Roman Military formations

[MonsGraupius]

Hi, I've been researching the battle of Mons Graupius and one of the key things that can be used to identify the site is the size of the Roman battle "front".

We know the numbers: 8000 auxiliary and 3000 cavalry. The auxiliary included Germans - but that may be too much detail.

I've read several places that the Roman army was in a checker-board formation - the checker boxes where three deep (hastati principes and triarii) and that each was 8 deep.

It sounds as if that is an established fact, but when I went to find the source, I couldn't find anything.

So this question is in four parts:
1. Do we actually have any source telling us how deep each formation was?
2. If not, what is the best guess
3. How would this figure be affected by German auxiliary troops

4. For battle re-enactors. If the line is too deep - don't you risk crush injuries, and if we really don't know, isn't this a good indication of the maximum depth?

[Frank]

1. Do we actually have any source telling us how deep each formation was?
2. If not, what is the best guess
3. How would this figure be affected by German auxiliary troops

1. it depends on the situation, military necessities and the decision of the commander
2. 4-10
3. see 1

[MonsGraupius]

Thanks. The Tactitus account suggests the line was "extended" .. "Agricola, fearing that from the enemy's superiority of force he would be simultaneously attacked in front and on the flanks, widened his ranks, and though his line was likely to be too extended".

I've seen a calculation using the idea that the line was extended from 8-deep to 6-deep. I tried to find the source of this, but was unable to find anything.

So, can the experts here tell me whether these seem sensible figures?

[ValentinianVictrix]

It might be worth having a look at Vegetius. He was writing much later but used military manuals much earlier than when he wrote his Epitome.

[Nathan Ross]

I've read several places that the Roman army was in a checker-board formation... when I went to find the source, I couldn't find anything... Do we actually have any source telling us how deep each formation was?

I think the 'checkerboard' or quincunx is an interpretation of the description in Livy of the arrangement of the republican manipular legion. This would be different (although how different we don't know) to a late republican or imperial arrangement. Ross Cowan's Roman Battle Tactics explains what we know (or think we know) pretty well.

For formation depth, Arrian's Array Against the Alans is a good early imperial source and says eight deep. We don't know how usual this was. It might be connected to marching column width or contubernium size (both of which seem to vary between 6 in the republic and 8 under the empire).

The Tactitus account suggests the line was "extended"... I've seen a calculation using the idea that the line was extended from 8-deep to 6-deep.

Unlikely the line depth changed - all the men in a formation would have part of the same unit, and asking the rear two of them to split off, run to the flanks and form a new unit while on a battlefield would be difficult to say the least.

Two possibilities - DB Campbell, in Mons Graupius (which you've probably read) translates the phrase as 'opened out' and suggests a change in formation. We could imagine the men on either flank spreading out and the rest of the formation opening after them, like a bellows, leaving larger gaps between each man and the next. Still difficult to do effectively though, I'd say.

Alternatively, if we imagine that Agricola had his c.16 cohorts of auxiliaries in two lines, each drawn up approx. 60 men wide by 8 deep, with eight cohorts in the first line and eight in reserve (or ten in front and six behind, etc), he could order four reserve cohorts to deploy on each flank to double the length of the front line.

[Robert Vermaat]

For formation depth, Arrian's Array Against the Alans is a good early imperial source and says eight deep. We don't know how usual this was. It might be connected to marching column width or contubernium size (both of which seem to vary between 6 in the republic and 8 under the empire).

Wheeler has argued that the 'falangial' 9as he refers to it) formation has always been in the roman options on the battlefield, and that it was used from time to time over the manipular organisation. Arrian describes sucha 'linear' formation, and this seems to become more common during the 3rd and 4th c. From Maurikios (later 6th c.) we get confirmation that Arrian's 8 deep (8-16) had become standard by the 4th or 5th c.
Maurikios also shows that a line could easily be extended, but by halving from 16 to 8 deep rather than 8 to 6.

Alternatively, if we imagine that Agricola had his c.16 cohorts of auxiliaries in two lines, each drawn up approx. 60 men wide by 8 deep, with eight cohorts in the first line and eight in reserve (or ten in front and six behind, etc), he could order four reserve cohorts to deploy on each flank to double the length of the front line.

Which is another solution.

[Nathan Ross]

it was used from time to time over the manipular organisation.

Do we actually have any evidence that the 'manipular formation' was still in use under the empire?

I was thinking recently (in relation to this thread) about Josephus and his description of the Roman army. It would seem to be a good eyewitness source, and his claim that the Romans marched 6 abreast has caused Wheeler and others to assume that Republican formations were still in use c.AD70.

But could Josephus have taken his details from a republican-era source? He saw the army himself, but it's doubtful he was taking notes at the time... Back in Rome writing the Jewish War, he might have looked up some sources to back up his memories, and reproduced details of an earlier army... :unsure:

[antiochus]

Nathan wrote:
For formation depth, Arrian's Array Against the Alans is a good early imperial source and says eight deep.

How do RAT members envisage how a legion is deployed? There are three troop types we know of under the empire. Maybe Arrian is only describing two in his legion deployment. Caesar mentions three battle lines so why is there a dilemma understanding how a legion is deployed? So is a 5000 man legion deployed 8 men deep suppose to have a frontage of 625 men? How are eight ranks deep suppose to be divided by three troop types? I have yet to see such questions being asked.

Nathan wrote:
Do we actually have any evidence that the 'manipular formation' was still in use under the empire?

Yes there is. But unfortunately most dismiss the evidence for one reason or another. Self defeating is how I would describe it. And that is the main problem, too many have a preconceived concept of what they believe a legion to be and therefore reject anything contrary to their view. They are happy to be wrong all their lives than be wrong for half their life.

Nathan wrote:
It would seem to be a good eyewitness source, and his claim that the Romans marched 6 abreast has caused Wheeler and others to assume that Republican formations were still in use c.AD70.

I’m not sure why many believe the Republican formation was 6 men deep. I’m not saying they are wrong. I’ve asked this question before and didn’t receive a reply, but what is the evidence for the legion in the Republic being 6 deep. A comment from Vegetius about a formation relating to 10,000 men being deployed one rank deep or six ranks deep and Josephus’ marching six abreast is not sufficient evidence to build a strong case. As a counter argument I could claim that Polybius’ 1200 hastati could be deployed 300 men wide by four men deep, and show these accords with Arrian’s description of the men being four ranks deep. I could then add the 1200 principes and have a combined depth of eight men. So who is providing building a better case now? If the above was correct, then Wheeler’s belief in a legion being a phalanx is making something out of nothing.

[Nathan Ross]

There are three troop types we know of under the empire. Maybe Arrian is only describing two in his legion deployment.

Three (or more) under the republic - but the empire? I don't think the few vague references to triarii etc in later imperial inscriptions allow us to be so certain in our knowledge...

Arrian says the first four ranks of his formation are armed with a thrusting weapon (spear, or heavy pilum?) and the rear four with a missile weapon (javelin, light pilum, or throwing spear?). Later the fourth rank possibly appears to swap to the missile weapon. So he could be suggesting that the armament of the legionaries is divided in two, or that all the soldiers carried both a thrusting weapon and one or more missiles, and used one or the other depending on their position in the formation.

So is a 5000 man legion deployed 8 men deep suppose to have a frontage of 625 men? How are eight ranks deep suppose to be divided by three troop types?

I don't imagine the legion was exactly 5000 men - but that would be the nearest round figure. Adding a larger cavalry component might bring it up to c.6000.

Yes there is. But unfortunately most dismiss the evidence for one reason or another.

As I've said before, I don't think anybody 'dismisses' evidence. Very little of the evidence we possess is either clear or comprehensive. Often it is contradictory. Therefore any attempt at analysis involves interpretation, and such interpretation tends to give primacy to certain sources over others. If everyone - or even most historians - agreed on any one interpretation we would not keep having these debates!

what is the evidence for the legion in the Republic being 6 deep.

The argument for the 6-deep republican formation is set out in Wheeler's The Legion as Phalanx... Part II, pages 160-164. Once again, it is not conclusive evidence - it is an argument (or a summary of various arguments) based on the interpretation of what little evidence we do possess.

Wheeler’s belief in a legion being a phalanx is making something out of nothing.

Quite possibly. His 'phalangical' legion does not appear to be a true phalanx, but rather a close-order formation of rather limited depth. I think his main point is that legion formation was more varied throughout its history that is often assumed.

[Bryan]

Since they are being brought up, does anyone have online PDF copies of Wheeler's papers, or at least a summary of what was discussed?

[Nathan Ross]

Since they are being brought up, does anyone have online PDF copies of Wheeler's papers, or at least a summary of what was discussed?

See this post...

:-)

[MonsGraupius]

Having spent much of the last week reading Roman accounts of battles, I've not seen one where there was any suggestion that the army had to be a certain formation. Indeed, the opposite is true, in each and every account the army is adapting constantly to the landscape and to the enemy.

I didn't read a single account saying "the army couldn't get through this or that gap or go along this or that road or fit this or that battlefield because the landscape didn't fit the formation of the army.

Take e.g. the simple expedient of walking down a road. As you will know Roman roads are different widths and when they started using native roads, there would be no defined width. So, the column would have to adapt its width to suit the road width. If you have ever walked in a group - this occurs very naturally.

So, changing formation would be as natural to an army moving through a landscape as changing gear in a car.

So, on some roads they would be six abreast, on others 5, 4, 3. At times they would be forced to pass carts or overtake others or bypass a fallen tree and would smoothly and easily change from 6 to e.g. 3 and then back again. And it would be the same on the line width - it would naturally and easily change - perhaps several times during a battle.

However, just as we easily change gear, it doesn't mean we wouldn't rather be in one particular gear, just because the Roman army would easily change formation (or just couldn't move anywhere off road - or any real battlefield) that doesn't mean they would not prefer a certain depth to their line.

[antiochus]

Nathan wrote:
Three (or more) under the republic - but the empire? I don't think the few vague references to triarii etc in later imperial inscriptions allow us to be so certain in our knowledge...

It nothing to do with the term triarii. Pilani is another term…hastatus, principes and pilani are centurion terms, therefore there is still three troop types.

Nathan wrote:
Arrian says the first four ranks of his formation are armed with a thrusting weapon (spear, or heavy pilum?) and the rear four with a missile weapon (javelin, light pilum, or throwing spear?). Later the fourth rank possibly appears to swap to the missile weapon. So he could be suggesting that the armament of the legionaries is divided in two, or that all the soldiers carried both a thrusting weapon and one or more missiles, and used one or the other depending on their position in the formation.

The hastati carried different weapons to the principes. I can’t see any difficulty in the Romans maintaining this doctrine during or after Arrian’s lifetime and the primary sources support such a doctrine.

Nathan wrote:
I don't imagine the legion was exactly 5000 men - but that would be the nearest round figure. Adding a larger cavalry component might bring it up to c.6000.

I following the primary sources here by using their rounding methods.

Nathan wrote:
As I've said before, I don't think anybody 'dismisses' evidence.

Actually many do. I have been informed many a time that Tacitus or Ammianus use of the term maniple must not be taken seriously. This is dismissing evidence. I have also been told on this list that the cohort did not exist before 210 BC. Prior to that the theory was 102 BC. All reference to cohorts before 210 BC have been dismissed and the reasons given are endless. Therefore, people do dismiss evidence and lots of it.

Nathan wrote:
Very little of the evidence we possess is either clear or comprehensive. Often it is contradictory. Therefore any attempt at analysis involves interpretation, and such interpretation tends to give primacy to certain sources over others. If everyone - or even most historians - agreed on any one interpretation we would not keep having these debates.

Can you show me one serious analysis of the empirical data that dissects every legion number, cohort sized or army number found in the primary sources? From everything I have read including Brunt’s study, there has not been a comprehensive study undertaken. Most study skim over such detail. So in a nutshell, the reason why we continue to have these debates is because no one has undertaken a serious analysis of the data. And the rest of the problem centres on the manner in how the information has been interpreted and the favouring of one ancient historian over another. These are all major shortcomings.

Nathan wrote:
The argument for the 6-deep republican formation is set out in Wheeler's The Legion as Phalanx... Part II, pages 160-164. Once again, it is not conclusive evidence - it is an argument (or a summary of various arguments) based on the interpretation of what little evidence we do possess.

It’s not just Wheeler, some on this forum are of the six deep school. It also has popped up in a few books I have read. As I have already stated, two vague pieces of information does not make a case.

Nathan wrote:
I think his main point is that legion formation was more varied throughout its history that is often assumed.

Mike wrote:
I didn't read a single account saying "the army couldn't get through this or that gap or go along this or that road or fit this or that battlefield because the landscape didn't fit the formation of the army.

Well the reason why is because the Romans knew the terrain would not adapt to them so they adapted to the terrain. That’s why the primary sources mention the Roman army at times deploying four cohorts wide….they were fighting in a restricted space.

Mike wrote:
Take e.g. the simple expedient of walking down a road. As you will know Roman roads are different widths and when they started using native roads, there would be no defined width. So, the column would have to adapt its width to suit the road width. If you have ever walked in a group - this occurs very naturally.

A 40,000 man army is entirely different to walking in a group. And how is it when people are escaping from a football ground they method of exiting does not occur naturally?

Mike wrote:
So, changing formation would be as natural to an army moving through a landscape as changing gear in a car. So, on some roads they would be six abreast, on others 5, 4, 3. At times they would be forced to pass carts or overtake others or bypass a fallen tree and would smoothly and easily change from 6 to e.g. 3 and then back again. And it would be the same on the line width - it would naturally and easily change - perhaps several times during a battle.

Have you thought this through? The width of a road column would be governed by the narrowest defile that the line of march anticipated encountering (bridge, village, etc.). Making the width of the column wider than any defiles that could be encountered would be senseless, and only slow the column to a halt in order to narrow its frontage in order to traverse the defile. Now with your army changing formation and every unit doing this, you have a traffic jam and unnecessary long delays. So do tell me how is it as easy as changing gears in a car?

Most Napoleonic formations marched four abreast and in Spain because of the poor roads they were reduced to two or three abreast. So how wide do you envisage the roads were in Britain when the Romans invaded? Did the Brits build them six abreast or even four abreast? I would say that when the Romans conquered Italy they found the roads no wider than that of a single horse and cart.