Appendix 3:
Defeats against Hannibal

De Sanctis, Storia, 3.² 2: 445 n. 28, argues that the two Roman defeats Livy records at Herdonia in 212 and 210—one of the praetor Cn. Fulvius Flaccus, the other of the proconsul Cn. Fulvius Centumalus, both against Hannibal—are in reality one and the same. He alleges that Livy, or the annalist he followed, found the battle dated to 212 in one source and to 210 in another and so mistakenly reported these as two distinct events. Crake challenged this view in his 1939 Ph.D. dissertation, "Archival Material in Livy," as did, independently, Toynbee, Hannibal 'sLegacy, 2:48. More recently, however, Brunt, Italian Manpower, 652, has accepted De Sanctis' arguments and rejected Flaccus' defeat in 212. This position is mistaken, however.

De Sanctis bases his case first of all on the fact that the commanders' nomina and praenomina are the same: Cn. Fulvius in each case. But as Toynbee points out, this resemblance is not so remarkable given the peculiarities of Roman nomenclature. Of greater weight is the fact that the far more distinctive cognomina are different—Flaccus and Centumalus—and securely attested as belonging to two different branches of the Fulvian gens. The fact that both battles took place at Herdonia, which De Sanctis likewise finds suspicious, Toynbee rightly explains by stressing the high military importance both sides attached to control of the region as well as the exposure of the legions to attack in open country that any attempt to seize the place entailed.

Crake (174, 187-88 n. 17) and Toynbee also emphasize the difference in the fates of the two commanders: Centumalus was killed; Flaccus survived to face trial on his return. De Sanctis regards the death of Centumalus as a fundamental point in favor of accepting the validity of his defeat and rejecting Flaccus', but, as

Crake argues convincingly, it is difficult to understand the charge against the latter apart from the defeat. De Sanctis assumes he was tried on other grounds and cites Val. Max. 8.4.3. But this source offers no support for the notion that Flaccus was not tried in connection with his defeat. In fact, this person may not even be the praetor of 212: he went into exile at Tarquinii to escape certain condemnation (Livy 26.3.12), whereas the Flaccus of Val. Max. 8.4.3 was found guilty of the charges against him. Crake (332 n. 26) also points out, in support of the reality of the defeat in 212, that it is difficult to account for Hannibal's ability to march on Rome in 211 if Flaccus' two legions had not been destroyed in the previous year.

Yet De Sanctis' strongest evidence for the alleged doublet is the parallels he finds between the details of the two events: in each the principal causes are negligence on the part of the commander and indiscipline among the soldiers; both involved rash decisions to attack Hannibal head-on; in both battles the left ala was placed in front together with a legion; and a cavalry envelopment in the Romans' rear decided each battle. But De Sanctis is largely mistaken, and closer examination turns up few such parallels. Only in 212 is the indiscipline of the soldiers a prominent cause of the defeat (Livy 25.20.6-7, 21.1-2, 21.5-8). On the contrary, in 210 the Roman soldiers fought steadfastly and with great determination (Livy 27.1.10-11), with no sign of indiscipline in the ranks. However, the carelessness and foolishness of Centumalus are highlighted in Livy's account of the causes of the disaster in 210 (27.1.4, 7), whereas Flaccus' faults largely center on his failure to enforce discipline and the fact that he corrupted his soldiers (Livy 25.20.7, 21.5, cf. 26.2.8, 11-16; but cf. 25.21.9, 26.2.15). The cavalry envelopment occurred only in 210 (Livy 27.1.8, 11); in 212 the Romans ran before they even made contact with the enemy (Livy 25.21.8, cf. 5; lightly armed soldiers previously hidden in ambush did attack the Romans from behind [Livy 25.21.10, cf. 3], but the role of the cavalry was to pursue the fugitives [Livy 25.21.4]). Both battles involved set-piece confrontations simply because it was normally how they were fought. Placing the left ala in the front lines hardly signifies much since there were only two to choose from. Note by way of contrast that the legionary numbers in each case were different—the first legion in 212; the fifth in 210 (Livy 25.21.6; 27.1.8). Altogether, the assertion of parallelism in the two accounts lacks substance, and each should be considered a distinct event.