Cannae - why bother with the cavalry?

[Frank]

Would this be too complicated for their standard fighting order to pull off given the commanders they had?

Probably not too complicated. Caesar did something similar with Labienus' cavalry at Pharsalos. But not every roman commander was like Caesar

[d.carmichael]

yeah thats what I was thinking about when I thought of the idea. Perhaps this could have turned the tides and made Cannae a roman victory.

[Macedon]

My opinion that the Romans fought in a single line is a result of various facts in the accounts of both Livy and Polybius as well as Appian, although the last one mentions as I write below the Romans forming in three lines.

1. Only Appian has the Romans draw up in three lines, Polybius often says so about the Romans in other battles, not in this case.

As I said, Appian has the Romans array in three lines and also mentions further reserves of cavalry but when he writes of the Celtiberians who had purposefully deserted to the Romans before the battle, he has them attack the Roman rear, who could not face about because it was frontally engaged, clearly, at least at the time, having the Romans fight in single phalanx. It is my personal opinion that he is wrong regarding the three lines -or that he is describing the formation before closing up the line for the charge-, but of course it is open for anyone to assume otherwise.

2. Both Polybius and Livy have the Roman infantry array in (abnormally) deep files. Had the Romans arrayed in three lines as normal, the first line would anyway have been about 2/7 of the infantry, which would be some 20,000 men, much less than the 40,000 (or about 32,000 without the Africans) of the enemy. In deep files, their frontage would have been even less than normal which would make the Carthaginian infantry line much longer unless, for some reason, it had also arrayed unusually deep. Furthermore, and I have already taken this into account for the above calculation, the Roman Triarii were probably, as commonly supported since the 10,000 troops are not actually identified in the sources, left as guard camp and that is also evidence that at least no third line was deployed.

3. When the center of Hannibal's line gave way, the accounts are clear in that the whole Roman infantry followed, funneling into the "gap". Had there been a second or a third line, it would have been possible for the Carthaginians to surround only the Hastati, since the lines of the Principes (and possibly the Triarii) would wait intact behind them. However, the Romans follow en mass and are encircled en mass. Also note that there is no mention of any retreat, relief or other type of interaction with any kind of infantry reserves.

In conclusion, the number of lines the Romans fought in at Cannae is of course a matter of debate and certainly open to interpretations but the fact that the Romans DID try to improvise and array in a manner that was different to what was considered normal, is not doubted. It is my personal opinion that they arrayed in a single line, an assumption that explains a number of tactical problems and issues met in the accounts of all sources and able to mathematically depict the numbers cited on the battleground.


Regarding the Ofanto (Aufidus). You might want to have a look at it in Google maps (or even in Wikipedia) to get an idea of its nature. Of course, this alone is not proof enough, although it is indicative (you can clearly see it is a rather small and short river), however, both texts also tell of multiple fords near both adversaries, which clearly shows that the river was in many places fordable. Its meandering nature is also evident and fully expected and of course the fact that the battle was fought in August also points at a minimum rather than a maximum flow.

I apologize for not making clear that the fact that the single line scenario was my opinion, not clearly supported by the sources, and I would be interested to know how the course of the battle could be explained with the Romans arraying in two or even three lines in your opinions, should you support the multiple lines scenario, especially considering the length of the Roman infantry lines as opposed to that of Hannibal's, the role of the rear lines in the accounts of the three historians and the course of the battle as described by them.

[antiochus]

Macedon wrote:
[i]"I would be interested to know how the course of the battle could be explained with the Romans arraying in two or even three lines in your opinions, should you support the multiple lines scenario, especially considering the length of the Roman infantry lines as opposed to that of Hannibal's, the role of the rear lines in the accounts of the three historians and the course of the battle as described by them."[/i]

Although the question is not aimed at me, I have decided to step into the fray.

I am a great supporter of Appian’s claim the Roman infantry were in three lines at Cannae as his description combined with Livy and Polybius are coherent. The first line consisted of the light infantry, the second line the hastati, and the third line, the principes and triarii. At this stage in Rome’s military development, the triarii are combined with the principes so as to give the line sufficient depth. The reason behind this is because by the time of Cannae, the emergency legion or the 5000 man legion had undergone a reform. In its early history, to increase the legion to 5000 men, extra triarii were added, termed by Livy as accensi. This practice was eventually dropped in favour of adding more principes, which accords with Polybius statement of the triarii remaining constant. Although the triarii are formed up with the principes, at Sentinum Livy mentions the last ranks being detached. These are the triarii. It will not be until Scipio’s African campaign that the triarii are once again deployed as a separate body. I believe this could be the result of the principes being rearmed with the pila. It could be that while in Spain, Scipio was influenced by the scutarii, and wanted to model the principes on them.

Macedon wrote:
[i]"When the center of Hannibal's line gave way, the accounts are clear in that the whole Roman infantry followed, funneling into the "gap". Had there been a second or a third line, it would have been possible for the Carthaginians to surround only the Hastati, since the lines of the Principes (and possibly the Triarii) would wait intact behind them. However, the Romans follow en mass and are encircled en mass. Also note that there is no mention of any retreat, relief or other type of interaction with any kind of infantry reserves."[/i]

Polybius does write the maniples were closer than normal and this can lend itself to the principes/triarii being closer to the maniples of hastati. One incident that is overlooked in many modern accounts of Cannae is Livy’s reference that the Africans ‘extended.’ Also during the last stages of the battle a tribune offers Paullus his horse so he can escape. Therefore, the Romans are not encircled en mass as there is still a way of escape. In Appian’s account it is only the last remnants that are encircled en mass and proving the most difficult to eliminate. It is at this point Appian remarks that Hannibal rode around his troops, and while encouraging them, abused them, “calling it shameful if they could not finish off this small group after their victory over the mass.” Appian (The Hannibalic War7 4 24)

There is no way Hannibal’s plan at Cannae was to encircle en mass the whole Roman infantry, especially in light of Frontinus’ comment that at the battle of Lake Trasimene, after the Romans had been enveloped, the Roman fought with greater ferocity. The ferocity of the Romans convinced Hannibal to open up his ranks to give the Romans the opportunity to escape so he could overwhelm them later without loss to his own men. Frontinus (Stratagems 2 6 4) As Livy also mentions bodies of Romans still escaping after his claim they were trapped, this would indicate an opening still remained for the Romans to escape through and take refuge in the Roman camp.

In any reconstruction of Cannae there is as lot of boxes to tick, and I feel confident I have achieved this. Polybius’ statement the maniples were much deeper than their frontage, lends itself to only one mathematical deployment possible for the Romans. Then by taking this I reconstructed the frontage of the Carthaginian line, which then accorded with Polybius’ statement the Carthaginian infantry were over 40,000 men. In fact the numbers and the depth of the Carthaginian infantry at Cannae follows the same mathematical pattern as it does at the Trebia, Illipa and Zama. I originally used the 12,000 mercenaries in Hannibal’s first line in relation to Scipio’s four legions and found a constant pattern throughout the Second Punic war for the Carthaginian infantry. The frontage I have for the Roman army (both infantry and cavalry), is the smallest ever given by a modern historian. The Romans are not doing anything ad hoc at Cannae, they are simply adhering to one of the deployment arrangements the legion is capable of doing in relation to what their organisational mathematics will permit. The frontage of a maniple at Cannae accords with the frontage of a century in the pre-maniple legion, so there is nothing new.

The plan at Cannae is for the Roman and allied cavalry to hold the Carthaginian cavalry at bay so the Roman infantry can break through the Carthaginian infantry. The Roman infantry are deployed so they can achieve a break in any section of the Carthaginian line. However, for the Roman cavalry on the right wing to hold the Carthaginian cavalry at bay so the infantry can achieve the break through, the Roman cavalry is deployed six squadrons deep. The first three lines of squadrons will work on a time basis in combination with the light armed infantry and the hastati. If the hastati do not make the break through, when they are withdrawn, the first three lines of squadrons are replaced by the remaining three lines of squadrons, who will hold the Carthaginian cavalry at bay, while the principes/triarii breaks the Carthaginian line. The allied cavalry on the left wing are also deployed many squadrons deep. However, on this front the allied cavalry and the Carthaginian cavalry seem to have been ordered not to be aggressive. This would have suited Hannibal because there are many references of the Roman and allied cavalry being able to defeat the Numidian cavalry.

If you read Livy and Polybius they have the Roman cavalry being defeated while the Roman light infantry are skirmishing. So much for the Roman cavalry giving the infantry the time they needed. It is at this point when the Roman cavalry are routed that the light infantry are recalled. At this point in time, Paullus knew it was now a race against time. And I believe with the complete collapse of the Roman cavalry in such a short amount of time, Paullus must have been shattered and deeply concerned. In my study of Cannae, even after the Africans extend, they still cannot encircle the Roman infantry. A lot more Roman infantry has to die or flee before the Africans can encircle the remnants.

So in a nutshell, I lay the blame for the defeat of the Romans at Cannae squarely with the Roman cavalry.

Steven

PS: Sorry for not bolding or Italic original questions but the BBC code editor will not load, and the colour of the topic bars are now blue instead of red.

[Macedon]

An interesting approach. I have some questions and some comments on it, although most have little to do with the number of lines in the Roman formation.

I am a great supporter of Appian’s claim the Roman infantry were in three lines at Cannae as his description combined with Livy and Polybius are coherent. The first line consisted of the light infantry, the second line the hastati, and the third line, the principes and triarii."

Appian's account is interesting and most probably drawing on a different tradition to that of both Livy and Polybius. However, I have to say here that according to him, all three lines he initially describes are heavy infantry, "each part of whom (here he refers to the lines, which he calls "parts") had the infantry in the middle, the light infantry and the cavalry on each side". The most probable reading here has the Romans indeed draw up in three lines of heavy infantry as well as three lines of cavalry and light infantry on the flanks. However, it could also read that the cavalry and light infantry were posted on the flanks of all three lines, which to me is more probable tactically and supported by Appian's further words, since later he writes "the Roman horsemen, being less in number than their enemy, advanced against them, extending their formation, making it thin". This would make no sense, should they have divided their horse in three lines, unless they then made them into one. There is also a problem with the light infantry, which Appian posts with the cavalry (whether in one or three lines) but then has them "run before the rest of the army in the interim space". Regarding your Triarii and the Principes being posted in the same line, I will only say that regardless the veracity of the claim that they were armed in the same manner and so probable to fight as a single formation, you need Appian to speak of two lines of infantry instead of three, should you of course choose to follow him in this.

Polybius does write the maniples were closer than normal and this can lend itself to the principes/triarii being closer to the maniples of hastati. One incident that is overlooked in many modern accounts of Cannae is Livy’s reference that the Africans ‘extended.’ Also during the last stages of the battle a tribune offers Paullus his horse so he can escape. Therefore, the Romans are not encircled en mass as there is still a way of escape. In Appian’s account it is only the last remnants that are encircled en mass and proving the most difficult to eliminate. It is at this point Appian remarks that Hannibal rode around his troops, and while encouraging them, abused them, “calling it shameful if they could not finish off this small group after their victory over the mass.” Appian (The Hannibalic War7 4 24)

There is no way Hannibal’s plan at Cannae was to encircle en mass the whole Roman infantry, especially in light of Frontinus’ comment that at the battle of Lake Trasimene, after the Romans had been enveloped, the Roman fought with greater ferocity. The ferocity of the Romans convinced Hannibal to open up his ranks to give the Romans the opportunity to escape so he could overwhelm them later without loss to his own men. Frontinus (Stratagems 2 6 4) As Livy also mentions bodies of Romans still escaping after his claim they were trapped, this would indicate an opening still remained for the Romans to escape through and take refuge in the Roman camp.

In my opinion, all sources stress the fact that Hannibal planned to surround the Romans (and thus cause maximum casualties). Polybius does not say that the speirae were closer to each other, so that we could assume that he meant the speirae of the different lines. He says that "they formed the speirae denser than before". This is very clear in the Greek text and different translations are plainly wrong. Regarding Livy, I of course have read that he has the Africans extend, but I cannot confirm or contest the accuracy of the translation, since I unfortunately cannot read Latin... What I have understood from the translation was that, according to him, the Africans managed to form a line on the rear of the Romans, a maneuver impossible to carry out, had there been a second (or third) line of any troops behind the victorious legions in front as I see it. According to Polybius and Plutarch, it was the Roman flanks that the Africans attacked and Appian makes no mention of them, nor of the crescent formation or its function. However, he especially stresses Hannibal's plan to encircle the Romans with his cavalry and the role of various stratagems he has him employ all around the battlefield, so he seems to agree more with Polybius in that matter.

As far as the encirclement is concerned, surrounding an enemy with cavalry, as was the case at Cannae (or partly surrounding them with cavalry as is the case in Appian’s account), is not the same as completely encircling them with infantry. Cavalry formations are much looser and enemy horsemen as well as infantry groups which keep in a body can escape. However, it is of course true that casualties during such attempts are monstrous as was the case here too.

In any reconstruction of Cannae there is as lot of boxes to tick, and I feel confident I have achieved this. Polybius’ statement the maniples were much deeper than their frontage, lends itself to only one mathematical deployment possible for the Romans. Then by taking this I reconstructed the frontage of the Carthaginian line, which then accorded with Polybius’ statement the Carthaginian infantry were over 40,000 men. In fact the numbers and the depth of the Carthaginian infantry at Cannae follows the same mathematical pattern as it does at the Trebia, Illipa and Zama. I originally used the 12,000 mercenaries in Hannibal’s first line in relation to Scipio’s four legions and found a constant pattern throughout the Second Punic war for the Carthaginian infantry. The frontage I have for the Roman army (both infantry and cavalry), is the smallest ever given by a modern historian.

So, in your opinion, what is the length of the Roman infantry formation in your reconstruction in relation to the Carthaginian one? How can 32,000 Celts in single line, those in the middle relatively thinly arrayed have the same frontage with, say, 50,000 Romans (assuming a mere 25% light troops out of the 80,000 total Roman infantry strength) deployed extra deep and in at least close order, if they arrayed in more than one line? Even if they arrayed in 2 and not 3 lines it would be a riddle. And should they array in 2 lines, as in your model, then wouldn't the first line actually consist of less men than the second line, as the latter would include both the Principes and the Triarii that followed the army to the field? Unless you believe that the Romans attacked in a line that was substantially shorter than that of Hannibal’s in order to keep the multiple infantry lines, which however play no role in any of the accounts.

The plan at Cannae is for the Roman and allied cavalry to hold the Carthaginian cavalry at bay so the Roman infantry can break through the Carthaginian infantry. The Roman infantry are deployed so they can achieve a break in any section of the Carthaginian line. However, for the Roman cavalry on the right wing to hold the Carthaginian infantry at bay so the infantry can achieve the break through, the Roman cavalry is deployed six squadrons deep. The first three lines of squadrons will work on a time basis in combination with the light armed infantry and the hastati. If the hastati do not make the break through, when they are withdrawn, the first three lines of squadrons are replaced by the remaining three lines of squadrons, who will hold the Carthaginian cavalry at bay, while the principes/triarii breaks the Carthaginian line. The allied cavalry on the left wing are also deployed many squadrons deep

How did you come to the six squadron depth of the Roman right and what do you mean by it exactly? Is this some system of cavalry fight you have devised yourself or do you base its existence on sources?

[antiochus]

Macedon wrote:
Regarding your Triarii and the Principes being posted in the same line, I will only say that regardless the veracity of the claim that they were armed in the same manner and so probable to fight as a single formation, you need Appian to speak of two lines of infantry instead of three, should you of course choose to follow him in this.

Appian is discussing the Roman infantry being deployment in three lines before the battle commenced. Why does he have to continue to speak of the light armed infantry after they are withdrawn? Does Polybius tell us anymore of the velites after they dealt with the elephants? No he doesn’t. The fact the light armed started the battle is attested by both Livy and Polybius. You cannot reconstruct Cannae by only employing Appian. How many lines does Livy or Polybius speak of? Who breaks through the Celts after the apex of the wedge is defeated? Unfortunately Appian, Livy or Polybius do not tell us.

Macedon wrote:
Polybius does not say that the speirae were closer to each other, so that we could assume that he meant the speirae of the different lines. He says that "they formed the speirae denser than before". This is very clear in the Greek text and different translations are plainly wrong.

I have other translations done for me and they have the maniple many times deeper than its frontage and the maniples closer together. I have provided the Frontinus reference of Hannibal creating gaps at Trasimene for the Roman army in part to escape to show he would have learnt from this and not to repeat it at Cannae. Even the Mongols knew that you must allow space for a surrounded army to try and escape through (giving hope) otherwise they will furiously fight to the death when they know there is no possibility of escape. I do not believe Hannibal wanted to completely trap the whole 14 legions in one pocket. As I see it, as the Romans were fleeing and being killed, the pocket got smaller and the Africans overtime kept extending until the gap was closed.

Macedon wrote:
According to Polybius and Plutarch, it was the Roman flanks that the Africans attacked.

Gary Brueggeman’s insightful website has a very interesting discussion on cavalry attacking the flanks of the infantry. It shows how only a few cavalry can actually attack a flank, with the majority of the cavalry having nothing to attack but thin air. At Cannae, a lot of the African infantry would be in the same position with nothing to attack. A flank attack is synonymous with attacking the flank and rear of a unit. Livy’s comment about the Africans extending does indicate they are being to encircle the Romans. In my book, I do not just state the Africans extend and leave it at that, I go into detail via diagrams as to how in the most simple and fast manner they could accomplish this. Here you will find I am strongly influenced by the Gary Brueggeman approach (don’t just state it, show it). This is where those military manuals come into play. Also in my study of Zama, I don’t state the Romans formed a single line, I detail it in diagrams how it could be done. This exercise produced a little surprise. Same with Ilipa, I give the numbers of the troops involved in the outflanking movement, their formations and how they wheeled, and most importantly, what the outflanking forces were trying to achieve.

Macedon wrote:
As far as the encirclement is concerned, surrounding an enemy with cavalry, as was the case at Cannae (or partly surrounding them with cavalry as is the case in Appian’s account), is not the same as completely encircling them with infantry.

There has been a tradition among historians of the cavalry completing the encirclement. What are the cavalry doing? Are they standing stationary horse shoulder to shoulder so to form an unbreakable line while dispatching the Roman infantry? This scenario lacks common sense and only possible in the mind of the beholder, like Sir Walter Raleigh. I have come to the conclusion that when the Carthaginian cavalry returned to the battle, they were running down those Romans fleeing the pocket, and were not creating a wall so as to prevent the Romans from fleeing. I also don’t have the Africans completely encircling the Roman infantry. My study shows the Africans don’t have the numbers, therefore the frontage to do this properly. It is only after the pocket has been severely reduced can the Africans accomplish this.

Macedon wrote:
So, in your opinion, what is the length of the Roman infantry formation in your reconstruction in relation to the Carthaginian one?

The Roman infantry have the same frontage as the Carthaginian infantry.

Macedon wrote:
How can 32,000 Celts in single line, those in the middle relatively thinly arrayed have the same frontage with, say, 50,000 Romans

Where did you get the figure of 32,000 Celts? Why do you believe the 32,000 Celts are deployed in a single line? What happened to the wedge formation? Taking your army numbers, if the field of Cannae can only allow a frontage of 1000 men, then your 32,000 Celts would be deployed 1000 men wide by 32 ranks deep. Your 50,000 Romans would be deployed 1000 men wide by 50 men deep.

At the Trebia, Hannibal had over 20,000 men, at Cannae over 40,000 men. At the Trebia the Romans had eight legions, at Cannae 16 legions. So both armies had doubled their infantry. At the Trebia the Roman broke through the Carthaginian line. This leaves the question as to why did they believe they had to deepen the depth of the maniples at Cannae.

Macedon wrote:
And should they array in 2 lines, as in your model, then wouldn't the first line actually consist of less men than the second line?

Yes the first line (the hastati) will have less men than the second line (principes/triarii). However, this does not affect the legion’s frontage. Following Polybius, if you have 1200 hastati in a legion and 1200 principes, and you decide to deploy them 200 men wide, the hastati and principes both have a depth of 6 men. This means the 600 triarii have a depth of three men, and when the principes are combined with the triarii, the principes and triarii have a combined depth of 9 men. The legion then has a frontage of 200 men and a combined depth of 15 men. If you include the 1200 velites, then they would be deployed 200 men wide by 6 men deep. Now the legion has a combined depth of 21 men. In board game jargon, all legions have a zone of control. With a legion having a frontage of 200 men, 14 legions have a frontage of 2800 men. Now you need to implement Polybius’ maniple arrangement for each legion and in doing so each legion and therefore each line within a legion will have the same frontage. So the hastati will end up having the same frontage as the principes and triarii, accept the principes/triarii have a greater depth. The formula I found is when the Romans increase the size of the legion; they increase the depth, not the frontage. In this manner, they do not have to increase the number of centurions or centuries.

Macedon wrote:
How did you come to the six squadron depth of the Roman right and what do you mean by it exactly? Is this some system of cavalry fight you have devised yourself or do you base its existence on sources?

This will upset a few of you. The answer is a combination of sources and Pythagorean geometry. This is combined with the research I have assembled on the legion from Servius to Vegetius, which shows some very illuminating military doctrines. Historians believe that Polybius’ comment that at Bagradas in 255 BC, the maniples were many maniples deeper was a one off arrangement and that is why Polybius mentioned it. Like many others I believed this was the case, but now I know it is not true. The formation adopted at Bagradas is a common formation when facing an enemy with a deeper formation, so Polybius never mentioned it again to avoid repetition. It is employed at Cynoscephalae in 197 BC. Philip’s right wing phalanx of 8000 men is deployed to match the frontage of the two Roman legions. I experimented with the available legion deployments and the one that match the Macedonian frontage was the exact deployment arrangement used at Bagradas. I’ve also done some exploration into Greek mathematics and found out why Aelian’s Hellenistic Military manual has increments of four. There is a lot more information floating around that will open more doors to the military organisation of the Greeks and Romans.

One item that surface during my research is I have found the Romans have a system for determining the frontage of a consular army. It makes sense they would have such an arrangement. The frontage of the cavalry and infantry in a consular army are combined and represent a specific number of legions. For example, the Roman cavalry when deployed could equate to the frontage of a half a legion, and the allied cavalry one and a half legions. Therefore, a consular army has the frontage equivalent to six legions. At Cannae, after redeploying the maniples according to Polybius, I compared this frontage with the legion’s normal frontage to see if there was a distinct mathematical pattern, which there is and to my delight it’s the standard Pythagorean ratio 3:2 (the perfect fifth) common to their military. The pattern that emerges is the Romans are reducing the frontage of four consular armies so that it conforms to #@!* consular armies. The next procedure is to determine if the infantry can form the so called quincunx formation. I know from the Servian times to Vegetius, this conforms to 3:2 ratio. Again it works perfectly and shows the incredible flexibility of Pythagorean geometry. The primary sources mention gaps between the maniples, between the ordines and between the cohort. At Cannae in order to conform to the Pythagorean ratio, which designates the number of intervals, the intervals are made by cohorts, which in the legion’s normal deployment would be equal to the frontage of a maniple. However, as most of you don’t believe in the existence of the cohort before 210 BC, then everything I am saying must be wrong. Regardless of your beliefs, the next stage is to determine if the Roman cavalry are doing the same thing. After doing a comparison of the space left, for the allied cavalry and the Roman cavalry, in order to get a correct fit, the allied cavalry can have squadron intervals, while the Roman cavalry cannot. For me this accords with Livy and Polybius comment the Roman cavalry did not have room to manoeuvre. The end result is the Roman cavalry are deployed six squadrons deep. But I will add the Roman cavalry can also deploy in the so called quincunx formation. This way the cavalry can replace squadrons with squadrons.

Steven

Again, I cannot get the BBcode editor to load, so I couldn't do the italic and bold. Must be something to do with the new internet service I have.

[Macedon]

Appian is discussing the Roman infantry being deployment in three lines before the battle commenced. Why does he have to continue to speak of the light armed infantry after they are withdrawn? Does Polybius tell us anymore of the velites after they dealt with the elephants? No he doesn’t. The fact the light armed started the battle is attested by both Livy and Polybius. You cannot reconstruct Cannae by only employing Appian. How many lines does Livy or Polybius speak of? Who breaks through the Celts after the apex of the wedge is defeated? Unfortunately Appian, Livy or Polybius do not tell us.

No problem with that. Just making clear that you also do not base your model on the three lines of heavy infantry described by Appian. I do the same thing. In my opinion, even if in the beginning they did deploy in their "traditional" triple acies, they later chose to form one line. In yours two.

I have other translations done for me and they have the maniple many times deeper than its frontage and the maniples closer together.

"πυκνοτέρας ἢ πρόσθεν τὰς σημείας καθιστάνων, καὶ ποιῶν πολλαπλάσιον τὸ βάθος ἐν ταῖς σπείραις τοῦ μετώπου·" The Greek text is very clear on this "He arranged the semaeae (maniples) more dense than before and the depth of the speirae (maniples again) (many) times greater than their frontage" So, we are talking about the soldiers in the maniples being arrayed closer to each other (which could be close or compact order) and practically in column. I still cannot see it as arranging the lines closer to each other, without of course claiming that they were not, I have them unite the lines myself. I only speak of the translation of the said text.

I have provided the Frontinus reference of Hannibal creating gaps at Trasimene for the Roman army in part to escape to show he would have learnt from this and not to repeat it at Cannae. Even the Mongols knew that you must allow space for a surrounded army to try and escape through (giving hope) otherwise they will furiously fight to the death when they know there is no possibility of escape. I do not believe Hannibal wanted to completely trap the whole 14 legions in one pocket. As I see it, as the Romans were fleeing and being killed, the pocket got smaller and the Africans overtime kept extending until the gap was closed.

Of course, I have scores of such examples and pieces of advice, However, in the texts, the terms "flank-attack" and "encirclement" are more vague than commonly perceived. It is especially clear in the later manuals that when troops attack the flanks of an enemy formation, they also often attack the rear too. So it is not about a long line engaging the few men forming the outermost file of the enemy. Wanting to completely trap the legions does not mean that you do that with infantry only. In this I personally choose the cavalry scenario, that is, that Hannibal completely surrounded the Romans with his cavalry (where his infantry was not engaged), who of course did not remain stationary as in some kind of phalanx but kept maneuvering back and forth as they used to do, giving the impression of hope as the space between them and the trapped Romans was empty, as was the space between the squadrons. The fleeing Romans, thousands and thousands of them, tried to escape through them, were mercilessly attacked and hacked down, resulting in the huge number of casualties we know of. Of course, some of them, those who were able and had the clear mind or experience to keep together in a body were able to retreat through the enemy cavalry formation as also did some riders of those who had not fled in the beginning. So, "a complete encirclement" does not of course tactically mean that they were "completely surrounded" by an uninterrupted chain of men.

Gary Brueggeman’s insightful website has a very interesting discussion on cavalry attacking the flanks of the infantry. It shows how only a few cavalry can actually attack a flank, with the majority of the cavalry having nothing to attack but thin air. At Cannae, a lot of the African infantry would be in the same position with nothing to attack. A flank attack is synonymous with attacking the flank and rear of a unit. Livy’s comment about the Africans extending does indicate they are being to encircle the Romans. In my book, I do not just state the Africans extend and leave it at that, I go into detail via diagrams as to how in the most simple and fast manner they could accomplish this. Here you will find I am strongly influenced by the Gary Brueggeman approach (don’t just state it, show it). This is where those military manuals come into play. Also in my study of Zama, I don’t state the Romans formed a single line, I detail it in diagrams how it could be done. This exercise produced a little surprise. Same with Ilipa, I give the numbers of the troops involved in the outflanking movement, their formations and how they wheeled, and most importantly, what the outflanking forces were trying to achieve.

I agree with this. I also have such an approach on tactical matters, very mathematical, diagrams, thousands of specially prepared 6mm miniature blocks to play with numbers and ratios, distances and angles. The question regarding Cannae is the shape of the Roman formation once they broke into the enemy. If their line became a crescent or if units broke off, then a flank attack in line is very understandable as there would have been Romans extending over a lot of space in a vertical axis to their original formation. And then we would also have the pushing advancement of the Libyans on the flanks as they marched towards each other. That some of them would be able to, and why not, form at an angle to enclose the enemy I find very possible. The question would be how many and how long they would extend. As much as I see it a natural consequence of events and of Hannibal's plan, I also see it as a secondary maneuver, taking into account that the cavalry had already adequately covered the "ground for escape", giving the Romans a false ray of hope, which is consistent with the tactics we mentioned above.

There has been a tradition among historians of the cavalry completing the encirclement. What are the cavalry doing? Are they standing stationary horse shoulder to shoulder so to form an unbreakable line while dispatching the Roman infantry? This scenario lacks common sense and only possible in the mind of the beholder, like Sir Walter Raleigh. I have come to the conclusion that when the Carthaginian cavalry returned to the battle, they were running down those Romans fleeing the pocket, and were not creating a wall so as to prevent the Romans from fleeing. I also don’t have the Africans completely encircling the Roman infantry. My study shows the Africans don’t have the numbers, therefore the frontage to do this properly. It is only after the pocket has been severely reduced can the Africans accomplish this.

They form a wall in squadrons, which is an "invisible" wall. The Romans think that the way is open, that they can flee, but as they try to run through the open space where they see no enemy cavalry, they are attacked by them. I have given more details on my opinion on this above.

The Roman infantry have the same frontage as the Carthaginian infantry.

Where did you get the figure of 32,000 Celts? Why do you believe the 32,000 Celts are deployed in a single line? What happened to the wedge formation? Taking your army numbers, if the field of Cannae can only allow a frontage of 1000 men, then your 32,000 Celts would be deployed 1000 men wide by 32 ranks deep. Your 50,000 Romans would be deployed 1000 men wide by 50 men deep.

At the Trebia, Hannibal had over 20,000 men, at Cannae over 40,000 men. At the Trebia the Romans had eight legions, at Cannae 16 legions. So both armies had doubled their infantry. At the Trebia the Roman broke through the Carthaginian line. This leaves the question as to why did they believe they had to deepen the depth of the maniples at Cannae.

The over 40,000 men at Cannae included the about 8,000 Libyans waiting behind the crescent. The rest are about 32,000 men. The wedge formation is in itself a battle-line, especially if it was a crescent shaped line as also described instead of detached smaller lines at a distance to each other. The question remains, these 32,000 men extended over a certain length of ground from the apex to the ends of the wings in two axes. Their frontage of course is only the axis parallel to that of the Roman frontage. So, in your model, is the total frontage of the whole Carthaginian infantry formation equal to that of the Romans? The field was quite extended, especially since the battle did not take place at an angle perpendicular to the general river flow (that is east-west) but more like parallel to it (north-south). So the lines themselves extended much longer than the distance between the river and the hills of Cannae. Anyways, we of course are unable to know the dimensions of the real battle-ground but at least according to all descriptions, it must have been quite extensive, whatever that could mean here. It certainly nowhere is described as narrow for such numerous armies.

Yes the first line (the hastati) will have less men than the second line (principes/triarii). However, this does not affect the legion’s frontage. Following Polybius, if you have 1200 hastati in a legion and 1200 principes, and you decide to deploy them 200 men wide, the hastati and principes both have a depth of 6 men. This means the 600 triarii have a depth of three men, and when the principes are combined with the triarii, the principes and triarii have a combined depth of 9 men. The legion then has a frontage of 200 men and a combined depth of 15 men. If you include the 1200 velites, then they would be deployed 200 men wide by 6 men deep. Now the legion has a combined depth of 21 men. In board game jargon, all legions have a zone of control. With a legion having a frontage of 200 men, 14 legions have a frontage of 2800 men. Now you need to implement Polybius’ maniple arrangement for each legion and in doing so each legion and therefore each line within a legion will have the same frontage. So the hastati will end up having the same frontage as the principes and triarii, accept the principes/triarii have a greater depth. The formula I found is when the Romans increase the size of the legion; they increase the depth, not the frontage. In this manner, they do not have to increase the number of centurions or centuries.

Yes, this more or less coincides with my calculations and understanding. I also believe that the extra men are used for added depth. However, the question as to the frontage of the Roman infantry line at Cannae remains, where the line was arrayed much deeper than usual, making the frontage maybe 1/3 to 1/4 what was "normal". This does not mean that there was no regulation ordaining such an event, of course there would have been, but if a line of Hastati (2/7 of a legion) was what made up the front of the Roman line alone, then we are talking about more or less 23,000 men, the rest standing in reserve behind them, a good distance away. If these 23,000 men arrayed in extra deep, dense files as the texts say, then we are talking about a frontage normally held by correspondingly less men. Against them Hannibal arrayed about 32,000 allies, so, in order for the two lines to have the same length of frontage, the Carthaginians should be arrayed even deeper/denser! This is a mathematical problem I cannot solve unless I assume that at least a good number of the men on the rear, if not all, joined the Hastati in the first line. Else, I have a Roman line for some reason much shorter than that of the enemy, which was a tactical possibility, but would have been commented on as a stratagem. How have you worked this issue out, so that the two lines are similar in length?




I will not comment on your theory regarding deployment mathematics, we have been through this before and of course it would be unfair to say anything before I can read it as a whole, although I still do not understand the 6 squadron depth. Do you mean 6 squadrons one behind another as in 6 successive lines? Do you mean that they joined in a single squadron that was 6 times larger and deeper than a single one?

I turned the quotes into italics to make them easier to read, so no worries.

[antiochus]

Macedon wrote:
The Greek text is very clear on this "He arranged the semaeae (maniples) more dense than before and the depth of the speirae (maniples again) (many) times greater than their frontage" So, we are talking about the soldiers in the maniples being arrayed closer to each other (which could be close or compact order) and practically in column. I still cannot see it as arranging the lines closer to each other, without of course claiming that they were not, I have them unite the lines myself. I only speak of the translation of the said text.

Thanks for this. I have been debating to some colleagues that it’s about the men in the maniples being closer than normal, but my colleagues slap me around and say it is also to be interpreted that the lines of maniples were also closer. The reason why I have the lines of hastati separated from the principes/triarii is because the allow intervals in the lines, the principes/triarii have to be further back from the hastati. This still does not prevent the Africans from being able to encircle the Romans. What it does mean is at the beginning the Africans cannot completely encircle the Romans.

Macedon wrote:
So, "a complete encirclement" does not of course tactically mean that they were "completely surrounded" by an uninterrupted chain of men.

Then we on are the same page. Is this the first time a RAT member is in agreement with me? You guys must be mellowing.

Macedon wrote:
So, in your model, is the total frontage of the whole Carthaginian infantry formation equal to that of the Romans?

I have a very different formation for the Carthaginian infantry that does not exactly follow the traditional view of the hollow lambada. If I had to give it a description I would call it a triangle. The base of the triangle is the main body of Carthaginian infantry, while the two sides form the wedge. The Romans will break through the wedge, but it is the role of the main body (the base of the triangle) to hold the Romans while the Africans encircle the Romans. It is the main body that starts to become concave at this point, and not the apex of the wedge as is the common theory. So to answer your question, only the main body (the base of the triangle) is equal in frontage to the Roman infantry. Actually when I look at it, the wedge could technically be also regarded as having the same frontage as the Romans, but stepped back.

Macedon wrote:
If these 23,000 men arrayed in extra deep, dense files as the texts say, then we are talking about a frontage normally held by correspondingly less men. Against them Hannibal arrayed about 32,000 allies, so, in order for the two lines to have the same length of frontage, the Carthaginians should be arrayed even deeper/denser! This is a mathematical problem I cannot solve unless I assume that at least a good number of the men on the rear, if not all, joined the Hastati in the first line. Else, I have a Roman line for some reason much shorter than that of the enemy, which was a tactical possibility, but would have been commented on as a stratagem. How have you worked this issue out, so that the two lines are similar in length?

See my answer above about the main body (the base of the triangle) being equal to the Roman frontage. I did an overview of Cannae some five years ago and could not mentally get my head around the depth of the Carthaginian infantry. I couldn’t understand why the Romans threw away the tactical advantage of having more men. Why couldn’t they form up in their normal deployment depth (and there is one), thereby forcing Hannibal to match it? Like at the Trebia, they would have achieved their break through. However, as Livy and Polybius remind us most of the Roman infantry were raw recruits, the importance of this piece of information cannot be played down. It could be the Roman commanders did not have any confidence in the majority of their army (both infantry and cavalry). A good investigator should not let his own prejudices get in the way, but by following Polybius I was struggling with the depth of the Carthaginian infantry, let alone the Romans. I experimented with expanding the frontage but the numbers I got for the Carthaginian infantry went beyond Polybius comment of just over 40,000 men. I have a good understanding of Polybius’ writing style and a good idea of what “just over” should amount to. Too much expansion of the Roman frontage and the Carthaginian infantry numbers go pear shape. The only way I can get the depth of the Carthaginian infantry to conform to the depth of the infantry I have worked out from the Trebia, Ilipa, the Great Plains and Zama is to adhere to Polybius’ description of the maniple deployment. After studying Bagradas and Cynoscephalae I abandoned by prejudices about the depth of the infantry at Cannae.

Also the Roman command structure at Cannae holds some interesting clues about what their intentions are. Unfortunately modern historians have not given this enough due investigation. Look at who commands what, and then investigate their military career previous to Cannae.

Macedon wrote:
Do you mean 6 squadrons one behind another as in 6 successive lines? Do you mean that they joined in a single squadron that was 6 times larger and deeper than a single one?

The 6 squadrons deep means one squadron behind another in 6 successive lines. For example the 30 squadrons were deployed 5 squadrons wide by six squadrons deep or six lines deep. So with the Roman cavalry being six lines deep, you would have thought they would have had some staying power. So much for Plan A.

Macedon wrote:
I turned the quotes into italics to make them easier to read, so no worries.

Thank you for doing that. Unfortunately I think it will be the norm. I haven’t a clue as to what is going on. I can't even get the smily face to work and I wanted to put in a few of them. I deleted RAT from the favourites list and reinstalled it but the BBcode still does not load. Also the title bars that were originally red are now always blue. This all started when I updated to the latest super doopper internet provider system ASDL premium, no expenses spared system.

Steven

[Macedon]

Your colleagues do have a point although for me the probability that they (and a number of other translators) are right is fairly low. You see, Polybius uses "dense" as an adjective and not as an adverb. The meaning, as I see it, is like in English :

he arranged the maniples denser

he arranged the maniples more densely

In the first instance, which is the case here, the probability that he means the intervals between the maniples is low and the probability that he means the maniples themselves as individual entities is high. In the second it is the exact opposite that is true. I am fairly sure that he means the men in the maniples and that he would have expressed it differently should he mean the intervals. And even in the second instance, he would have meant the intervals, if any, between the maniples of the same line and not those of the lines, where, he would under no circumstances use the word "denser", as they weren't normally dense in the first place. He would have said something like "more closely" instead.

So, in order to make the numbers fit, you chose to array the Carthagnians in practically 2 "lines" instead of one. This is a very bold suggestion that cannot be easily supported by the texts at hand. Is there any hint in the sources to use in your support or is it a conclusion you have come to solely based on other battles of the Carthaginians? I understand that one major reason for you to come to it would be the same why I concluded that the Romans fought in a single line. Mathematics. If Hannibal had arrayed in one line, then the Romans had to also array in one to make the descriptions fit. If you have Hannibal array in two lines, then the Romans are allowed to also array in two but, again, not in three.

How in your model do you explain the fact (with which you agree), namely that the Libyans attacked (flanks and/or rear) the Romans who had funneled into the gap created by the retreating/fleeing Celts? The distance between the two Roman lines was considerable, unless you support that they were for some reason very close together but why would they have done so if they were not joined in one line? So, as the Hastati gradually lost their order in the pursuit, lessening their front as the wings tried to keep in touch with their advancing middle, the second line would still have retained their front and order and would have counterattacked any attempt of the Libyans to engage, especially if they at the same time they extended parallel to the them, practically showing them their backs. Even if the interim distance was not that long, they would not have dissolved their order to just run behind the Hastati. How do you support your opinion that "they would not have been able to stop the Africans"? Why wouldn't they? Wasn't it one of their main functions? They would not have been pinned by the enemy skirmishing cavalry yet, they would have received those of the hastati in flight, they would be fresh and ready.

Regarding the reasons for deploying deeper, they are a number and yes, to help less experienced troops retain order is of course one. In my understanding of battlefield tactics, a deeper formation does not directly affect "attacking" power. It does so indirectly by protecting its order during combat as well as during the charge and by prolonging its staying power by providing more reserves for the tired and the wounded or dead. So, a line in double depth has two first-rankers and two rear-rankers, more men who were supposed to be of the same fighting calibre and quality of arms. One in triple or quadruple depth correspondingly more. I also commented on this in my first post.

[antiochus]

Macedon wrote:
Your colleagues do have a point although for me the probability that they (and a number of other translators) are right is fairly low.

The distance between the lines to me is not really important. Anything anyone, including myself writes about the distances is conjecture. What is not conjecture is Polybius’ description of the maniple deployment that is mathematics, and extremely important mathematics. From it I can determine the frontage of the Roman legions.

Macedon wrote:
If Hannibal had arrayed in one line, then the Romans had to also array in one to make the descriptions fit. If you have Hannibal array in two lines, then the Romans are allowed to also array in two but, again, not in three.

I am finding this a little confusing. I have the Carthaginians deployed in what would resemble a wedge, with a base…. a triangle so to speak. So are you referring to the wedge proportion as the first line, and the base of the triangle as the second line?

Macedon wrote:
How in your model do you explain the fact (with which you agree), namely that the Libyans attacked (flanks and/or rear) the Romans who had funneled into the gap created by the retreating/fleeing Celts? The distance between the two Roman lines was considerable,

How is the distance between the two Roman lines “considerable?” We are not told. Are you are expecting me to defend my theory against your theory, which you believe your theory to be the benchmark?

Macedon wrote:
Even if the interim distance was not that long, they would not have dissolved their order to just run behind the Hastati. How do you support your opinion that "they would not have been able to stop the Africans"? Why wouldn't they? Wasn't it one of their main functions?

My theory does not centre on the Romans not being able to stop the Africans, the ancient sources tell us the Romans were defeated. So their attempt at trying to stop the Africans obviously failed. In the primary sources the triarii are mentioned as being camp guards, ditch diggers and the reserve. It is not recorded if their function was to protect the rear. Why are you of the belief that if the Romans were in two lines (hastati first line, principes/triarii the second line), that the Africans could not encircle them.

At the first battle of Philippi in 42 BC, the Romans are deployed in three lines, yet Brutus’ legions managed to attack Octavian’s flank and defeat him. Brutus praises the valour of his soldiers in defeating Octavian’s famed fourth legion “on which their wing placed its reliance.” Here the third line failed to protect the army. At Pharsalus Pompey’s third line failed to prevent his left flank from being attacked. At the Allia, it would appear the triarii failed to stop the Celts attacking them in the flank and rear.

What do you envisage when the Romans at Cannae are attacked in the flank and rear? Are you expecting my theory of a second line to do some amazing and speedy redeployment? I have no problems with those troops closer to the threat trying to change direction to met the situation, but those in the centre, will they also react with the same speed, ease and orderly manner as miniature wargamers portray ancient warfare? In reality, you now have to factor in confusion, and there are too many accounts in the primary sources of armies being thrown into confusion to ignore. Also we have no idea that after breaking through the apex of the wedge if the hastati were replaced by the principes/triarii. There is absolutely nothing that tells us this, a point I make clear in my book.

No matter how hard I try, no matter how many sheep and bulls I sacrifice to the gods, no matter how many saecular games I perform, when it come to the Roman military I cannot shake of the Pythagorean influence or geometry. At Cannae, because there are only 14 legions facing Hannibal, with the other two attacking Hannibal’s camp, I thought I had a good example of the Romans not adhering to Pythagorean geometry but the Romans proved me wrong. Why? Because the foundations are Pythagorean so it’s pointless trying to get away from it. Stupid me! But in my ignorance I knew the 14 legions at Cannae do not accord with the 3:2 ratio, which is the most defining ratio governing the tribes and the legions. However, when I examined in depth the Roman command structure, only then did I realise the 14 legions conform to the other sacred Pythagorean ratio 4:3, which is the perfect fourth. The perfect fourth represents the 8:6 of the Pythagorean tetrachord 6:8:9:12 which adds up to 35 (termed harmonic 35). So the 8:6 part adds to 14, or the 14 legions. Now I married this with the Roman command structure and deducted of the 14 legions, eight legions can be designated in the centre for a breakthrough and the three legions on each flank are the flank guard legions. So the 8 legions to 6 legions when proportioned to two, equates to 4:3.

Like you I use mathematics to reconstruct what could have transpired. As Hannibal is reacting to the Roman deployment, I based the apex of the Carthaginian wedge to match the frontage of those eight legions.

I have the Romans break the apex of the wedge first (the eight legions), and then the Carthaginian flank guards. The next stage is as Livy writes the Romans:

“following up the Celts (the eight legions) and pressing on to the centre and that part of the enemy's line (the base of the triangle or the main body), which was giving way, progressed so far that they (the eight legions) now had the heavy-armed Africans on both of their flanks.

The Roman formation is now a wedge, with eight legions forward. The three legions on the flanks of the wedge are breaking through the Carthaginian flank guards and therefore, falling behind the eight legions. I believe Hannibal formed PART of his army into a wedge with the sole purpose of it being defeated and in the process the victorious Romans would end up advancing in a wedge shape. If we take the premise the hastati from the eight legions that did the break through and continued advancing, the principes and triarii of the eight legions would through their standard doctrine, keep in contact with their units of hastati. We see this doctrine not obeyed at Zama, when the principes and triarii were ordered not to advance in support of the hastati through fear of being disordered.

Those eight legions in the centre, because they were forward of the flank guard legions would have realised with the main body engaging them and the Africans now on the flanks of the army, there were in an extremely precarious position. Plutarch has the centre retire, while Appian has the left flank rout first followed by the right flank.

That’s my theory. However, I will add that although I have said the Carthaginian infantry formation was a triangle, the actual formation I have is nothing like what other historians have reconstructed. I like what I have because it makes sense of all the primary sources.

On a side note have you read Goldsworthy’s account of Cannae (Cassell and co)? Goldsworthy states that “The Romans surged forward until they were level with the starting position of the flanks of Hannibal's main line and kept going, for there was nothing to oppose them.” Here the Romans have broken through and there is nothing to stop them. Then suddenly he writes “The Libyan's charge stopped the Roman advance dead, robbing it of all momentum. There were now no organized reserves in the Roman army to feed into the combat and renew the surge forward.”

First is there is nothing to stop them, then why does Goldsworthy believe there now was no reserves to renew the surge forward? I’m really puzzled as to how the Libyan attack on the flank stopped the Roman advance dead. At the Trebia when the Romans broke through, realising they could not help the rest of the army they made their way to Placentia. They did this because there was no one facing them to stop them. But at Cannae, with Goldsworthy declaring there is nothing to stop them, they are stopped by the flank charge. Goldsworthy claims “The two phalanxes of Libyans gripped the Romans like a vice, and around them the surviving Celtic and Spanish troops, joined perhaps after a while by some of the routers, pressed round to add to the fighting line.” So we now have the whole Roman, with nothing in front of them to stop them, locked in a vice. Are they being jammed up shoulder older to shoulder along the line that is the reason they can’t move? In this scenario wouldn’t a centurion order some of his men to turn sideways and step out of this vice and give everyone else some room so they can move forward? Now in order to explain why the Romans don’t move, Goldsworthy wants us to believe those routed Celts and Spanish troops came running back to rejoin the army and add to the Romans misery.

This rather poorly understood scenario has occurred because too many historians fail to acknowledge Hannibal main line formed the base of the wedge and this main line had the same frontage as the Romans. Every time I read about the battle of Cannae by an author following the doctrine the Carthaginians deployed in a hollow lambada (every one’s doing it so it must be right), in order to make it work, they end up producing a scenario devoid of command sense.

Steven

[Timianus]

This thread need pictures.

[d.carmichael]

This discussion is way beyond me. I would need pictures and diagrams to understand it. Sounds very mathmatical. I'm modern day military and I just wonder how we can rely on mathmatics when it comes to a battle. Yeah sure probably with the army structure theres probably mathmatical fomulas with how it was organized. From my experiance being a soldier on the battlefield the story changes dramatically. Once the battle is commenced communication can sometimes become troublesome even with our modern day tech. I can only immagine how difficult it must have been on the battlefield at cannae. I can easily see a commander back in those days form his troops up for battle but once it starts and the chaos sets in, for him to nearly loose most control and just pray what efforts he made before battle was commcenced works in his favour. I read somewhere in the tons of reading on this forum that the roman troops were mostly raw recruits. That must have made things even worse for their commanders. Too many times have I seen confusion and communication break down very quickly while on an exercise or hell even on the parade square with new soldiers. It can take a year or more of training to get a soldier to a point where he can be effective. We can speculate all day on how the lines were drawn up but at the end of the day we can't know how the battle was actually carried out. Just my thoughts.

[Timianus]

This discussion is way beyond me. I would need pictures and diagrams to understand it.

I am a little lost as well. I had always assumed (perhaps incorrectly) that the main body of the roman army at Cannae was stacked unusally deep--i.e. there were more than 3 "ranks" (or lines) in each "file." Maybe I my terminology is wrong, but when speaking of "lines", I always assumed that to mean a single horizontal line or "rank" and that there were 3 lines in each maniple (Hastati, Princepes, and triarii)

Again, I am probably wrong, but I always thought the problem at Cannae was that the romans arranged in more of a rectangular phalanx formation rather than their traditional 3-lined maniples. In other words, they stacked their maniples. But, this obviously doesn't account for the view that the triarii were left back to guard the camp. if this were the case, then there would be only 2-line maniples, which still could have been stacked.

[Macedon]

I am finding this a little confusing. I have the Carthaginians deployed in what would resemble a wedge, with a base…. a triangle so to speak. So are you referring to the wedge proportion as the first line, and the base of the triangle as the second line?

Yes, what you are describing is practically a two "line" formation, one in wedge and one in line. This is what I mean.

How is the distance between the two Roman lines “considerable?” We are not told. Are you are expecting me to defend my theory against your theory, which you believe your theory to be the benchmark?

My theory does not centre on the Romans not being able to stop the Africans, the ancient sources tell us the Romans were defeated. So their attempt at trying to stop the Africans obviously failed. In the primary sources the triarii are mentioned as being camp guards, ditch diggers and the reserve. It is not recorded if their function was to protect the rear. Why are you of the belief that if the Romans were in two lines (hastati first line, principes/triarii the second line), that the Africans could not encircle them.

No, I make no comparisons here, I just want to understand the mechanics you used and the answers you gave to specific tactical questions. I find tactical issues extremely interesting and I am always eager to understand how other same-minded people understand them. The normal distance between lines in such multiple line formations is usually described in hundreds of yards and one thing all accounts have in common is that the auxiliary (rear) lines keep their shape and do not just follow the units in front of them, dissolving the integrity of their own lines in the process until they are called upon to reinforce the front either as units or as a whole line (again depends on theories and circumstances) . To me, assuming that the rear line was present and that it had not already dispersed in flight, it would be impossible to effectively destroy the Roman first line, as the men in flight would find refuge in the second line or else the rear line would advance against the maneuvering Africans and practically force them away, being a fresh, orderly formed line advancing against an opponent in movement, extending or not, receiving an attack against its flank or rear. I cannot see them so close to the front line in order to also be trapped by the Africans. Of course, many hypotheses could be voiced as to how things could fare in order for the outcome to be the same, but such action would be so critical that I would have expected some relative mentioning in the accounts.

Regarding the triarii, I also believe that they were the 10,000 men tasked with camp guard duty but I have to say that unfortunately they are not positively identified in the sources.

At the first battle of Philippi in 42 BC, the Romans are deployed in three lines, yet Brutus’ legions managed to attack Octavian’s flank and defeat him. Brutus praises the valour of his soldiers in defeating Octavian’s famed fourth legion “on which their wing placed its reliance.” Here the third line failed to protect the army. At Pharsalus Pompey’s third line failed to prevent his left flank from being attacked. At the Allia, it would appear the triarii failed to stop the Celts attacking them in the flank and rear.

It is one thing to somehow fail to prevent a flank attack and another to be trapped with the first line. In the many accounts of the role of auxiliary lines, the countering of such action either with force or by just receiving the fugitives as they retreat is among the tactical reasons offered for their existence. As I already said, of course they could fail, but I would have expected some relative action in the accounts, unless there is such and I just do not link it with it as, undoubtedly, I do have strong preconceptions because of my work as I guess you and most others like us do. This is why such discussions can be interesting as they can offer a fresh look into things.

What do you envisage when the Romans at Cannae are attacked in the flank and rear? Are you expecting my theory of a second line to do some amazing and speedy redeployment? I have no problems with those troops closer to the threat trying to change direction to met the situation, but those in the centre, will they also react with the same speed, ease and orderly manner as miniature wargamers portray ancient warfare? In reality, you now have to factor in confusion, and there are too many accounts in the primary sources of armies being thrown into confusion to ignore. Also we have no idea that after breaking through the apex of the wedge if the hastati were replaced by the principes/triarii. There is absolutely nothing that tells us this, a point I make clear in my book.

As I hope I made clear above, I do not really disagree with the possibility of something like that happening. However, tactics and deployment rest on mathematical probabilities too. It is not just the lack of supporting evidence that troubles me, it is the tactical risk included for Hannibal's Africans as they extend before or behind a fresh, unpinned Roman line, as you describe the action, making it a very dangerous maneuver to have been organized beforehand. It is more a decision I would expect a general to have made as a reaction to exploit an opportunity, rather than a preplanned one.

the principes and triarii of the eight legions would through their standard doctrine, keep in contact with their units of hastati. We see this doctrine not obeyed at Zama, when the principes and triarii were ordered not to advance in support of the hastati through fear of being disordered.

Yes, this is a key difference between our understandings and the center of my "objections". As I perceive it, the doctrine of the rear lines was to keep their line and NOT follow the hastati in front of them unless ordered to. Keeping in mind that up to that point the hastati were fully victorious in their eyes, I expect the rear line to not have sent many reinforcements to them, so their integrity cannot have suffered much. If, as you suggest, each unit (of whatever predetermined size) kept a constant distance with the men in front of them, then the shape of the auxiliary line formation would have changed in the same manner the front line did making your suggestion much more probable as parts of the rear would even be as advanced if not more than the wings of the front line.

That’s my theory. However, I will add that although I have said the Carthaginian infantry formation was a triangle, the actual formation I have is nothing like what other historians have reconstructed. I like what I have because it makes sense of all the primary sources.

Yes, I know, I have read many modern descriptions, Daly's included, and most have many obvious problems ranging from a lacking mathematical approach to over-reliance on other people's translations leading to pre- (mis) conceptions. The one line scenario is mentioned by a number of scholars as a possibility, although a very few have an approach that is even remotely close to ours and by that I mean our mathematical reasoning.

[Macedon]

I am a little lost as well. I had always assumed (perhaps incorrectly) that the main body of the roman army at Cannae was stacked unusally deep--i.e. there were more than 3 "ranks" (or lines) in each "file." Maybe I my terminology is wrong, but when speaking of "lines", I always assumed that to mean a single horizontal line or "rank" and that there were 3 lines in each maniple (Hastati, Princepes, and triarii)

Again, I am probably wrong, but I always thought the problem at Cannae was that the romans arranged in more of a rectangular phalanx formation rather than their traditional 3-lined maniples. In other words, they stacked their maniples. But, this obviously doesn't account for the view that the triarii were left back to guard the camp. if this were the case, then there would be only 2-line maniples, which still could have been stacked.

When referring to "lines", usually "battle-lines" are meant and not ranks. So, the triple acies system is one with three lines of x ranks each. In this discussion, this is how we used the term. In order for them to be called "lines" a rectangular shape is not required. A line could have any shape as long as there was tactical "cohesion" among its units. Even gaps could exist in them. When the term "phalanx" is used, usually gaps are not supposed to exist unless very narrow and/or concealed with shallow lines of men. There was no "problem" in Cannae for the Romans. They had a plan, they lost. It is us who have the problem reconstructing with certainty the various accounts into a single proposal so that we can identify the tactics that led to their defeat. My discussion with Steven is centered on the role of the rear line as I believe there was none while he believes there was one. Both formations (and any other) could have led to the Roman defeat but the tactical challenges of course are different.

From my experiance being a soldier on the battlefield the story changes dramatically. Once the battle is commenced communication can sometimes become troublesome even with our modern day tech. I can only immagine how difficult it must have been on the battlefield at cannae. I can easily see a commander back in those days form his troops up for battle but once it starts and the chaos sets in, for him to nearly loose most control and just pray what efforts he made before battle was commcenced works in his favour. I read somewhere in the tons of reading on this forum that the roman troops were mostly raw recruits. That must have made things even worse for their commanders. Too many times have I seen confusion and communication break down very quickly while on an exercise or hell even on the parade square with new soldiers. It can take a year or more of training to get a soldier to a point where he can be effective. We can speculate all day on how the lines were drawn up but at the end of the day we can't know how the battle was actually carried out. Just my thoughts.

Mathematics was according to the ancients a key skill for commanders. Contrary to military action today, pitched battles were very mathematical in their nature as far as the shapes of the formations are concerned. Distances, intervals and other mathematical dimensions are very important to understand how and why armies formed for battle. Chaos could ensue but was not nearly as usual and extended as suggested in movies, at least according to the sources, my and many others' understanding. Usually, chaos, that is full disorder, meant defeat and it was a signal for flight. Many tactical decisions, among them added depth as discussed here, were taken to guard against disorder.

Surely, images would be helpful, if you are really interested in following this discussion I could make some during the next days.