The mechanism of Roman line relief

[Erpingham]

Alas, my knowledge of the fighting techniques of the Imperial legion is thin (lack of similar level of exposure to debate than on the Republicans ) but it seems to me that the appearance of auxilia in self contained small units and the fact that most of these units could form line of battle will have complicated the formation model.  Take, for example, Mons Graupius where the auxilia are deployed in front of the legions - could line relief work in those circumstances? Or would you get something more on the lines Ammianus aludes to of individual units being relieved/reinforced?

Going briefly back to Ammianus, Rance (whose view of Ammianus' literary flourishes is probably influentual in my own view looking back) does not say definitively that the Romans adopt a foulkon (or fulcum, as the original Latin would have it) because their is no explicit evidence of Late Western Roman use, though the term must have been introduced between the late 3rd and early 5th century.  He is clear that the formation described is a lineal descendent of Roman formations, in its anti-cavalry form dating back to the late Republic.  For those who haven't read it for a while, the paper is well worth another look.

[Patrick Waterson]

Alas, my knowledge of the fighting techniques of the Imperial legion is thin (lack of similar level of exposure to debate than on the Republicans ) but it seems to me that the appearance of auxilia in self contained small units and the fact that most of these units could form line of battle will have complicated the formation model.  Take, for example, Mons Graupius where the auxilia are deployed in front of the legions - could line relief work in those circumstances? Or would you get something more on the lines Ammianus aludes to of individual units being relieved/reinforced?

Join the club - knowledge of the precise functioning of early, middle and late Imperial legions is sketchy to say the least, largely because our sources for the period tend not to have a Polybian descriptive approach or even a Livian assemblage of clues.  However I believe it would be fair to say that you make a good point in that committing auxilia to the main line of battle requires a different approach to relief than the usual three-line legion system.  Mons Graupius (AD 83 or 84) was an engagement where, as Mick Hession has pointed out, once the British (Caledonian) chariotry was out of the way the infantry of the Caledones were really no problem for their trained Roman opponents, even though the latter were 'merely' auxilia.  Hence one suspects the question of line relief did not arise.

The German invasions of the 2nd century AD that gave Marcus Aurelius such a hard time may have been a different matter, the Marcomannii in particular being a tough nut to crack.  When the crisis of the 3rd century AD broke, the Empire seems to have begun increasing its cavalry establishment, perhaps a sign that the infantry could no longer entirely cope.  By the 4th century, heavy - even armoured - cavalry had become a major component of Roman armies, and in the 5th century it seems to have become the mainstay of the Empire's remaining forces, although a larger infantry component still existed.  In the 6th century, especially during Belisarius' campaigns of reconquest, infantry is definitely the subordinate arm, at some times almost superfluous in battle, though still essential in sieges and useful in engagements such as Taginae (where Narses dismounts his cavalry and supports it with infantry, especially archers).  We may be able to surmise from this that the Romans were finding their infantry system had weaknesses, and these weaknesses may have had to do with the inability to provide effective relief for auxilia, or it may simply be that their opponents were now using more cavalry (the increasing Sarmatian presence being a case in point) and the Romans were having to raise more and heavier cavalry to counter them.  Somewhere along the line, however, there would have been a need to relieve auxilia in mid-fight, German tribes tending to have more fighting ability and staying-power than Caledonians.  Hence one could posit that with increasing proportions of auxilia in the battle line a reduction of the legions to single-line formations would have allowed a more or less uniform unit relief system instead of having to operate two different systems side-by-side.

Whatever the roots of the change, by the mid-4th century AD the Romans seem to have had a system that worked really well against Germans when not let down by bad leadership, and could even work effectively against Sassanid/Sasanian Persians (Parthians and Persians were a problem for the earlier system), though this took good leadership to achieve.

'Foulkon' seems to derive from the Greek 'phylax' or 'phulax' (depending on how one transcribes the upsilon), meaning a guard.  'Phylakein' or 'phulakein' were guard forces of any sort, e.g. to cover foragers, so the 'foulkon' seems by extension of meaning to have been a formation or configuration whereby infantry could, so to speak, guard themselves.  Ammianus refers to Roman infantry at Argentoratum being protected by a 'nexam scutorum compagem', a close-joined array of shields, 'in modum testudinis', i.e. in the fashion of a tortoise.  He seems to be describing either a testudo or, more likely, a shield wall arrangement.  Whether this equates to a Byzantine or even Rancian 'foulkon' I leave others to judge.  Functionally I see no reason why the concept should not fit: doctrinally and from a detail standpoint it may be a more open question.

Patrick

[aligern]

I would place the dominance of cavalry in Roman armies as a change that occurs in the 580s when fighting the Avars and not earlier. Even then a large Roman army operating against the Persians or Arabs  might well have a majority of infantry.  The reason for the use of all cavalry armies in the Strategikon  of Maurice is really because the infantry cannot keep up and pin the Avars. At one point in Maurice's campaigns a general is referred to as having ' the infantry army'.
Belisarian armies are cavalry armies by default and are not typical. The army at Dara, at Callinicum, at Taginae, atvCasilinum, is an army that is largely infantry, but with a substantial cavalry component. I would go so far as to say that the cavalry has danced to being a decisive arm, but still uses an infantry base. At Taginae and at Casilinum infantry and dismounted cavalry are used to blunt the opposing attack and then cavalry surround and pursue. If anything that is a sophisticated all arms strategy rather than a cavalry dominant one.

Roy

[Erpingham]

'Foulkon' seems to derive from the Greek 'phylax' or 'phulax' (depending on how one transcribes the upsilon), meaning a guard.  'Phylakein' or 'phulakein' were guard forces of any sort, e.g. to cover foragers, so the 'foulkon' seems by extension of meaning to have been a formation or configuration whereby infantry could, so to speak, guard themselves.  Ammianus refers to Roman infantry at Argentoratum being protected by a 'nexam scutorum compagem', a close-joined array of shields, 'in modum testudinis', i.e. in the fashion of a tortoise.  He seems to be describing either a testudo or, more likely, a shield wall arrangement.  Whether this equates to a Byzantine or even Rancian 'foulkon' I leave others to judge.  Functionally I see no reason why the concept should not fit: doctrinally and from a detail standpoint it may be a more open question.

Patrick

Rance has a couple of different derivations of foulkon .For detailed discussion see the paper, but both would involve the term coming into Greek from Latin.  As I said, a whole other topic of debate More interestingly, there is the question of whether in modum testudinis means not "in the fashion of a tortoise" but "in the fashion of (a military formation called) a tortoise".  Rance would go for the latter, and would view this , as you suggest, as a type of interlocked shieldwall i.e. a formation in which soldiers held their shields interlocked as they would in a testudo.  Which is also what troops in a foulkon do too, so this maybe the same or an ancestral version.

[Patrick Waterson]

Well stated, Roy.

I would agree the 580s as the time when Byzantine infantry starts to become an optional extra on the battlefield rather than an integral part of the army.  From the 3rd century AD cavalry was increasing in status and importance, and my understanding of the beginning of cavalry dominance would date to such actions as the Milvian Bridge, where the cavalry engagement seems to have been expected to decide the battle - and did.  We can agree on cavalry becoming the arm of decision somewhere along the line.

Anthony, Ammianus' usage ('in modo testudine') could be read either way (tortoise or testudo), though a full testudo involves shields all round and on top, and is really an assault formation for sieges rather than a battlefield formation (albeit Crassus' men appear to have used it at Carrhae once they realised the Parthians were not going to run out of arrows).  This may be why Ammianus has them 'in modo testudine' (in the fashion of an overlapping scaly thing) rather than simply 'in testudo' (or rather 'in testudine').  The governing criterion for their formation is that they need to be able to use their personal weapons.  There is of course a certain attraction in the idea that rear ranks are holding up shields as a 'roof', given that friendly (and possibly enemy) missiles are coming over for much of the battle.

[andrew881runner]

I have made a video which shows clearly how relief system worked. I studied a lot of sources.
http://m.youtube.com/watch?v=LzTgdz7vmPY

[Mark G]

That is one of the interpretations.

Your Greeks are obligingly passive throughout, especially when the complex formation changes are in progress.

And the absence of velites and pila stands out, as does what appears to be the pvermanned triatii
Fun to watch tjough

[Prufrock]

Nice illustration of your ideas, Andrew.  I thought the push and shove seemed quite realistic; it didn't have the massive casualties of RTW: 1.

But as Mark G says, the enemy don't advance and the legions are able to move how they wish, so I'm not convinced.  Good to watch though.  Thanks for sharing!

[Justin Swanton]

I have made a video which shows clearly how relief system worked. I studied a lot of sources.
http://m.youtube.com/watch?v=LzTgdz7vmPY

I love RTW2 graphics. It would be great to be able to tweak in detail the behaviour of troops so as to give a precise demo of the mechanism/theory one is trying to illustrate. Do you know if there is any way of doing that? A mod or add-on or something?

I enjoyed the video. Mille grazie!

[andrew881runner]

there are actually hundreds mods for rome 2. I made this video when there was no mod yet. I could not use velites cause I had limited number of units. For the Pila throwing, there are actually very good mods which add it. At that time there was Noone. If you go into my channel you will see many more videos with examples of how cohortal legion and manipular legion could act against different enemies, like pike phalanx or barbarians.
The idea or relief system I show in the video is based on the assumption that during a battle there were a lot of minor pauses, because soldiers could not sustain the physical and emotional stress of fighting for  long. This is is most accepted theory of relief system I have found and the most practical. We know in fact that each manipulum had a front centuria and a back centuria. Some author speaks about a legion with no gaps, some talks of gaps among maniples of the dimension of a maniple. So this tactic I have shown would melt together all these different things.
Sorry for my English.

[andrew881runner]

I have made a video which shows clearly how relief system worked. I studied a lot of sources.
http://m.youtube.com/watch?v=LzTgdz7vmPY

I love RTW2 graphics. It would be great to be able to tweak in detail the behaviour of troops so as to give a precise demo of the mechanism/theory one is trying to illustrate. Do you know if there is any way of doing that? A mod or add-on or something?

I enjoyed the video. Mille gratia!

as I said before there are many excellent mods on the steam workshop, but one of the best is Divide ET Impera, which changes the skin of units and their behaviour in a realistic direction. So if you will buy the game try that mod, the game alone is a bit unfinished and lacks some epic feeling, but is surely a very good game.

[Patrick Waterson]

The idea or relief system I show in the video is based on the assumption that during a battle there were a lot of minor pauses, because soldiers could not sustain the physical and emotional stress of fighting for  long. This is is most accepted theory of relief system I have found and the most practical. We know in fact that each manipulum had a front centuria and a back centuria. Some author speaks about a legion with no gaps, some talks of gaps among maniples of the dimension of a maniple. So this tactic I have shown would melt together all these different things.

Again, the ability of troops to sustain fighting for long periods seems to have depended upon the troop type, their equipment and their training.  Italians generally and Romans in particular seem to have used heavy shields (held close to the body and perhaps supported on the shoulder like the Greek hoplite shield), light weaponry and a lot of training, giving them considerable endurance.  Gauls seems to have relied heavily on initial impact and to have been poor at sustained fighting; Germans seem to have been better than Gauls at sustained action but still inferior to Romans, especially in an Italian climate.

There is also the consideration that if an opponent is tired enough, or lax enough, to stand by and not interfere while one carries out these relief manoeuvres, then he is probably sufficiently tired or lax that a good push would break him, making relief unnecessary.

The converse also applies: that is an excellent video, but if I were commanding the hoplite opponents the hastati would be massacred as they tried to withdraw, the more so as the first part of their withdrawal is spent exposing their flanks and getting in the way of the engaged hastati who are still trying to retreat.  The same goes for the principes.  And why do the hoplites put themselves into complete confusion before attacking the triarii?

I applaud your exploration of the Roman relief system - this has been extensively discussed here among Society of Ancients members and it should be a good subject on which we can compare notes.

[Prufrock]

I could not use velites cause I had limited number of units. For the Pila throwing, there are actually very good mods which add it. At that time there was Noone.

My apologies, I didn't realise the constraints you were working under.  Given this, it's an even bigger 'well done'.  I've posted a link to it on the SoA facebook page so that others can see too.

Thanks again,
Aaron

[andrew881runner]

what you say of endurance can be true but does not take into account there pauses in fighting existed and that was sure. I have read a lot and if you want I can give you some links, in Italian unfortunately of very good websites who list all sources.
As for the fact that enemies would have used the relief moment to attack, they would not have had time, since the timing I show in the video must not be taken as accurate, they probably did it in less tkme. Secondly even if they attacked in the little time of rank change they would have found the principes line running towards them. And the full line of principes would have deployed in seconds. So they would have had something like an estimated 20 seconds (not my idea) to attack.
Considering that they were probably resting some meters far from line of battle, they would have had excellent reflexes to understand the complex manoeuvre which was happening and attack in that very moment. Unlikely to happen. Even if they did it, they would have had no real effect, since they would have simply slowed the retreat of hastati line and the advance of principes.
I really think that this is the relief system more likely to happen.

[Patrick Waterson]

I used to think the relief system worked this way until I began thinking about how and why the Romans would want to relieve lines.  There is no point relieving a line which is doing well: relief is required only when the enemy is starting to gain superiority over the hastati, and then it is necessary to have a relief system which works well under pressure.

If the relief system depends upon a break in the fighting, then a) any relief system (not just the Roman one) will work during a break so anyone can do it, and b) if there is no break in the fighting the system cannot work.

Considering that they were probably resting some meters far from line of battle, they would have had excellent reflexes to understand the complex manoeuvre which was happening and attack in that very moment. Unlikely to happen. Even if they did it, they would have had no real effect, since they would have simply slowed the retreat of hastati line and the advance of principes.

The argument that the enemy would be baffled by what was happening might work for the first battle they ever had against Roman forces, but by the time of the second battle they would have seen the system in action and thought about what to do.  The proposed system of pulling back alternate centuries would create gaps in a tired line and allow the opponent to surge round the flanks of the prior centuries.  The principes line would have to deploy 'in seconds' and rush in, but that would result in pockets of tired hastati being forced against the enemy by the intact principes line, so the hastati would be sacrificed rather than relieved.

It was this problem of trying to relieve a line under pressure that made me take another look at the sources, and this led to a couple of clues.  Livy in his Book VIII description of the battle between the Romans and Latins on 340 BC refers to the hastati as retro cedentes (retiring rearwards or backwards) in order to be relieved by the principes.  Polybius, in his Book II description of Rome's victory against the Insubres, notes that the Roman deployment at the edge of a river left no room for the customary Roman epi poda relief manouevres.  Epi poda means stepping backwards.

This suggests that the intact hastati line withdrew backwards through the line of principes.  Each file (except one, on a flank) would pass between two files of principes, who would pick up the combat as it reached them.  One intact line handed over the fighting to another.  At no point were there gaps, nor was the formation disrupted in any way.  It was unspectacular, but it could be done even when the fighting line was under strong enemy pressure.

I have also concluded that the 'quincunx' formation was never used in battle: the only apparent example of quincunx deployment in our sources is at Zama, and even there Scipio places his subunits one behind the other and fills the gaps with velites.  In the battle against the Insubres (223 BC) the Romans have a continuous front line, as is evidenced by equipping the leading ranks of hastati with triarii spears.  This would seem to be the norm and not the exception.