Published online by Cambridge University Press: 24 September 2012
When in the autumn of 44 Octavian embarked on his perilous journey as heir to Caesar's mantle he and his soldiers were branded as brigands or extolled as saviours of their country. What was their legal status ? Most modern historians, fascinated with naked power, tacitly dismiss this question as utterly frivolous. They are fully satisfied with Ciceronian epithets. Octavian and Cicero were not. The Roman doctrine of the genera militiae afforded Octavian a comfortable legal niche at each stage of his career. It allows us to comprehend the intricate manoeuvres in the senate at the turn of 44 and 43 B.C.; it also sheds light on the crowning coup of Octavian when he led Italy as a dux against Antonius and the Queen. But this antiquarian doctrine, attested only in late authors, could easily be dismissed as an artificial construct far removed from real life. Historical puzzles lie dormant and insoluble until a spark of insight creates an instant enlightenment. And the spark comes from Aphrodisias, from the new documents published in an exemplary way by Joyce Reynolds, Aphrodisias and Rome (1982). It comes in the shape of a new puzzle.
1 Watson, G. R., The Roman Soldier (1969), 43Google Scholar, and R. W.Davies, ‘Joining the Roman Army’, BJ 169 (1969), 214, take this phrase to refer to rolls or records. But in a fundamental article Gilliam, J. F., ‘Enrolment in the Roman Imperial Army’, Eos 48, 2 (1956), 212,Google Scholar perspicaciously observed that in numeros distribuere (or per numeros distribuere in Trajan's reply) ‘is not equivalent to in numeros referre’. In the latter phrase numeri means rolls, but in the former probably units. The tirones, Gilliam suggests, ‘were still at some recruiting center and had not yet been divided among or at any rate dispatched to units’.
2 The Letters of Pliny. A Historical and Social Commentary (1966), 601.
3 As H. H. Scullard writes, Scipio's preparations for his expedition to Africa are ‘shrouded in doubt’ (Scipio Africanus (1970), III). For the story itself, see Passerini, A., Le coorti pretorie (1939), 6 ffGoogle Scholar.
4 The custom of providing vicarii may have come into being during the civil wars, but it is doubtful if any conclusions can be drawn from the story in Macrobius 2. 4. 27 (brilliantly elucidated by Cichorius, C., Römische Studien (1922), 282–5)CrossRefGoogle Scholar: ‘exclamavit ingenti voce veteranus: at non ego, Caesar, periclitante te Actiaco bello vicarium quaesivi sed pro te ipse pugnavi’.
5 On the cohortes voluntariorum, see Kraft, K., Zur Rekrutierung der Alen und Kohorten an Rhein und Donau (1951), 82–95Google Scholar, who criticizes the idea of Cheesman, G. L., The Auxilia of the Roman Imperial Army (1914), 65–7, 186–7Google Scholar, that the majority of these units were the cohortes libertinorum. The inscription of the ‘captor of Decebalus’ clearly supports Kraft's view; cf. Speidel, M., JRS 60 (1970), 151Google Scholar.
6 The Greek sense of ἀντιστρατιώτης is not in dispute: ‘soldier of the enemy’, as LSJ duly records. The Roman development thus parallels that of ἀντιστράτηγος from ‘enemy's general’ to ‘acting commander’ or ‘governor’. But ‘acting soldier’ does not take us very far: see below in the text.
7 cf. C. Atzert in his Teubner edition, pp. XXVII–XXIX.
8 In any case Cato will not have written two identical letters, one in 173 when his son served as a tiro in Liguria under the consul M. Popillius Laenas, and the other in 168 when Licinianus was in the army of Aemilius Paullus and took part in the battle of Pydna. Cf. Drumann-Groebe, , Geschichte Roms v, 160–1Google Scholar.
9 Clearly a later addition reflecting imperial practice. Smith, R. E., Service in the Post-Marian Roman Army (1958), 29–33Google Scholar, rightly observes that until Augustus the length of service was not explicitly specified and the sacramentum contained no reference to it.
10 cf. Bell. Alex. 56. 4; Tac., Ann. 16. 13. 3; Hist. 1. 5. 1; Flor., 1. 22. 23; Iuv., 16. 35–6.
11 Römische Forschungen II (1879), 247–57;Google ScholarEph. Ep. 5 (1885), 142–5 = Ges. Schr. VIII, 446–9; StR I3 (1887), 695–6.
12 ‘Die Evocati’, Hermes 14 (1879), 322–31Google Scholar.
13 ‘Il sacramentum militiae nell'ambiente culturale romano-italico’, SDHI 29 (1963), 1–25.
14 ‘Coniuratio’, Jahrb. f. Numismatik u. Geldgeschichte 13 (1963), 51–70Google Scholar.
15 On dilectus, see the brilliant exposition by Brunt, P. A., Italian Manpower (1971), 625–44Google Scholar. But cf. also the objections by Rawson, E., PBSR 39 (1971), 15 ffGoogle Scholar.
16 It is important to keep apart the length of the legal obligation to serve and the length of the actual service. Only the soldiers whose stipendia were emerita or confecta had a legal claim to a missio; cf. Livy 34. 56. 9; 39. 19. 4; 39. 38. 12; 40. 35. 11; 43. 14. 9, and the passages adduced by Smith, Service, 35 n. 3, whose illuminating discussion (27 ff.) dispersed many common misconceptions. See also Brunt, , JRS 52 (1962), 80–2;Google ScholarManpower, 399 ff.; Harmand, J., L'armée et le soldat à Rome de 107 à 50 avant notre ère (1967), 245 ffGoogle Scholar. The idea of the twenty legitima stipendia in the late Republic has no source authority.
17 And moreover this imperium-holder had to be authorized by the senate (or the people) to hold the levy. Mommsen believed (StR I3, 119) that the magistrate cum imperio did not need any permission from the senate for dilectus, but see the convincing critique of this theory by Brunt, ZPE 13 (1974), 162 ff. The antiquarians connect coniuratio with tumultus, but in the annalistic tradition tumultuary levies are as a rule ordered by the senate, and the milites tumultuarii are often sacramento rogati, cf. esp. Livy 32. 26. 10–12; 40. 26. 7; 41. 5. 11 (in conjunction with 41.5. 4). We can put forth the following scheme: the regular dilectus was always accompanied by sacramentum; the tumultuary dilectus was accompanied by a sacramentum when it was conducted on express orders from the senate by a magistrate cum imperio. The coniuratio comes into the picture when a magistrate acted on his own initiative or when the dux was a privatus.
18 cf. O. Fiebiger, RE 6 (1909), 1146; Neumann, A., Der Kleine Pauly 2 (1975), 471Google Scholar.
19 For sources and discussion, see Botermann, H., Die Soldaten und die römische Politik in der Zeit von Caesars Tod bis zur Begründung des zweiten Triumvirats (1968), 36 ffGoogle Scholar.
20 Appian (BC 3. 40) reports that the veterans collected by Octavian in Campania marched ὑφ'ἑνἰ σημείῳ, under one vexillum: a fair description of a manus tumultuaria.
21 On 4 November 44 Cicero wrote (Att. 16. 9): Octavian ‘rem gerit palam, centuriat Capuae, dinumerat’. A. Alföldi, Oktavians Aufstieg zur Macht (1976), 108 n. 401, maintains that ‘centuriat Capuae’ does not refer to the formation of military units at all: ‘Centuriare und dinumerare sind Ausdrücke der stadtrömischen Wahlbestechung’. This is true of decuriatio (cf. Linderski, J., Hermes 89 (1961), 106Google Scholar ff.), but centuriare is not attested in this sense. The explanation in OLD (s.v. ‘dinumero’): he ‘is giving the soldiers their pay’ is marred by ‘pay’, which introduces a wrong emphasis. Bailey's, D. R. Shackleton rendering is exemplary (Cicero's Letters to Atticus vi, 189)Google Scholar: ‘he's … forming companies at Capua and paying out bounties’.
22 cf. Livy 22. 28. 1: ‘inter sese decuriati equites, centuriati pedites coniurabant’; Caes., BC 1. 76. 3: ‘centuriatim producti milites idem iurant’.
23 On the evocati under the empire, in addition to the works of Mommsen, Schmidt and Fiebiger quoted in notes 11, 12, and 18, see Domaszewski, A. v., Die Rangordnung des römischen Heeres (1908Google Scholar; 2nd ed. by B. Dobson, 1967), 75–8; Durry, M., Les cohortes prétoriennes (1938), 117–26Google Scholar; E. Birley, ZPE 43 (1981), 25–9.
24 cf. Syme, R., Roman Revolution, 160–1Google Scholar; Helleguarc'h, J., Le vocabulaire latin des relations et des partis politiques sous la république (1963), 95–7, 123–5Google Scholar.
25 Brunt, , JRS 52 (1962), 81,Google Scholar reads (through a lapsus calami) ‘bello confectae’ (sc. legiones), which of course affected his argument.
26 Brunt, P. A. and Moore, J. M., Res Gestae Divi Augusti (1967), 38–9Google Scholar, opt for I January, but see the discussion by Stein, P., Die Senatssitzungen der ciceronischen Zeit (Diss. Münster, 1930), 80–3Google Scholar.
27 App., BC 3. 65; cf. Cic., Phil. 14. 26–7; Fam. 10. 30. 1; 11. 19. 1.
28 Botermann, , Die Soldaten, 42, 202–3Google Scholar; , Brunt, Manpower, 481–2Google Scholar.
29 Livy, Per. 85; Plut., Pomp. 6. 3–4.
30 Kromayer, J., Die rechtliche Begründung des Prinzipats (1888), 16 ffGoogle Scholar.; Syme, Rom. Rev., 284 ff.; Premerstein, A. v., Vom Werden und Wesen des Prinzipats (Abh. Münch., 1937), 36Google Scholar ff.; Herrmann, P., Der römische Kaisereid (1969), 78 ffGoogle Scholar. (cf. J. Briscoe, CR 21 (1971), 260–3); Brunt-Moore (above, n. 26), 67–8; Fadinger, V., Die Begründung des Prinzipats (1969), 18 ffGoogle Scholar.; Benario, H., Chiron 5 (1975), 301–9Google Scholar.
31 Livy 2. 32. 1; 3. 20. 3–5. See also 28. 29. 12, where the phrase in verba iurare refers (as follows from 28. 27. 4 and 12) to the renewal of sacramentum. The iurare in verba and dilectus: 6. 2. 6; 22. 11. 9. Cf. also 7. 16. 8; 45. 2. 10. At 22. 53. 12 the phrase does not refer to sacramentum, but to an oath of the type ‘ut victor revertar’, and in four other passages it appears in a non-military context (6. 22. 7; 7. 5. 5; 32. 5. 4; 41. 15. 11). Cf. Herrmann, (above, n. 30), 42 n. 75; 81 n. 89. On Tac., Ann. 1. 7. 2 (the oath in verba Tiberii Caesaris), see the judicious remarks by Goodyear, F. R. D., The Annals of Tacitus I (1972), 138–9Google Scholar.