The Battle of Cannae, on August 2, 216 BCE, pitted genius against overwhelming odds, brains vs. brawn, a master tactician facing overconfident commanders. It was Hannibal’s masterpiece, perhaps the most astonishing military victory in history, still assiduously studied in military academies worldwide, and the inspiration for a number of relatively recent war engagements, from the Schlieffen Plan in WWI to General Schwarzkopf’s envelopment of Saddam Hussain’s army in the first Gulf War. It has generated an extensive body of literature, ranging from specialized books and studies published in many languages to chapters and discussions in numerous textbooks. And yet, despite the level of interest generated, many aspects of the clash between the multi-ethnic Carthaginian forces and the largest army Rome had ever raised remain obscure and controversial. This may be in part the result of the loss or destruction of the Carthaginian records of the event, coupled with the distortions and contradictions in the pro-Roman accounts, amounting to what, in today’s terms, we would call a campaign of disinformation.
The first point of controversy concerns the actual strength of the armies that took part in the celebrated battle. After being defeated in the cavalry encounter at the Ticinus river (218 BCE), at the battle of the river Trebia (218 BCE), and at Lake Trasimene (217 BCE), the Romans decided to raise a massive army to get rid of Hannibal once and for all. Polybius (3:107) tells us that the Romans recruited no less than eight legions, to be matched by an equal number of legions from their Italian allies, for a grand total of 16 legions, and that the numerical strength of each legion was increased from 4,000 to 5,000. Consequently, the size of the infantry forces marshaled by the Romans totaled 80,000. Hannibal’s army at Cannae, on the other hand, numbered 40,000 foot soldiers, giving the Romans a 2:1 numerical advantage. These numbers are usually accepted as valid by most historians and there is no reason to challenge them. The only additional factor is the number of troops staying back in the Roman and Carthaginian camps and therefore not actually participating in the battle. The 10,000 number usually given seems too large, as this was a decisive conflict and both sides would have wanted to field the largest possible number of actual combatants in the field. It must also be remembered that the camps would have included a large number of support personnel that typically accompanied ancient armies in the field, and these may have led to the inflated number reported by the sources.
The question, though, is the size of the cavalry forces that fought in the battle of Cannae, and especially the total strength of the Roman horse. According to the classical sources (Polybius and Livy), the Roman defeats at the Ticinus, Trebia, and Trasimene battles were in part the direct result of numerical superiority of the Carthaginian horse. It stands to reason, then, that in recruiting cavalry complements for their enlarged legions, the Romans would also have increased the size of the equestrian forces accompanying each legion, probably from 200 to at least 300, or perhaps even 350 or 400. Polybius (3:107) actually states that “on occasions of exceptional gravity” the Romans increased the size to 300, and that the numbers of the allied cavalry were required to be “three times as numerous as the Roman.” Hannibal’s successive defeats of three Roman armies certainly constituted an “occasion of exceptional gravity,” for the very survival of the Republic was at stake. It makes sense to assume that the cavalry, whose weakness had contributed to the previous defeats, was particularly strengthened. Polybius indicates elsewhere (4:25) that a typical legion of his day included 300 horse riders (rather than 200) as standard number, which makes an enhanced figure of 400 for “special occasions” even more likely. If we use the figure of 400 per Roman legion (or a total of 3,200 for the eight legions at Cannae), and triple that number for the allied legions, giving each a strength of 1,200 (or 9,600 total for the eight allied legions), we arrive at a combined cavalry number of 12,800 riders. This would mean that the Romans not only outnumbered Hannibal’s infantry forces, but also held numerical superiority in cavalry (Hannibal’s numbering 10,000), contrary to the usual narrative. Even if we accept only an increment for “occasions of exceptional gravity” of 300 rather than 400, this would total 2,400 (300 x 8) Roman and 7,200 (900 x 8) allied cavalry, or a total of 9,600. It can be concluded, therefore, that the Romans had at least practical parity, if not superiority, to Hannibal’s horsemen.
Polybius (3:107), though, lists the strength of the Roman horse only as being “over 6,000.” Why this discrepancy? Perhaps one reason can be found in the fact that the Romans consistently rationalized their defeats by attributing them to Hannibal’s superior cavalry numbers, so that accepting that they actually held superiority, or at least parity, in cavalry, at the moment of their greatest disaster, would have made their defeat that much more shameful. Polybius wrote 50 years after the battle, and although he may have interviewed some survivors, he did not witness the event. The Romans prided themselves of being the best warriors of their time, a self-image challenged by Hannibal’s victories. While necessarily recognizing and even reluctantly admiring Hannibal’s genius (for how could they have been defeated by a lesser general?), they needed a way to excuse and rationalize their own failure. Hannibal’s alleged cavalry superiority of 10,000 against the “over 6,000” Roman riders provided such an explanation.
Incidentally, Gregory Daly has argued (in his book on Cannae) that the larger numbers for the Roman horse are unlikely because “early encounters between Roman and Numidian cavalry suggest no significant qualitative difference between the two” and that therefore “the annihilation of the citizen cavalry at Cannae can be largely explained by their being greatly outnumbered.” As we will see in a following article, though, there was a different reason for the defeat of the Roman and allied horse, which did not preclude their actual numerical superiority. It can also not be argued that the Romans would not have been able to raise a larger cavalry force due to a shortage of horses or riders, for even after Cannae they continued recruiting large cavalry contingents for scores of legions; clearly, no such shortage existed. We must therefore conclude that at Cannae Hannibal’s army of 40,000 foot soldiers and 10,000 riders faced the vastly numerically superior Roman force of 80,000 infantry and 12,800 (or at least 9,600) cavalry.
A second point of controversy concerns the question of who was actually in command of the massive Roman army that fought at Cannae? Following the disastrous defeat of the Roman army and the death of its commanding officer, Consul Gaius Flaminius, at Lake Trasimene in 217 BCE, Fabius Maximus was appointed dictator for six months. Realizing that he was no match for Hannibal on the battlefield, he wisely followed the crafty Carthaginian with his army at a distance, harassing the enemy but refusing to face him in battle (an approach that has become known as “Fabian tactics”). This earned him the derogatory nickname of “cunctator”, or “delayer.” Roman pride and bellicosity simply could not tolerate the situation for long, and at the end of his term the Senate replaced him with two new consuls who were given the specific assignment of dealing with Hannibal once and for all. The traditional account names the consuls as the prudent and aristocratic Lucius Aemilius Paullus and the brash and impulsive Gaius Terentius Varro, a plebeian, the son of a butcher. Varro was a demagogue and a braggart, and it was supposedly he who was in command at the time of the battle, and thus deserves the blame for the worst military disaster in the history of Rome.
But was Varro really in command? And did Aemilius Paullus advise against engaging Hannibal in battle, as the classical accounts (Polybius and Livy) claim? The consuls alternated command on a daily basis, and we are told that Aemilius Paullus refused to commit himself to battle the previous day, but on August 2, when Varro was in charge, the hot-headed plebeian consul ordered the Roman army to deploy for battle. In the first place, this does not make sense, because the stated mission of the largest Roman army ever assembled was to face and defeat Hannibal in battle—clearly both consuls were interested in engaging and destroying the Carthaginian invader. But perhaps Aemilius Paullus’s alleged reluctance resulted from the battlefield being favorable to the tactics of Hannibal, a factor overlooked by the supposedly impulsive Varro?
Was the plain of Cannae, next to the Aufidus River (today’s Ofanto) really a poor choice of battleground for the Romans? Hannibal was a master of the ambuscade, and had surprised the Romans both at the Trebbia and at Lake Trasimene, where he hid his entire army in ambush. Clearly, the Romans needed a battlefield where Hannibal could not hide forces or spring a surprise attack from some unexpected angle. Cannae was perfect for this—a plain allowing no shelter for hidden forces to launch an ambuscade. The argument that the plain offered ideal terrain for the deployment of cavalry, Hannibal’s strongest contingent, and that although in infantry the Romans outnumbered the Carthaginians 2-to-1, Hannibal had numerical superiority in horse, is also flawed. We have already established that such superiority was fictitious, and that in reality the Romans had parity or even numerical superiority in cavalry as well. The plain of Cannae offered the Romans the perfect field to use their enhanced equestrian forces and their massively superior infantry, with no worries of a surprise ambush. Both Aemilius Paullus and Varro would have approved of the battlefield.
A key factor that points toward Aemilius Paullus having been in command is his position at the start of the battle. As Seibert points out, the traditional place for the commanding consul would have been with the Roman horse on the right wing, which is precisely where Aemilius Paullus was. He could also have been with the infantry, but most certainly as commanding consul he would not have been on the left wing, with the allied cavalry (the horse supplied by the Italian allies of Rome), a position of less honor and distinction. It was Varro who was positioned on the left wing, a clear indication that he was not the commanding consul that day.
There is an additional factor that points to Aemilius Paullus, and not Varro, as the consul in command, namely that the former, injured in battle, when offered the opportunity to escape when it was clear that defeat was inevitable, refused to leave. His choice of certain death makes sense if he felt disgraced and responsible for having caused the greatest calamity in the history of Rome. Varro, on the other hand, fled with a small surviving contingent of the allied horse, and, upon returning to Rome, was received with open arms and thanked for “not despairing of the Republic,” hardly the reception he would had received had he been in command and thus responsible for the massacre. That this was not an uncharacteristic act of largesse on the part of the Roman Senate is confirmed by the fact that Varro held further command during the war years that followed—it is unthinkable that further forces would have been entrusted to Varro, had he been considered the incompetent and disgraced commander responsible for the crushing defeat of the Roman legions at Cannae.
But why do the classical records distort the identity of the commanding Roman general? In the case of Polybius, our otherwise most reliable source, it must be remembered that he was in the employ of the Aemilian family. It is possible that he may have intentionally changed the name of the consul in charge to protect the honor of the patrician family he served, as well as the memory of Aemilius Paullus, who, in his version of the battle, appears as a wise and prudent man, a hero who sacrifices himself to die with his soldiers. Varro was the obvious scapegoat on whom to blame the defeat. And as for Livy, who wrote two centuries after Cannae, his patriotic zeal in praising the virtues of the Republic to inspire the Roman youth of his day would naturally have inclined him to follow Polybius, protecting the honor of an aristocrat and blaming the calamity on a plebeian demagogue.
A third controversial point concerns the different casualty figures given by Polybius and Livy for the Battle of Cannae. Which, if any, is credible? Most historians agree that the casualty figures recorded in various ancient sources for battles in Antiquity are greatly exaggerated. The exception is Cannae, due to the nature of the double envelopment that resulted in the destruction of the Roman army. Even Delbrueck, who in his History of Warfare is very skeptical of such numbers, accepts as credible the enormity of the losses at Cannae.
Polybius (3:117) states that only 70 of the allied cavalry managed to escape with Varro and that 300 others “reached different cities in scattered groups.” He further indicates that some 10,000 Romans were captured, “but not in the actual battle, while only perhaps three thousand escaped from the field to neighboring towns.” He adds: “All the rest, numbering about seventy thousand, died bravely.” As for Hannibal’s losses, Polybius lists “about four thousand Celts, fifteen hundred Spaniards and Africans, and two hundred cavalry,” or a total of 5,700. Livy (22:49), on the other hand, gives a lesser figure for the Roman losses (about 50,000) and a larger for the Carthaginian dead (8,000). How believable are these figures, and why is there such a large discrepancy between Polybius’s numbers and the ones reported by Livy?
Some historians, such as Goldsworthy (2001) and Daly (2002), in their respective books on Cannae, have balked at accepting the Polybian figures, because they seem larger than possible if one assumes that the Romans had only 6,000 cavalry. The total strength of the Roman army would then have been 80,000 infantry plus 6,000 horse, or 86,000, and Polybius’s figures, 70,000 fallen, plus 10,000 prisoners, plus up to 10,000 survivors, would total about 90,000, which would constitute an impossibility. But we have already established that the real cavalry strength of the Romans at Cannae was, in all likelihood, somewhere between 9,600 and 12,800—quite a bit more than the 6,000 cited by most authors (who fail to heed the hint in Polybius’s qualifier “over,” indicating that the number was greater than 6,000). Taking the revised figures into account, the total strength of the Roman army at Cannae was between 89,600 and 92,800m and Polybius’s casualty numbers create no contradiction.
Livy, on the other hand, followed Polybius’s 6,000 figure literally, and, faced with the apparent contradiction, gave a much lower number for the Roman fallen, about 50,000 (45,000 foot soldiers and 2,700 horsemen). Livy does not actually name his source, and some think it may have been the now lost account of Quintus Fabius Pictor. Nevertheless, it must be remembered that Livy’s explicitly stated motivation in writing his history of Rome was to inspire patriotism in the Roman youth of his day. The creation of a fictional numerical superiority in the Carthaginian horse, plus a sharp reduction in the number of Roman dead, allowed for the greatest shame of Roman arms to be substantially diminished.
Due to the overall greater reliability of Polybius’s account over Livy’s, it seems reasonable to agree with Delbrueck and accept that, indeed, 70,000 Romans and a little over 5,000 of Hannibal’s men died on the plain of Cannae, on that fateful day of August 2, 216 BCE. The Roman dead included the consul in command, Lucius Aemilius Paullus, two quaestors, 29 military tribunes, the proconsuls Servilius and Minucius, and at least 80 men of senatorial rank. It was the worst disaster Rome had ever experienced. Incidentally, it should be added that the 10,000 Romans that were captured, mostly when they surrendered in the Roman camps following the destruction of the army at Cannae, are not indicative of that many soldiers having been kept in the camps and not actually participating in the battle. Instead, it is reasonable to assume that many of them managed to escape from the battlefield before the completion of the fatal encirclement, and then took refuge in the camps.
The scale of the losses at Cannae is staggering. Over 75,000 men perished in a single day of combat. Not until the scale of destruction of the major battles of World War I in the 20th century, over 2,100 years later, were these terrible numbers surpassed. And those modern battles inflicted their death tolls with machine guns, bombs, cannons, tanks, and poison gas. Hannibal’s tactics at Cannae resulted in losses comparable to the casualties resulting from the atom bombs dropped on Hiroshima and Nagasaki, in a single day of hand-to-hand combat with swords and spears.
We will analyze how Hannibal managed to achieve his devastating victory against the overwhelming odds stacked against him in another article.
Bibliography
Barcelo, Pedro. Hannibal. Stratege und Staatsmann. Stuttgart, Germany: Klett-Cotta, 2004.
Bender, Peter. Weltmacht Amerika-Das neue Rom. Stuttgart, Germany: Klett-Cotta, 2004.
Christ, Karl. Hannibal. Darmstadt, Germany: Primus Verlag, 2003.
Daly, Gregory. Cannae: The Experience of Battle in the Second Punic War. London, England: Routledge, 2002.
De Beer, Gavin. Hannibal: Challenging Rome’s Supremacy. New York: Viking, 1969.
Delbrueck, Hans (Walter J. Renfroe, translator). History of the Art of War Within the Framework of Political History, Volume I: Antiquity. Westport, CT: Greenwood, 1975.
Dodge, Theodore Ayrault. Hannibal: A History of the Art of War Among the Carthaginians and Romans Down to the Battle of Pydna, 168 BC, With a Detailed Account of the Second Punic War. New York: Houghton Mifflin, 1891.
Egelhaaf, Gottlob. Hannibal. Ein Charakterbild. Stuttgart, Germany: Carl Krabbe, 1922.
Ewert, Michael. Amerikas punische Kriege. Niedergang, Terror und Gehirnwaesche. Neu Isenburg, Germany: Melzer, 2005.
Goerlitz, Walter. Hannibal. Eine politische Biographie. Stuttgart, Germany: W. Kohlhammer, 1970.
Goldsworthy, Adrian. Cannae. London, England: Cassell, 2001.
Groag, Edmund. Hannibal als Politiker. Rome, Italy: “L’Erma” di Bretschneider, 1967.
Healy, Mark. The Battle of Cannae: Hannibal’s Greatest Victory. Oxford, England: Osprey, 1994.
Kromayer, J. Roms Kampf um die Weltherrschaft. Leipzig, Germany: B. G. Teubner, 1912.
Lancel, Serge. Hannibal. Paris, France: Fayard, 1995.
Lehmann, K. “Das Cannae-Raetsel.” Klio: Beitraege zur alten Geschichte, 24:1, 1930, 70-99.
Livy (B. O. Foster, translator). History of Rome, Books XXI-XXII (Loeb Classical Library). London: Harvard, 1929 (reprinted 2001).
Maestre Yenes, Maria A. H. “Una explicacion matematica del triunfo militar de Anibal en la batalla de Cannas.” Estudios Clasicos, 15, 1971, pp. 27-48.
Mosig, Yozan. “Review of The Ghosts of Cannae: Hannibal and the Darkest Hour of the Roman Republic by Robert L. O’Connell.” The Journal of Military History, 74:4, 2010, pp. 1261-1263.
Mosig, Yozan and Imene Belhassen. “Revision and reconstruction in the Punic Wars: Cannae revisited.” The International Journal of the Humanities, 4(2), 2006, pp. 103-110.
O’Connell, Robert L. The Ghosts of Cannae: Hannibal and the Darkest Hour of the Roman Republic. New York: Random House, 2010.
Polybius (W. R. Patton, translator). The Histories, Volume II (Loeb Classical Library). Boston, MA: Harvard, 1922 (reprinted 1999).
Russo, Renato. La battaglia di Canne e la campagna annibalica in Puglia nel contesto della seconda guerra punica. Barletta, Italy: Rotas, 1999.
Samuels, Martin. “The Reality of Cannae.” Militaergeschichtliche Mitteilungen, 47, 1990, pp. 7-31.
Shean, John F. “Hannibal’s Mules: The Logistical Limitations of Hannibal’s Army and the Battle of Cannae, 216 B.C.” Historia, 45:2, 1996, pp. 159-187.
Seibert, Jakob. Forschungen zu Hannibal. Darmstadt, Germany:Wissenschaftliche Buchgesellschaft, 1993.
Seibert, Jakob. Hannibal. Darmstadt, Germany: Wissenschaftliche Buchgesellschaft, 1993.
Seibert, Jakob. Hannibal. Feldherr und Staatsmann. Mainz am Rhein, Germany: Philipp von Zabern, 1997.
© Yozan Mosig, 2012.
- Hannibal: Victories of the Great Hero from Ancient Carthage - December 19, 2018
- Alexander the Great and Hannibal Barca: A comparison - March 9, 2017
- A Note on Hannibal’s Losses During the Crossing of the Alps - July 18, 2016