The Frontage of the Roman Army at Cannae

[DBS]

I've not time to check Livy but from memory the last big campaign was Battle of Telamon in 225BC .

Oh, the campaigns against the Gauls in northern Italy continued for several years after Telamon, the Romans never missing the opportunity to kick an enemy when they were down, and 220-219 of course saw the Second Illyrian War with trans-Adriatic expeditionary fun.  Not massive forces - standard consular armies it would seem - but still meaning up to four legions and four alae in the field each year in all probability, so the usual annual tempo of Republican state violence visited on the neighbours.

[stevenneate]

Cannae - the battle that just keeps on giving! This would make the Punic Wars Republican Roman army the army that just keeps on giving?!

Enjoying the discussion gentlemen. Whilst we can use the old adage to respectively agree to disagree, I love that we are not only still discussing it but that we are still so passionate about it.

Hannibal would be so proud that we are still talking about him. Even Varro's ghost might take some comfort that he is still remembered?!

[Monad]

Pardon me for expecting some precision in language, especially when used in connection with a fallacious extrapolation to a supposed Pythagorean ratio.

Oh what is your problem now? Many a scholar and musicologist are well aware of the two basic Pythagorean ratios, so they are not what you describe as "supposed." David, can we have an agreement not to communicate with each other.

Sure, but an interpretation based on what?

Based on my research. I do believe I am entitled to form my own conclusions. Based on some of the replies here, I am doing nothing out of the ordinary.

I disagree. A mid-republican legion deployed about 200 yards wide. In this case a legion deployed 100 yards wide, i.e. at twice its habitual depth for half its habitual width. My take is that the frontage of the battlefield at Cannae was a little under 3km, allowing infantry and cavalry to deploy deep, the depth of the latter hopefully negating the Carthaginian numerical superiority in cavalry. But we can argue about this.

We don't have to argue. We just disagree.

Nobody is claiming exact numbers. Elsewhere in his History Livy gives numbers for legionary strength as 4,000, 4,200, 5,000, 5,200 and 6,000. About 4,000 means the legions were augmented by about 1,000 men each which is all we need to know.

If I said to you, I want you to financially invest in a machine that I have invented that no matter what you put in it, even your household garbage, it will turn it into gold. Logically, before making such a financial investment, you would want proof. And yet, when it comes to Polybius and his various sizes for the legion, there has been no proper investigation into the matter. It is not like there is a paucity of data available to use.

Does Polybius' legion sizes hold up? Have you actually tested to see if the additional 1,000 troops are actually 1,000 troops? What if the number was 1,200 troops. What if the figure of 4,000 troops is also a rounded number? What period of Rome's military history has Polybius acquired such numbers from? Could Polybius' legion of "about"4,000 infantry could belong to a garrison legion? Could Polybius' legion of "about" 4,000 infantry be related to the camp guards being omitted?

I have meticulously examined all of Polybius' legion sizes and have compared with the mass of data we have available at our fingertips, and what I found is it shows that Polybius is very confused about the machinations of the Roman legion, especially in regard to Roman military procedures.

His legion of "about" 4,000 infantry is missing a detachment that generally are assigned to garrison duty. His legion of 5,200 infantry is his legion of 4,000 infantry, which omits the camp guards, having the 1,200 velites added to it. His legion of 5,000 infantry has been rounded from 5,040 infantry. As to the legion of 6,000, can you provide a reference, as I'm not sure if you are referring to the Third Macedonian War period or something else.

Cannae - the battle that just keeps on giving! This would make the Punic Wars Republican Roman army the army that just keeps on giving?!

Steven, sorry about not quickly replying to your message on academia. I wanted to ask you if I could send to you (electronic), Volume I of my research. I can also send you the full Cannae section, which explains or shows how Livy arrived at 45,500 infantry causalities for the Romans and other numbers I purposely left out of my paper on the Roman army at Cannae. I am keeping somethings tight.

[DBS]

My problem is when you think that numerical ratios ascribed on a rather dodgy basis to Pythagoras had anything, anything, to do with military organisation and deployment. Armies comprised the number of men that were available and thought necessary for the task in hand, recognising that there were severe penalties to having too many men in arms, let alone in the same spot.

That is the whole point of Cannae; the Romans had been humiliated in the previous two years, so broke all the rules by raising an eight legion army to destroy Hannibal. And failed.

There were eight legions at Cannae. The allied alae were not legions; the Romans did not class them as such, and did not call them that. To then claim that there was a Pythagorean ratio between eight legions and six legions is palpable nonsense as a) the six "legions" were not legions, and b) there were eight alae anyway.

Furthermore, no general is going to think that he ought to put two of his formations the other side of the river to align with mathematical precepts attributed to a long dead Greek, who probably had never come up with such precepts anyway but had simply had them foisted on his memory by later generations...

Returning to the question of frontage, Justin makes some excellent points. I would further observe that we do not know how the alae were organised or deployed. We cannot simply assume that they matched the Roman legions in this regard, since all we are told is that they matched approximately the Romans in terms of numbers of infantry levied, and triple cavalry. The number of cavalry, necessarily drawn from the better off, alone suggests that maybe, maybe, the infantry was less well equipped and that concepts such as triarii and principes were not necessarily observed. We just do not know. My point is that this uncertainty is going to be a significant qualification on any calculations of frontage.

[Monad]

My problem is when you think that numerical ratios ascribed on a rather dodgy basis to Pythagoras had anything, anything, to do with military organisation and deployment.

Would you like me to send to you Volume 1 of my research?
 

[DBS]

No thank you, because the very idea that Pythagoras had anything to do with military organisation and deployment is self evidentially nonsense.  Look at enough numbers, discarding those you do not like, from enough different sources without regard to the context in which those sources were written, and you will see ostensible patterns.  Those patterns are likely false, or at best driven by other factors.  Throughout mankind's history of violence towards mankind, the manner in which armies have been organised has been driven by a lot of different, but often fairly constant, factors: availability of manpower; availability of equipment; competing demands (such as agriculture); social classes and structures; environmental considerations; authority of leaders; practicalities of command and control.  Unit sizes are dictated by what is achievable and what works for your style of warfare.  Mathematical principles do not come into it.  Furthermore, you must be well aware that considerable doubt has been cast as to whether Pythagoras was actually that interested in mathematics, as opposed to more general and rather wierd social philosophy, and the suspicion that a lot of his supposed mathematical precepts were actually conjured up by later generations and posthumously attributed to him.

[Justin Swanton]

Sure, but an interpretation based on what?

Based on my research. I do believe I am entitled to form my own conclusions. Based on some of the replies here, I am doing nothing out of the ordinary.

You mistake me, my good sir! This is a free discussion forum and conclusions are welcome within the normal bounds of common courtesy. But evidence for those conclusions is also welcome so don't feel shy about giving it. 

Nobody is claiming exact numbers. Elsewhere in his History Livy gives numbers for legionary strength as 4,000, 4,200, 5,000, 5,200 and 6,000. About 4,000 means the legions were augmented by about 1,000 men each which is all we need to know.

If I said to you, I want you to financially invest in a machine that I have invented that no matter what you put in it, even your household garbage, it will turn it into gold. Logically, before making such a financial investment, you would want proof. And yet, when it comes to Polybius and his various sizes for the legion, there has been no proper investigation into the matter. It is not like there is a paucity of data available to use.

Does Polybius' legion sizes hold up? Have you actually tested to see if the additional 1,000 troops are actually 1,000 troops? What if the number was 1,200 troops. What if the figure of 4,000 troops is also a rounded number? What period of Rome's military history has Polybius acquired such numbers from? Could Polybius' legion of "about"4,000 infantry could belong to a garrison legion? Could Polybius' legion of "about" 4,000 infantry be related to the camp guards being omitted?

I have meticulously examined all of Polybius' legion sizes and have compared with the mass of data we have available at our fingertips, and what I found is it shows that Polybius is very confused about the machinations of the Roman legion, especially in regard to Roman military procedures.

His legion of "about" 4,000 infantry is missing a detachment that generally are assigned to garrison duty. His legion of 5,200 infantry is his legion of 4,000 infantry, which omits the camp guards, having the 1,200 velites added to it. His legion of 5,000 infantry has been rounded from 5,040 infantry. As to the legion of 6,000, can you provide a reference, as I'm not sure if you are referring to the Third Macedonian War period or something else.

It is Livy who gives the legion sizes although Polybius details how a legion was numerically expanded. Here are the references from Livy:

4,000 twice (History: 21.17; 28.28),
5,000 four times (23.34; 26.28; 39.38; 41.21),
5,200 five times (40.18; 40.36; 41.9; 42.31; 43.12),
6,000 twice (42.31; 43.12),
and 6,200 once (24.34).

Polybius makes clear that a legion was expanded by a proportional increase in the numbers of hastati, principes (and by implication velites), but not of triarii. The smallest size for a legion was 4000 men:

The division is made in such proportions that the senior men, called triarii, should number six hundred, the principes twelve hundred, the hastati twelve hundred, and that all the rest as the youngest should be reckoned among the velites. And if the whole number of the legion is more than four thousand, they vary the numbers of these divisions proportionally, except those of the triarii, which is always the same.- Histories, 6:21.

That means 1000 velites in a 4000-man legion. A legionary occupied a frontage of one yard. If the legion was 200 yards wide that meant 6 ranks of hastati, 6 ranks of principes, 3 ranks of triarii and 5 ranks of velites. The velites were divided amongst the other three types, which suggests 2 ranks of velites with the hastati, 2 ranks with the principes and 1 rank with the triarii.

A veles was in effect a peltast, which meant he was capable of hand-to-hand combat at a pinch, and if he occupied the rear of each of the three lines that would bring the hastati and principes up to 8 ranks - the minimum depth for a typical infantry line, (necessary to prevent cavalry charging right through it amongst other reasons).

The triarii were in a shallower line but they were the line of last resort and in any case would have the depth of the hastati and principes behind them.

5000 men would be achieved by adding 2 ranks of hastati, 2 ranks of principes and a rank of velites (attached to the triarii), bringing up the depth of the hastati and principes (including velites) to 10 ranks - which incidentally was the depth of the first two lines of the later Marian legion.

5200 men would be achieved by adding another rank of velites.

6000 men would be achieved by adding another 2 ranks of hastati and 2 ranks of principes.

6200 men would be achieved by adding another rank of velites.

All very practical.

[Erpingham]

Not particularly wanting to get too moderatorial but I think the side discussion of the historicity and relevance of Pythagoras might be generating more heat than light. Steven has a well-developed and densely argued thesis on this elsewhere and it doesn't directly impact a discussion of frontage. 

All seem agreed that 14 legion-sized bodies are lined up next to each other.  We know the formation is deeper than usual, partly through tactical preference but also perhaps to fit in the space available (see the epic discussion on "Where the hell was the Aufidius in 216?" elsewhere on the forum (short answer we don't know but we can make some guesses).  The discussion, for those interested, is here but it meanders a bit (sorry, couldn't resist  ) before really getting into the mapping part.

Justin has laid out his rationale for his version of the frontage. Steven has given his but in less detail.  They give very different estimates. I note from the Aufidius discussion (yes, I refreshed my memory) that Goldsworthy also makes an estimate similar to Justin's (and put the river in a similar place).  As this battle has been discussed by many, many historians there may well be other estimates (with rationales) out there that can be considered and critiqued to help us scope the problem better.

[Monad]

Polybius makes clear that a legion was expanded by a proportional increase in the numbers of hastati, principes (and by implication velites), but not of triarii. The smallest size for a legion was 4000 men:

First time inserting an image. Don't know how to control the size. I calculate a legion of the Second Punic War has having 4,800 infantry made up of 1,200 velites, 1,800 hastati, 1,200 princeps and 600 triarii, arrayed five maniples wide by six maniples deep:

You cannot view this attachment.



Princeps and triarii belong to Class I. Each battle line has 600 infantry organised into 10 centuries each of 60 infantry, and all under the command of a military tribune. Well, Polybius does tell us that there were six military tribunes to a legion, that is why I call them tribune cohorts, so as to stop any confusion. And in my legion, they are not glorified messengers are many academics like to tell me. Now of the 1,200 velites, I allocate 200 to each tribune cohort (20 per century), so each tribune cohort amounts to 800 infantry. And then I have the triarii tribune cohort guard the camp, and hello, I am left with a legion of 4,000 infantry.

What about Polybius' legion of "ABOUT" 4,000 infantry. Well, I am now going to cut the legion horizontally. I want a detached to go and do something else, and like Napoleon's corps system, to be a mini army, so let's cut up the legion horizontally. Now in order not to upset some, I will avoid the Pythagorean principle that governs what I am doing, and that is by now employing the 10-cohort organisation, I am deducting two cohorts of 480 infantry, which in the time of Augustus would be called a vexillation and take it away from the 10 cohorts of 480 infantry, leaving me with 8 cohorts of 480 infantry or 3,840 infantry, consisting of:

 

960 velites
1440 hastati
  960 princeps
  480 triarii
3840 infantry

Following the premise that a century had 80 infantry, 3,840 divided by 80 equals 48 centuries. Now to my many admirers that claim I cherry pick information, let's go for check mate. In 209 BC, Livy writes that the consul Quintus Fulvius Flaccus ordered his son to give the proconsul M. Valerius Laevinus in Sicily a body 4,344 men, taken from the consular army of Quintus Fulvius Flaccus. Now what happens when we add the officers, musicians and cavalry, the result is:

 

960 velites
1440 hastati
  960 princeps
  480 triarii
3840 infantry
  48 centurions
  48 optiones
  48 musicians
3984 infantry
  120 Roman cavalry
  240 allied cavalry
4344 men

All of you have no idea of how prevalent the 48-century legion is in the ancient sources. I have tons of evidence and all supporting each other about the Roman army. Masses of data has been examined, processed, examined again and again, and all the pieces continually show the same thing. The end result is there is no legion of 5,200 men, which I am able to prove countless times via Livy's mass of data relating to the replacement system, which no one has bothered to examine. And there is no legion of 6,200 infantry. However, there is a legion of 5,400 infantry. And to support all this, I have studied all the data concerning the Roman navy and again, it all interlocks. Oh, and there is a period when the Roman legion had no light infantry. They believed that role could be performed by the hastati of Class IV, which being the youngest, would be suitable. Well, later the Romans realised their mistake. The Class IV hastati are the same age bracket as the velites, and that is why Polybius omits them, he just got confused between them.

Justin, you have produced a book and will go to great lengths to defend it, and your reputation, which means we will never agree. However, I believe I have got it right. I provide far greater detail than any other academic, and this is partly due to uncovering the internal structure of the Roman tribes, which has allowed me to reconstruct the Roman levy system from 513 BC to the end of the republic. That is my advantage over many. Join the list of people who hate me for it.

[Justin Swanton]

My overall impression is that this construct requires essentially a rejection what major sources like Livy and Polybius say about the size and organisation of the legion. Size - they give a variety of sizes except 4800 men. Organisation - nowhere does anyone affirm there were more hastati than principes, and above all that the legion consisted of 6 lines of hastati, principes and triarii. The earlier legion had 5 lines if you include the rorarii and accensii but that was a different beast.

My own approach is to assume the sources knew a great deal more about their subject matter than we do and so I try to reconcile their apparent contradictions as much as possible, only conceding a source is wrong if it is clearly demonstrated to be so. When it comes to the legion, I'm quite happy that the sources give enough coherent information to enable us to reconstruct the legion without us having to reject whole swathes of what they affirm.

Oh yes, my book - I don't in the least mind being proven wrong (I've already abandoned a few assumptions I made in it) but you're going to have to prove me wrong by reference to the sources, not by an appeal to Pythagorean mathematics, sorry.

That figure of 4344 men piqued my interest. I found the reference in Livy: "The consul Fabius ordered his son Quintus to take to M. Valerius, the proconsul in Sicily, the remains, so far as they had been got together, of the army of Fulvius. They amounted to 4344 men." - History, 27:8.13. This is the remnants of an army, not a regular structured legion!

[DBS]

There is no mention whatsoever of Classes in the Punic Wars.  Nor should there be; the distinction between velites and the heavy infantry is by age and wealth, and then within the heavy infantry the three types are distinguished by broad age bands.  What Steven is proposing is a hopeless melange of the supposed "late monarchy" army (leaving aside the question of whether Servius Tullius ever existed and whether Livy had any reliable basis for his description five and a bit centuries later) with the contemporary description by an experienced military observer, Polybius, of the mid second century BC legion.   There is absolutely no evidence whatsoever for a 48 century legion.  Just dividing supposed strengths of forces by a number to come up with 48 is not evidence, because it completely ignores the fact that throughout military history, hardly any unit has maintained its theoretical establishment once on campaign, even assuming that it was raised to theoretical establishment in the first place.

As Justin says, just ignoring sources like Polybius (or worse, as Steven does, dismissing him as a fool) is a very poor method and would be laughed out of court in any academic environment.

[Erpingham]

Just running through this again, I've noted that, for all there differences, Steven and Justin assign the same frontage to the legions at Cannae (100 men).

Justin, using Polybius' narrower frontage (I believe), equates this to 100yds, giving 1400 yds. He switches to km for the battleline length, but 3km is about 3,300 yds, so 1,900 yds for the cavalry.

Steven hasn't given a frontage figure for his legions but it must be less than the figure used by Justin.  In order for a 1400 yd battle line, the cavalry must also be at a narrower frontage than Justin allows.  Or have I just misunderstood the proposed deployment? Perhaps Steven can confirm the figures he has used for his calculation, so we can compare the two better?