How continuous was combat?

[Erpingham]

From the quotes earlier mentioned, particularly about men who had previously fallen (from fatigue, blood loss or both) getting up to rejoin the fighting and the fact that mounds of bodies managed to pile up during Caesar's fight against the Nervii.  Neither seems practicable if casualties were being withdrawn/removed or if men were falling back to take breaks and re-engaging on a new line. 

I think I'd have two caveats here.  One is the "mounds of dead" description is also applied in medieval battles (most famously Agincourt) where we know that the units at the front were constrained in some way from withdrawing or operating flexibly (crushed together, unable to use weapons etc.).  The statement about all hope being gone suggests that may have been the case here and so the situation may not be normal.

Secondly, there is a difference between dead casualties and wounded casualties.  Just because bodies are piling up doesn't mean ambulatory wounded men are not exchanging places with unwounded comrades.  It seems, from evidence shared so far, wounded men stayed with their units.  We don't have the evidence that they stayed in the front rank. 

[RichT]

Agreed there's not good evidence (aside from the scattered cases eg Appian) for routine replacement of wounded, or tired, or generally fed up, men within a formation. On the other hand, while walking wounded might have kept fighting if they were able, if someone fell, fainted, or was otherwise hors de combat, it would be ludicrous to think that the man in the rank behind would just stand and do nothing,  rather than stepping forward and taking up the fight. I expect that in this as in all other periods, if a badly wounded man was able to get out of the way, or be dragged out of the way, he would - but how possible this would be depends on vexed questions like the intervals between files, the amount of pushing and shoving going on, etc, and no doubt varied widely from case to case.

This proves nothing about the continuity (or duration) of combat.

There is this too:

Caesar BG 3.4 "A short time only having elapsed, so that time was scarcely given for arranging and executing those things which they had determined on, the enemy, upon the signal being given, rushed down [upon our men] from all parts, and discharged stones and darts, upon our rampart. Our men at first, while their strength was fresh, resisted bravely, nor did they cast any weapon ineffectually from their higher station. As soon as any part of the camp, being destitute of defenders, seemed to be hard pressed, thither they ran, and brought assistance. But they were over-matched in this, that the enemy when wearied by the long continuance of the battle, went out of the action, and others with fresh strength came in their place; none of which things could be done by our men, owing to the smallness of their number; and not only was permission not given to the wearied [Roman] to retire from the fight, but not even to the wounded [was liberty granted] to quit the post where he had been stationed, and recover."

But this is defence of a fortified camp, which might have had different rules from open battle (not least, its duration).

[aligern]

I suppose hat the difference between a siege and a battle in the open field is that between the opposing sides there is a wall or palisade that forms a boundary that cannot easily be crossed.mHence a proponentnof thebthoery that men stayed n pkace until death or flight relieved them might say that it becomes possible to manoeuvre in front of a fortification because the opponent cannot react aggressively. Believing as Zi do that there is more depth and complexity on an Ancient battlefield than some allow for, I would see the relief of tired troops indicate in Richard's pist as a drill normal to the battlefield. The line wall only forms a boundary and protection what goes on behind it is the normal rotation of units that would occur in the open field.
Roy

[Patrick Waterson]

I think I'd have two caveats here.  One is the "mounds of dead" description is also applied in medieval battles (most famously Agincourt) where we know that the units at the front were constrained in some way from withdrawing or operating flexibly (crushed together, unable to use weapons etc.).  The statement about all hope being gone suggests that may have been the case here and so the situation may not be normal.

Although the piles of bodies would have to be started long before 'all hope was gone' in order to reach the height they did.  Caesar's praise for the courage of the Nervii (Gallic War II.27) would ring hollow if their valour and tenacity had been solely the result of being poorly positioned - the implication is that there was a voluntary element.  That the Nervii were using their weapons freely right up to the last is evidenced by their  famous last stand atop the mounds of bodies.

Secondly, there is a difference between dead casualties and wounded casualties.  Just because bodies are piling up doesn't mean ambulatory wounded men are not exchanging places with unwounded comrades.  It seems, from evidence shared so far, wounded men stayed with their units.  We don't have the evidence that they stayed in the front rank.

Or that they left it.  If they fought until they dropped, the new front ranker would of course step over them to deal with his opponent.  What we do know is that Romans were very averse to letting people leave their places on the battlefield (sons of consuls are on record as having been executed for leaving the ranks to duel and defeat an enemy champion), but they also had a Civic Crown awarded for saving a fellow citizen's life (the saved man recommended the award, which apparently made it quite rare), so one can hypothesise around that.  One would logically expect that a man with a disabling would would not be expected to stay in the front rank, but until we find a pertinent source statement this remains conjectural.

This proves nothing about the continuity (or duration) of combat.

Which 'this' is my learned friend referring to?

But this is defence of a fortified camp, which might have had different rules from open battle (not least, its duration).

One can surmise that in the defence of a fortified camp a reserve would ordinarily be kept and wearied/wounded defenders replaced with fresh ones if at all possible when under sustained assault.  Manning a rampart, a static obstacle with easy communications behind it, is indeed a different matter to holding steady a line of battle in the open.

The line wall only forms a boundary and protection what goes on behind it is the normal rotation of units that would occur in the open field.

Perhaps Sir should start producing some examples of this purported 'rotation ... in the open field'.

[RichT]

As Anthony suggested earlier, this is an argument over the tiniest details of the internal anatomy of the creature in front of us, when we can't agree whether that creature is an elephant or a giraffe.

Reading something like Caesar's account of the Sambre (which is a poor example since it's not a formal pitched battle, but here we are) it is obvious to me that what is described is a fast moving, mobile running battle, with a mix of sword play, hurling missiles, attacks, retreats, engagements and disengagements,  small units acting independently, redeployments and all. It may be equally obvious to Patrick that it involved continuous unbroken battle lines standing in fixed positions hacking at each other for hours on end. It's pretty pointless quibbling over the details when there is such a fundamental divergence on the big picture.

To go from the sublime to the ridiculous, it's quite hard to form a mental picture or model of ancient battle, and this is somewhere where wargames should be able to help; ignoring all the minutiae of rules and factors and whatnot, they could give a feel for the dynamics of battle (this is Phil Sabin's hope). I haven't found toy soldier games help at all to be honest, but the other night I pulled out Sid Meier's Gettysburg again, and it struck me how good a feel for a how battle (might have) worked this gives - I don't know if it's accurate or not (not my period), but it feels entirely plausible, and the way units move, fight, cluster,  disperse, fall back, rush forward made me think of the fighting at the Sambre. Off topic, but I think there is scope for this sort of computer re-creation to provide insights that can't be achieved by quibbling over words or banging heads about fundamentally divergent mental models.

[Mark G]

That sounds like a cue for Justin to re enter the fray.

[aligern]

Unfortunately wargames generally include mst of the more minor tactical aspects within the £granularity ' of the two battle lines. I recently plated a game that used the full three rank plus velites formation of a Roman legion. Normally all of this  detail is subsumed in the one or two ranks of Roman legionaries. Certainly in mst rules the Romans do not get down to the Triarii because these ceterans are eithe subsumed or put on the flanks or some other unique task. Even in the game I was playing the hastati, principes and triarii of each legion operated as seperate units,rather than a whole line of the army falling back to reveal a new line. Perhaps it is because working out all of the interactions of minor units, such as the fate of cohorts takes so much time and the rule writer wants to concentrate on grand tactical manoeuvre, breaking the cavalry on one wing and then outflanking the enemy, fir example.
Like Richard  rather despair of wargames telling us much about the minir tactical details. This is largely because the rule writer has to decide how combat works in order to replicate it with rules. Thus we are not likely to learn anything beyond what the author already imagined.
As an exampke. let us take the pilum. This could be
1) A deadly volley that kills many and cannot be defended against because when it pirces a shield its design allows it to penetrate through to the holders body.
2) A device for de shielding opponents prior to their being engaged with the sword that means that The thrower has the advantage in combat.
3) Merely a heavy javelin, The Romans have two each and one or both sides can spend a considerable time throwing them with not too devastating effect, before closing with swords.
All three situations have evidence to support them and rules can be written for all three, once that has been done, playing the game will not tell us which is the most likely way that the pilum was used, though some games will be quicker, some more devastating.

[Patrick Waterson]

Reading something like Caesar's account of the Sambre (which is a poor example since it's not a formal pitched battle, but here we are) it is obvious to me that what is described is a fast moving, mobile running battle, with a mix of sword play, hurling missiles, attacks, retreats, engagements and disengagements,  small units acting independently, redeployments and all. It may be equally obvious to Patrick that it involved continuous unbroken battle lines standing in fixed positions hacking at each other for hours on end. It's pretty pointless quibbling over the details when there is such a fundamental divergence on the big picture.

In addition, the Sambre was effectively three battles in one.

1) On the Roman left the Atrebates, who had pushed themselves too hard sprinting across the river and up the hill, were easily routed by the Romans and fled back down the hill with the Romans cutting them down.  Those who made it across the river rallied on the nearby steep slope while the Romans were crossing, but their heart was not in it and they were soon routed again, this time for good.

2) In the Roman centre (sorry this is all Romanocentric but so are our sources) the Viromandui advanced with more restraint and resisted with more effect, being driven into but not across the river - Caesar notes that the Romans were fighting on the riverbank itself, the Gauls perforce being in the water.  He does not return to this part of the field in his narrative.

3) The Roman right was pitted against the Nervii, who fought to the last and impressed even Caesar by their athleticism and stamina as well as their courage:

"... it ought not to be concluded, that men of such great courage had injudiciously dared to pass a very broad river, ascend very high banks, and come up to a very disadvantageous place; since their greatness of spirit had rendered these actions easy, although in themselves very difficult." - Gallic War II.27

Another translation:

"Only heroes could have made light of crossing a wide river, clambering up the steep banks and launching themselves on such a difficult position."

So here we have three limbs of our pachyderm (or cameleopard): one which gives way rapidly, one which is lost to sight half-way through, and one which stands for the whole battle, albeit at the cost of amputation.  This is what makes it impossible to establish a general rule which fits every battle, because not only is every battle different but parts of the same battle can show very different characteristics.  What we do have are accounts of the Roman military system and what it trains for: we can see many manifestations of this, both in the observance and the breach, on campaign and on the battlefield.  I suggest we study the known classical descriptions of the system to find the intended norm.

And as Roy notes for the pilum, we have different engagements in which the weapon's lethality ranges from the marginal to the considerable.  All are correct: what matters as much as the weapon is its use, individually and collectively.

One thing I would not recommend is trying to superimpose engagement patterns from completely different centuries, cultures and military systems.  The American Civil War did not even properly match its own century - look at the contrasting battlefield approaches in the Crimean, Franco-Austrian (1859), Austro-Prusso-Danish (1864), Austro-Prussian (1866) and Franco-Prussian Wars - let alone provide any useful comparisons for the classical period.

[aligern]

Just to clarify that I think that pila generally do not cause large casualties, inded missile weapons do not cause massive casualties provided the opponents have sheilds and/or armour.  This is evidenced in that the victors in Roman versus Roman battles do not suffer large casualties, nor are the opponents in Roman versus say Carthaginian or Macedonian battles become destroyed in the missile phase. In those examples where ( in my belief untypically) Romans throw pila all day casuaties are obviously not huge.  In theory, if there are 1,200 velites  in  a legion with 7 javelins each, so that is 7000 javelins and a further 2,400 or 4,800 pila and yet We do not hear of opponents being destroyed by missiles.
German armies , following Tacitus, throw huge quantities of missiles (though they might not be the highest quality weapons) yet the Germans do not do huge damage to the Romans, or even other Germans.  The Parthians  spend all day with unlimited supplies of arrows shooting at legionaries, but the target units are not killed, they lose morale. Caesar's centurion has 100 arrows in his shield .
Generally, we might say that  neither missiles nor arme blanche weapons create great casualties until the opponent either runs away or is too crushed together to fight back effectively,the Cannae or Adrianople effect. We are not well informed as to what causes armies to break, though we have reams of speculation. That feeds through into our wargames as they generally elide from assuned cause to known effct. So unit A shoots at unit B , it causes casualties and as those mount unit B finally breaks. Unit X fights unit Y hand to hand and causes casualties which in the end break Y.  As a moderating device we throw dice which lessen or increase the killing effect, but it is kilks that determine the outcome. The units often have 'factors' such as armour or weapon type or skill level which the author makes more or less arbitrary judgements about, but which all effectively or lessen kills. Yet one thing we can be reasonably certain about is that units are not broken by kills, or rather not broken by kills alone.
Roy
Following Richard's point about how useful our modelling of battles goes and

[Erpingham]

Unit X fights unit Y hand to hand and causes casualties which in the end break Y.  As a moderating device we throw dice which lessen or increase the killing effect, but it is kilks that determine the outcome. The units often have 'factors' such as armour or weapon type or skill level which the author makes more or less arbitrary judgements about, but which all effectively or lessen kills. Yet one thing we can be reasonably certain about is that units are not broken by kills, or rather not broken by kills alone.

At the risk of drifting off topic, rules have been using casualties as a proxy for unit degradation since Kriegspiel.  The danger is really being too literal.  I think the charts in Early Modern rules like the WRG series gave a spurious feeling of exactitude, giving numbers of dead which had to be totalled into 20s to remove figures.  More modern rules, I think, do move more towards outcomes, using steps to represent not dead men but general reduction in will to combat, wrapping killed, wounded, exhausted, demoralised together in casualties.

To go back to the original suggestion of testing different ways to model combat by adjusting rules does rather require us to know enough in detail about combat to know which one comes closest.  As we spend a lot of our time wrangling about what combat was like in the ancient and medieval period, this is going to be hard.  I think it will be particularly hard to model a small group ebb and flow model for our Romans because it will need a much greater operational cohesion of battle lines than many rule sets can muster.

[Prufrock]

the other night I pulled out Sid Meier's Gettysburg again, and it struck me how good a feel for a how battle (might have) worked this gives - I don't know if it's accurate or not (not my period), but it feels entirely plausible, and the way units move, fight, cluster,  disperse, fall back, rush forward made me think of the fighting at the Sambre. Off topic, but I think there is scope for this sort of computer re-creation to provide insights that can't be achieved by quibbling over words or banging heads about fundamentally divergent mental models.

But this is the problem, isn't it. If no one knows, if there is no agreement, no computer simulation can help. All it can do is demonstrate the assumptions that have gone into the model, much like this thread.

The sources show periods of missile exchange, charges, withdrawals, follow ups, last stands, leader speeches, reform, formation changes and various other phases of battle. Sometimes there was intense melee combat; at other times it was not so intense. Sometimes there was no melee combat at all.

The question 'was battle continuous?' Depends entirely on the definition of battle and continuity. Terms have not been defined, so the argument is all over the place, and we have got no further than 'sometimes' or 'it depends' as an answer.

If this turns up as a six page summary in Slingshot I'm cancelling my subscription.

[Patrick Waterson]

The question 'was battle continuous?' Depends entirely on the definition of battle and continuity. Terms have not been defined, so the argument is all over the place, and we have got no further than 'sometimes' or 'it depends' as an answer.

That is not really the central point, although it would have been nice if people had established some goal areas at the ends of the playing-field.  The more alert among us will have noticed that the actual point at issue is whether such combat as existed when two sides met on the battlefield was habitually, that is in the majority of cases and what we could consider the norm, punctuated by frequent and/or regular mutual disengagements and withdrawals for rest breaks or whether it was not.

What we seem to be getting is a lot of source accounts which do not mention any sort of break and a few which do, albeit even these have a hard time fitting the 15-20 minute bracket which I understand is considered by the proponents of discontinuous combat to be the maximum for which a man could possibly fight.  There seems to be at least a partial failure to appreciate that this depends heavily upon the man and the fight.

If this turns up as a six page summary in Slingshot I'm cancelling my subscription.

To paraphrase Frederick the Great, you and I can desert together.

[aligern]

Re the mechanism by which combat is resolved.
I distinctly remember that in WRG rules many years ago Phil Barker expressed that the view that following a bad dice throw something must have happened . on the field such as a general getting killed or a standard falling, that the use of a die to randomise elements of the morale  process included and allowed for such events. That is generally the view of rules writers.
Erpingham's point is, if I read aright, that casualties are not representing actual dead or incapacitated men, but rather the degradation of fighting ability. Thus a unit becomes  nearer dissolution, or at least disruption (by pushbacks and overlaps) and eventually dissolves.
What exercises my mind is whether we should be looking to replicate the actions that Ancient generals took and the standard operating methods that the troops used in order to generate the outcomes.
To an extent fighting methods are taken care of by the factors allocated for  attack and defence. Put a unt of Romans against a unt of Germans and they will go through various processes and the rule writer has to decide on how those are replicated, as I said, mostly by factors being allocated.There I suppose the best we can do is examine reported battles to verify the actual relative lethality of fighting styles. The other possibility is replicating the fighting styles in more detail on the tabletop, but would this create too many interactions and take up too much time.
Another area of decision that operates is those things that a general can arrange in order to advantage his men, such elements as being uphill, fighting over broken ground, or arrangng to clear the ground for your scythed chariots, or using pathways in marsh your chaps are familiar with, being motivated by an exhortation...that one more  step, arranging a flank attack, depriving the opponent of breakfast or lunch, sending in reinforcements.   These vary in effect, but I suspect they were important to generals desperate to get an edge and maybe we do not arrange to replicate them sufficiently Should a player be able to spend on increased weapon supplies? Should a rousing 'one more step' be on the shopping list? Should a a die roll determine whether you can catch the opponent before he has taken his meal?
The remaining area  is to look at what actually does cause units to break at a point where relatively few casualties have been caused, but fighting may well have gone on some time. The ancients were generlly of the opinion that battle was risky and often luck dependent. As said ealier, a standard bearer tripping up could be the event that tips the balance and that cannot be trained or prepared for. However, that is unusual, surely the determinant of which side routs should be something discernible and replicable. Here I would raise something which I recall 7th edition WRG attempted, which was to see the determinant as exhaustion or loss of cohesion and then reward generals who managed to arrange for their enemy to lose cohesion more rapidly than their troops. In such a mechanic there do not need to be a large number of 'kills' as these are only contributory to cohesion and most importantly the result of its loss....free hack here we come. In this model the job of the general (the player) is to arrange for the enemy to suffer the most disorganisation until a limit is reached. The limit itself could be dice moderated so that it was not completely predictable. combat would still contribute to disorganisation and still be variable, contributing a greater or lesser amount to degradation, but it would not be the only or greatest modifier. Troops would tire at different rates during combat...we. could always allow for break off, rest and recovery during the fight.
I do recognise that many sets of rules do try to take into account other factors, but are these factors matters of devising by the player? and are 'kills' too dominant.

Roy

[RichT]

But this is the problem, isn't it. If no one knows, if there is no agreement, no computer simulation can help. All it can do is demonstrate the assumptions that have gone into the model, much like this thread.

Sure - a computer simulation can't tell you what actually happened. What it can do is tell you whether your conceptual model looks and feels plausible when you see it in action, and also whether, when given inputs as close as you can make them to what the sources describe, it produces outputs close enough to be satisfactory to what the sources describe.

I think those are worthwhile goals - otherwise we have conceptual models but no way of testing them other than by arguing over whose conceptual model is more closely based on the sources, which is a sterile, circular argument, as evidenced by this thread. Between Patrick's model of static immobile lines banging away at each other for hours, and the rest of the world's model of something more subtle and complex going on ( ), we have no way of judging which is better, in the absence of the sources clearly ruling out one or the other (which they don't, or this would all have been settled decades ago and we wouldn't be having this discussion). If we could see these models in action, play with them, adjust parameters, it might help. Or not, it's just a suggestion.

It has all been unfocused but I think we know that what we are talking about is the dynamic standoff model of Roman infantry combat, and it might have been better to stick to that more closely.

If this turns up as a six page summary in Slingshot I'm cancelling my subscription.

How about a summary in six parts?

Re: Roy's thoughts - manual toy solider (or cardboard) wargames have too much abstraction (rightly, if they are to be games) to be useful as combat models (IMHO) - though they might (Lost Battles) be useful as battle models. Whether losses are tracked as cohesion, fatigue, or casualties doesn't really make much difference - different words for the same thing (in model terms, not in real life terms - ie gradual unit degradation). Agree very much about player choices - I would love to see wargames where players still had something to do, decisions to take, once units are in combat, rather than just rolling dice, adding factors or checking tables.

[Erpingham]

What we seem to be getting is a lot of source accounts which do not mention any sort of break and a few which do, albeit even these have a hard time fitting the 15-20 minute bracket which I understand is considered by the proponents of discontinuous combat to be the maximum for which a man could possibly fight.  There seems to be at least a partial failure to appreciate that this depends heavily upon the man and the fight.

These are probably a fundamental point of disagreement.   Firstly the two sides take the lack of direct evidence for breaks in different ways.  To the discontinuous cohorts, it is an absence of evidence.  To the continualists, it is evidence of absence.  More significant is the second point.  The discontinualists place the issue of human endurance centrally and assume that human endurance then can be compared to human endurance now.  Patrick (I can't speak for other continualists) holds modern comparisons of athletes endurance as invalid and that the endurance of ancient fighters was much greater.  This point, I think would undermine any attempts at computer simulation because the base assumptions would be so different. 

On Aaron's point of not defining our terms, I think he is right.  What to me is clear evidence of discontinuity, as battles have phases of activity, is not discontinuous to others, because continuity/discontinuity is only about low-level activity once units have contacted.  I'm not sure it would have made a difference to reaching conclusions however.  This is partly because we are taking on the whole ancient and medieval period (albeit usually concentrating on a particular time period at a time) which fails to allow for different practices in different cultures.  But to return to a point made several times by Richard in many different epic discussions like this - we can't reach a definitive conclusion because we don't have the evidence for one.  All we can do is attempt a reasonable theory.  And all these discussions can do is explore the field, throw up ideas and let members form their theories.