fill in Diodorus's gap with most of Strabo's information.

²¹ Desjardins loc. cit. (note 14). Tierney, J. J., ‘The Map of Agrippa’, P.R.I.A. 63 (1963), 151–66.Google Scholar

²² Strabo 4. 1. 1.

²³ Strabo 4. 4. 3.

²⁴ Strabo made some curious mistakes about the Belgae. He assumed that the Armoricans were Belgae when they were certainly not (Strabo 4. 4. 1.), though by calling these ‘Belgae’ ༔θνη τῶν παρωκ∊αντῶν he echoes Caesar's epithet for the Armoricans (BG 7. 75. 4; 8. 31. 4), and both are rough translations of the Celtic words behind the name Armorican. This could be yet another case of Strabo's misunderstanding of his material as he described the actual Belgae (4. 3. 4–5) without calling them by that name. However, the fact that Strabo thought of the Belgae as primarily coastal peoples (4. 4. 1; 4. 4. 3) might be significant and directly or indirectly due to a much earlier source than Caesar such as Pytheas, whom Poseidonios used for the Atlantic coast. It is just possible that the coastal peoples on the north-west of Gaul around the Seine and Somme mouths had a common name at least so far as foreigners were concerned. Coin evidence of the third century suggests that they had a peculiar development and seaborne relations with the Mediterranean (Scheers, S., ‘Le premier monnayage des Ambiani’, Revue belge de Numismatique cxiv (1968), 45–73)Google Scholar, and Hawkes, C. F. C. (‘Celtes, Gaulois, Germains, Beiges’, Celticum 12 (1965), 1–9Google Scholar and ‘New Thoughts on the Belgae’, Antiquity xlii (1968), 6–16) suggested grounds on which the Belgae may be recognised as a distinct entity from an early period in the north-west of Gaul. Strabo's error with the Armoricans could then be due to mis-application of the name Belgae to these other people on the coastal route between the Mediterranean and north-western Europe, though it is sometimes a waste of energy to try to save Strabo's veracity.Google Scholar

²⁵ Such details are Caesar's account of the origins of the Belgae and their resistance to the Cimbri and Teutones (BG 2. 4), which cannot on present evidence be shown to come from a source earlier than Caesar and certainly cannot be attributed to Poseidonios without confirmation.

²⁹ Some source used by Strabo, which cannot on present evidence be called Poseidonios, mentioned some Belgae around the Rhine and Alps (4. 1. 1). Perhaps it was Asinius Pollio (cf. 4. 5. 3), although Strabo may be responsible for introducing the Alps into the description.

²⁷ Tierney pp. 200, 216.

²⁸ Tierney p. 212, cf. Suidas s. Poseidonios; Strabo 14. 2. 13; [Lucian] Macrobius 20; Cicero, Att. 2. 1. 2; Athenaeus iv. 36. p. 151E and ix. p. 369c.

²⁹ Tierney p. 206 n 87. Poseidonios at Massilia: Diodorus 4. 20. 2–3. Cf. Gallia 1972, 519 where nailed severed heads of similar kind to those found at Entremont were recorded as being found at Les Pennes-Mirabeau 15 km from Marseille. Poseidonios may have provided some of Strabo's information about Gallia Narbonensis (Strabo 4. 1. 3–14) but he by no means provided all of it, as Strabo mentions ‘other’ authors (4. 1. 3).

³⁰ D. Nash, ‘The chronology of Celtic coinage in Gaul: the Arvernian “Hegemony” reconsidered’, NC 1975, 212 f. Strabo 4. 2. 3 implies that the information was hearsay from the Arvernians themselves.

³¹ Artemidorus, on whom Poseidonios draws, seems to have known the Rhone valley: Strabo 4. 1. 11.

³² Strabo 4. 1. 13.

³³ Strabo 4. 2. 1.

³⁴ Thompson, E. A., The Early Germans (Oxford, 1965).Google Scholar

³⁵ See note 15 above. In this case even the order chosen for Caesar's account is treated as a garbling of Poseidonios's text.

³⁶ Diodorus v. 29; Athenaeus iv. 36 p. 151 E–152D, iv. 37 p.152 D–F. Tierney p. 214: ‘In ch. 13 [BG 6] the word plebes denotes a lacuna with respect to the original.’ There is no clear evidence that Poseidonios in fact gave a systematic treatment of Celtic social structure rather than an anecdotal account of Celtic customs in general. Yet on p. 203 Tierney held that Poseidonios made distinctions between companions and bards that were ‘presumably’ accurate. If so, Poseidonios would be virtually unique among ancient historians in his use of accurate terminology. The acid words of A. E. Housman could well be applied to the pursuit of Poseidonios in the text of Caesar: ‘The sacred name of Posidonius, if I remember right, is not once mentioned in my notes; and when I come to 11 96 I shall not pretend that Manilius, or Catullus either, imbibed from the manuals of that Rhodian sage the daring theory that the moon's light is borrowed. If anyone is enamoured of speculations for which no material exists, he is welcome to pursue them, atque idem iungat volpes et mulgeat hircos; but the lot is fallen unto me in a fair ground, I have a goodly heritage, and I leave the sands of the sea to be ploughed by others’ (Housman, A. E., M. Manilii Astronomicon liber primus (London, 1903), pp. lxxii–iii).Google Scholar

³⁷ Greene, D., ‘Early Irish Society’, in ed. Dillon, M., Early Irish Society (Dublin, 1954, repr. 1963), 22–35Google Scholar; Jackson, K. H., The Oldest Irish Tradition: a Window on the Iron Age (Cambridge, 1964), 23Google Scholar f. Cf. Maquet, J. J., The Premise of Inequality in Ruanda (Oxford, 1970), 103, 124 on the functions of their ‘learned class’.Google Scholar

³⁸ Caesar, BG 6. 13. 4–6. 14. 6; Diodorus v. 31. 2–5, cf. Appian, Celt. 12, which is not proved to be from Poseidonios. Druids were mentioned by a Mediterranean writer as early as 200 B.c: ap. Diogenes Laertius, Vit. Philosoph. proem. 1, pp. 6 ff. (Tierney p. 197).

³⁹ Diodorus v. 20. 2–4.

⁴⁰ A distinction must probably be made in Poseidonios's observations between the local Celts he observed directly, such as those at Entremont who had major settlements with substantially constructed buildings by the early first century, and the Aedui, Arverni and other more northerly Celts whom Poseidonios was describing in the context of 125–1, when they had no such institutions.

⁴¹ Jackson op. cit. (note 37). In some respects, early Ireland and Gaul in Poseidonios's day were not strictly comparable, for instance in that the Irish never developed a coinage. But it may still be argued that in political structure there were important similarities.

⁴² E.g. Aedui (Caesar, BG 1. 15–18, 7. 32); Arverni (BG 7. 4. 1–3); Helvetii (BG 1. 2–5).

⁴³ Cf. Cicero, de Div. 1. 90; cf. note 38 above.

⁴⁴ Sherwin-White, A. N., Racial Prejudice in Imperial Rome (Cambridge, 1967), 10 f.Google Scholar

⁴⁵ Ceasar, BG 6. 13.

⁴⁶ Especially the immunity from tribute, for instance in Ruanda (Maquet, op. cit. (note 37), 124), or their exceptional mobility between the territories of the peoples (Greene, op. cit. (note 37)).

⁴⁷ Tierney pp. 214, 222 f.

⁴⁸ Caesar, BG 7. 33. 3.

⁴⁹ Nock, A. D., ‘The Roman army and the Roman religious year’, in Stewart, Z. (ed.), Arthur Darby Nock: Essays on Religion and the Ancient world (Oxford, 1972), 759–60. In fact by the early principate the most savage practices had been reduced to symbolism, according to Pomponius Mela 3. 2. 18.Google Scholar

⁵⁰ It was not difficult to make the leap from the existence of a similar institution in all the Gaulish or Irish peoples or tribes, especially when it had the special mobility of the learned class, to the conclusion that it was a centrally organized system capable of conspiracy.

⁵¹ Diodorus v. 31. 1; Caesar, BG 4. 5. 3; Strabo 4. 4. 5.

⁵² Jackson, loc. cit. (note 37); Greene, op. cit. (note 37), 86 f.

⁵³ Holmes, T. Rice, Caesar's Conquest of Gaul (Oxford, 1911), 831.Google Scholar