Hostname: page-component-586b7cd67f-r5fsc Total loading time: 0 Render date: 2024-11-24T09:16:10.506Z Has data issue: false hasContentIssue false
  • English
  • Français

Arrian at the Caspian Gates: a Study in Methodology

Published online by Cambridge University Press:  11 February 2009

A. B. Bosworth
A. B. Bosworth
Affiliation:
University of Western Australia
Get access Rights & Permissions [Opens in a new window]

Extract

In a recent article Professor Brunt has made an eloquent plea for greater rigour in handling the remains of non-extant authors. When the original is lost and we depend I upon quotation, paraphrase or mere citation by later authorities, we must first establish the reliability of the source which supplies the fragment. There is obviously a world of difference between the long verbal quotations in Athenaeus and the disjointed epitomes provided by the periochae of Livy. As a general rule, the fuller and more explicit the reproduction in the secondary source, the more confident we can be that it approximates to the original. Our doubts should increase as the references become less precise and resort to paraphrase rather than direct quotation. The wider context is also important. One always needs to know why the secondary author is making his citation and what interest he has in a strictly literal reproduction. These principles are unexceptionable, but they are difficult to maintain in practice. One rarely has the opportunity to make a sustained experiment, checking an author's techniques' of quotation and digest against sources which are now extant. As a result the historian all too often feels constrained to squeeze the last drop of meaning out of testimonia which are by their very nature imprecise. The standard work, Jacoby's Fragmente der griechischen Historiker, is a pitfall for the unwary. It presents all citations, whether full quotations or the vaguest of references, on the same status as ‘fragments’; and the context is necessarily reduced to the barest minimum, so that the reader's attention is focussed directly upon the lost original and diverted from the machinery of transmission.

Type
Research Article
Information
The Classical Quarterly , Volume 33 , Issue 1 , January 1983 , pp. 265 - 276
Copyright
Copyright © The Classical Association 1983

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 Brunt, P. A., ‘On historical fragments and epitomes’, CQ 30 (1980), 477–94CrossRefGoogle Scholar.

2 Wirth, Gerhard, ‘Arrian und Traian -Versuch einer Gegenwartsdeutung’, Studii Clasic 16 (1974) 169–209, esp. 189 f.Google Scholar; Wheeler, Everett L., Flavius Arrianus: a political and military biography(Diss.Duke University, 1977), pp. 27–31Google Scholar; Stadter, Philip A., Arrian of Nicomedia (Chapel Hill, 1980), pp. 9, 135–43Google Scholar. (These three works will henceforth be referred to by the author's name alone.) The hypothesis had been mooted briefly in the past, most categorically by Domaszewski, Alfred von (SB. Heidelberg 16 (1925/1926), Heft 1, 5)Google Scholar, who argued that Arrian served as an equestrian officer. This evoked brief expressions of dissent; cf. Schwartz, E., RE ii. 1236Google Scholar; Jacoby, F., FGrHist IID. 567, 575Google Scholar; Upper, F. A., Trajan's Parthian War (Oxford, 1948), p. 2Google Scholar. None of these authors, however, questions the postulate that ten books of the Parthica were devoted to Trajan's campaigns.

3 Joh. Lyd. de magistratibus populi romani 3. 48, p. 142. 6. The page references here and subsequently are to the standard Teubner text by R. Wuensch (Leipzig, 1903), and, unless it is otherwise stated, all references to John are to the de magistratibus. This work was begun in A.D. 554. See in general Martindale, J. R., Prosopography of the Later Roman Empire II (Cambridge, 1980), pp. 612–15Google Scholar.

4 τοιο⋯τοσ μ⋯ν οὖν πε περ⋯ τ⋯ν κασπ⋯ων πυλ⋯ν τοῖσ 'Rho;ωμα⋯ων συγγραπε⋯σιν ⋯ λ⋯γος (ὂν) 'Aρριαν⋯ς ⋯π⋯ τ脴ς 'Aλανικ脴ς ἱστορ⋯ας κα⋯ οὐχ ἥκιςτα ⋯π⋯ τ脴ς ⋯λδ⋯ης Παρθικ膂ν ⋯κριβ⋯στερον διεξ⋯ρχεται, αὐτ⋯ς τοῖς τ⋯ποισ ⋯πιστ⋯ς οἱα τ脴ς χώρας αὑτ脴ς ⋯γησ⋯ μενος ὑπ⋯ τραἲαν膃 χρηστ膃. This is the limit of the fragment in Roos', edition (Parth. F 6Google Scholar, Flavii Arriani quae exstant omnia II (revised Wirth, G.: Leipzig, 1967), p. 229)Google Scholar; Jacoby, FGrHist 156 F 37Google Scholar continues it for a sentence (on which see below, p. 270). Wirth, p. 189 n. 79, interprets the flnal clause as an explanation of Arrian's later Cappadocian command: ‘he himself commanded the area, seeing that he had been in charge of that territory (before) under Trajan’ (this is accepted i by Wheeler, p. 30, and less explicitly by Stacker, p. 143). I find this interpretation very forced. In the context ⋯πιστ⋯ς should be understood simply in the sense of proximity, as in the exact parallel from Polybius: ποιησ⋯μενος δ⋯ τ⋯ν πορε⋯αν δι⋯ τ脴ς Δαυν⋯ας… ⋯π⋯στη το腐ς κατ⋯ τ⋯ 'P⋯λιον τ⋯ποις ⋯νυπ⋯πτως (Polyb. 9. 7. 10; see further LSJ 2 s.v. ⋯π⋯στημι B. iii). John's statement now becomes perfectly simple; Arrian's description was more accurate because he had actually visited the scene in his capacity as commander of the area.

5 Photius, , Bibl. cod. 58Google Scholar = Parth. F I Roos: σ avyypaufnTai 8eKOX TO. Kara ‘AXavovs, fy >. eiriypau/ifv ‘AXaviKr/v. The only other possible allusion to the work is a passing reference in Procopius(de bell. Goth. 4. 14. 47 f.; cf. Roos, Mnemosyne 54 (1926), 116). V

6 Steph. Byz. s.v.'EXiyua = Parth. F 5 Roos.

7 Gutschmid, A. von, Philologus 8 (1853), 435–39CrossRefGoogle Scholar. His results served as Roos’ framework for his reconstruction of theParthica (Roos, A. G., Studia Arrianea (Leipzig, 1912), p. 36)Google Scholar, which, was in turn endorsed by Jacoby, (FGrHist IID. 566 f.)Google Scholar. Since that time the identification has been taken as an ‘absolutely fixed point’ (Longden, R. P., JRS 21 (1931), 12 n. 3)Google Scholar.

8 For the dimensions of the History of the Successors see the digest of Photius (cod. 92 =Succ. F 1 Roos) and the brief characterisation by Stadter, pp. 144–52.

9 Historia 14 (1965) 354Google Scholar =Danubian Papers (Bucharest, 1971), p. 237Google Scholar.

10 Eutrop. 8. 2. 2; Festus, , Brev. 20. 2Google Scholar; Jerome, , Chron. p. 194Google Scholar Helm. For digest and discussion see Lepper (above, n. 2), pp. 13–15.

11 For the operations of Lusius Quietus see Themistius, , Oral. 16, p. 294Google Scholar. 21 Downey; Arrian Parth. F 87 Roos, and for those of Bruttius Praesens see Arrian Parth. F 85 Roos (located by Syme, , Historic 18 [1969] 352Google Scholar = Roman Papers (Oxford, 1979), p. 774)Google Scholar.

12 AE 1968, no. 510–11: IV Scythica was present in force in late 116, and there were vexillationes from I Italica andVI Ferrata. See further Chaumont, Marie-Louise, ANRW II. ix. 1 (Berlin, 1976), 137Google Scholar.

13 The error recurs at 1. 34, p. 34. 16 ff., which is a free adaptation of Pomponius (Dig. 1. 2. 2. 24) masquerading as a translation. The passage is again ascribed to Gaius.

14 John claims that Tullus Hostilius made the office of quaestor compulsory, whereas Ulpian (Dig. 1. 12. 1) merely states that it was established fact that Tullus had quaestors. John omits the previous sentence (which evinces doubt whether quaestors existed under Romulus and Numa), so that the text as it stands is unintelligible.

15 1. 32, p. 33. 13 ff.; cf. Persius 1. 73–75. Compare the reference to Lucan at 3. 46, p. 135. 14, where John claims that Dyrrhachium got its name from an eponymous Cretan king and cites the second book of the Pharsalia in support. Lucan says nothing about a Cretan king or the foundation of Dyrrhachium. He merely notes that Brundisium was once occupied by Cretan exiles (Phars. 2. 610); Epidamnus (not named Dyrrhachium) appears fleetingly at line 625.

16 Athenaeus claims that the fish was served at Rome to the accompaniment of pipes and garlands; in John's version the accompaniment is pipes and cymbals.

17 Pliny, HN 9. 61Google Scholar: ‘postea praecipuam auctoritatem fuisse lupo et asellis Nepos Cornelius et Laberius poeta mimorum tradidere.’

18 Pliny, HN 9. 62Google Scholar: ‘inde [sc. from the Carpathian sea] advectos Tiberio Claudio principe Optatus e libertis eius praefectus classis inter Ostiensem et Campaniae oram sparsos disseminavit.’

19 Macrob, . Sat. 3. 16. 10Google Scholar

20 Diod. 1.98. 1 (cf. 77.5,79.4 for other Solonian borrowings). The triple division of Egyptian society comes at 1. 28. 4–5 and is adduced as evidence that Athens was colonised from Sais in Egypt.

21 De mensibus 3. 1, p. 37. 14 = Arrian Parth. F 1 Roos.

22 1. 47, p. 49. 13–20. The list ends with a reference to book 1 of Claudian's extant encomium on Stilicho. Once more there is no reference to veterans in the text, but Claudian (1. 384) does refer in passing to ‘veterum…lauros’, the triumphs of men of old.

23 1. 47, p. 50. 2: τοὺς λεϒομένους Τριβαλλούς οὔτως δ⋯ τοὺς Bέσσους Ἀρριανòς ⋯⋯ τοȋς περì Ἀλεξάνδρου προσηϒόρευσε.

24 In the Roman period the Bessi were the predominant tribe of the Haemus (cf. Strabo 7. 5. 12, 318). They were the descendants of Arrian's autonomous Thracians.

25 Arr. 1. 1. 5–6 (on this passage see my Historical Commentary on Arrian's History of Alexander I (1980), pp. 53 f.)Google Scholar. Alexander only encountered the Triballi after he crossed the Haemus and pressed towards the Danube (Arr. 1. 2. 1).

21 Arr. 1. 1. 4–5: ⋯λαύνων ⋯πἱ Θρᾴκη, Θη Τριβαλλοὺς καì Ἰλλυριούς … δεκαταȋος ⋯ϕίκετο ⋯π⋯ τò ðρος τòν ΑΙμον

27 τοιύτους ϒ⋯ρ ἄρχοντας ⋯κεȋνος ἒσχεν, οȋ τοȋς τε ἒρϒοις εἰς τοσαύτην εὒκλειαν τ⋯⋯ πολιτείαν ⋯⋯⋯στησαν(p. 142. 21–23).

28 Themistius, Oral. 34. 8Google Scholar (T 13 Roos). The passage describes both Arrian and Q. Iunius Rusticus, the mentor of Marcus Aurelius, and conflates their achievements. The two careers, however, can easily be disentangled. For discussion of the context and its historical implications see HSCP 81 (1977), 229–32Google Scholar.

29 Pompey passed through the centre of Iberia in his campaign of 65 B.C., forcing the pass at *. Mtshkheta, immediately south of the Caspian Gates (cf. Dio 37. 1. 4–5). For Crassus' invasion of 36 B.C. see Plut, . Ant. 34. 10Google Scholar; Dio 49. 24. 1. Strabo 11. 3. 5, 501, states that he used the same k, passes as Pompey.

30 Tacitus, (Ann. 6. 33. 3)Google Scholar mentions both the Darial pass and the coastal pass of Derbent (cf. Jos, AJ 18. 97)Google Scholar.

31 Pliny, HN 6. 40Google Scholar; Tac, . Hist. 1. 6. 4Google Scholar; Dio 63. 8. 1; Suet, . Nero 19. 2Google Scholar. There is controversy about Nero's intentions (for reviews of the problem see Wheeler, pp. 117–23; Chilver, G. E. F., Historical Commentary on Tacitus’ Histories I and II (Oxford, 1979), pp. 55 f.)Google Scholar; but all sources agree that the Caspian Gates were his geographical objective.

32 For the evidence see Bosworth, , Antichthon 10 (1976), 73–76CrossRefGoogle Scholar; Wheeler, pp. 131 ff. It is significant that Statius (Silv. 4. 4. 63 f.) regards the Caspian Gates as a natural arena for the operations of the Cappadocian army.

33 He gave a detailed account of the interview, apparently including speeches by both Trajan and Parthamasiris (Parth. F 37–40 Roos; cf. Roos, , Studia Arrianea, p. 36Google Scholar; Stadter, p. 139).

34 Dio 71. 2. 1; cf. Lucian, Alexander 27Google Scholar; de hist, corner. 21, 25 f.; Groag, , RE iiA. 1009–10Google Scholar (‘auf einem in der Geschichte Armeniens oft genannten Schlachtfelde’).

35 Pliny, HN 5. 84Google Scholar (the context is difficult to interpret, but Elegeia is clearly placed on the Euphrates); Ptolemy, Geogr. 5. 13. 12Google Scholar Nobbe.

36 The favourite choice is Ilica, between modern Erzurum and Askale (cf. Baumgartner, RE v. 2258; Mitford, T. B., ANRW 11. vii. 2 (Berlin, 1980), 1198 f.)Google Scholar. Magie, D., Roman Rule in Asia Minor ii. 1465Google Scholar, was rightly sceptical.

37 Note Trajan's line of march in A.D. 114 (Dio 68. 19. 5; cf. Lepper [above, n. 2], pp. 7, 207; Chaumont [above, n. 12], p. 134).

38 Mitford (above, n. 36), p. 1176, ‘perhaps near Erzurum, but more plausibly at Satala'. Chaumont, pp. 100–1, argues that Corbulo spent the winter of A.D. 57/8 in the plateau of Erzurum after moving from Cappadocia. In that case he will have passed by or near Elegeia.

39 Tac, . Ann. 13. 39. 1–5Google Scholar. Volandum is often identified with the fortress of Olane (cf. Chaumont, p. 103 n. 171), which Strabo (11. 14. 6, 529) locates in the vicinity of Artaxata. If so, Tacitus has no t given a single place nam e between the Cappadocian border and the Armenian capital.

40 Pliny's reference t o Elegeia occurs in a context in which he cites Corbulo an d C. Licinius Mucianus as sources for the geography of the Euphrates, (HN 5. 83Google Scholar = Corbulo F 3 Peter).

41 Thuc. 8. 60. 1–2, 95. 1, 3–4. There are other passing references to Oropus a t 3. 91. 3, 4. 96. 7 and 7. 28. 1. The reference t o Sollium is also to its first appearance in Thucydides (2. 30. 1; cf. 3. 95. 1, 5. 30. 2). Unfortunately Stephanus is no t consistent. There is n o apparent reason why the reference to Anactorium is t o Thucydides’ account of its capture by the Athenians in 424 (Thuc. 4. 49) an d no t t o the similar note in boo k I recording its capture an d resettlement by the Corinthians (1. 55. 1; cf. also 2. 80. 3, 3. 114. 3, 5. 30. 2, 7. 31. 2).

42 444. 18,470. 13, 564. 1 (these and subsequent references are to pages and lines of Meineke's edition (Berlin, 1849; repr. Graz, 1958)). The most frequent corruption is ις or ⋯κκαιδκάτψ (9. 8, 239. 16, 278. 19, 296. 11, 308. 1, 355. 6). At 445. 3 the reading is ι and at 477. 13 there is a variant ς.

43 Book XVI: right – 111. 6, 368. 9, 477. 3, 594. 8; wrong – 20. 19, 224. 1, 250. 14, 378. 11, 433.2,476. 12, 574. 6 (variants in two manuscripts). Book V: right – 115. 17,671. 19;wrong – 564. 16. 712. 19.

44 83. 8(ις for ις); 448. 3 (ς for ⋯).

45 Parth. F 8–9 Roos. For Phalga see Isidore of Charax, FGrHist 781 F 2. (i) (фάλιϒα), and for Choche Amm. Marc. 24. 5. 3, 6. 2. Similarly Naarda, mentioned in Book XI (F 10 Roos), was a rich and flourishing city of south Babylonia, the scene of a Jewish uprising under Artabanus III (Jos, . AJ 18 311 ff., 369)Google Scholar. It could have been mentioned in almost any context, as could the obscure fortress of Thebetha (F 11 Roos: cf. Lepper [above, n. 2], pp. 124 f.).

46 καì βασιλεὺς ς ⋯ξελαύνει ⋯κ Σελευκείας οὐ πρόσω τοȗ Τίϒριδος ⋯ς κώμην ᾑ τινι Χωχ⋯ ðνομα (F 10).

47 F 10–14 Roos. For Trajan's operations in Adiabene see Lepper. pp. 9–10, 129–36.

48 Note for instance Josephus’ interest in the principality during the reigns of Izates and Monobazus, (AJ 20. 17 ff)Google Scholar. Tigranes ravaged the territory extensively in A.D. 62 (Tac, .Ann. 15. 1. 2)Google Scholar, and the consequence was a Parthian threat to the frontier, Euphrates (Ann. 15. 5. 2–4)Google Scholar, which is a possible context for the reference to Olbia and the satrapy of Chazene (F 14 Roos). I am not arguing that book XIII did cover the events of 62, merely that Trajan's campaign is not the only possible home for the fragments.

49 F 16 Roos; cf. Dio 68. 28. 4.

50 Schwartz, , RE ii. 1236Google Scholar; Jacoby, , FGrHist 11D. 567Google Scholar. See the strictures of Stadter, p. 142.

51 οἱ ϒράψαντες οὐ μόνον τ⋯ ἒρϒα ΤραΪανοȗ ⋯λλ⋯ καì τ⋯ βουλεύματα (F 73 Roos).

52 F 46 (not explicitly attributed to Arrian), 60–62,85. These fragments were among a number amassed by Wirth, (Historia 13 (1964), 228Google Scholar; Phiologus 107 (1963), 288–300)Google Scholar, and they are accepted by Wheeler (pp. 30 f.) as proof of autopsy. Stadter, pp. 143 f., is more cautious: ‘ certain fragments … suggest autopsy without guaranteeing it.’