Recoil - when did it historically happen?

[aligern]

I would take the logic if your last statement, Patrick, as meaning that the lines had closed up already.
There is plenty of time in those Tacitan battle descriptions for the two lime system to operate and be commented upon. Like you say there is no such comment. Are we then to assume that the deployment is one thick line.

The frontage question is no problem, it simply means that all cohorts deploy to the same frontage and that what varies is depth.

Interesting that  4 up three back deployment allows the three to fill in the three spaces left when the four are deployed.
I don't know that you can argue that there are no gaps because they are not mentioned when my whole contention is that the second line does not perform line relief because the sources do not mention it!
There is a further problem with your reconstruction, Patrick. That is that the second line of your 4:3:3 has only three cohorts. If they are to perform line relief they will suddenly be facing an opponent who has driven back one Roman line with only 60% of the troops that line one had. If they are deployed in some way as standard cohorts the opponent will be round the flanks of the three.
When we see the third line used it is to go to where there is trouble and the Romans are being pushed back. This is not line relief, it is adding more men to a melee .
Again, none of your examples fail to fit with there being two lines for manoeuvre and just the one (lines one and two combined) for the fighting with line three as a reserve that moves to either wing.

Roy

[Patrick Waterson]

I would take the logic if your last statement, Patrick, as meaning that the lines had closed up already.

A thought: if the Italica and Rapax in Tacitus' account had 'closed up already', i.e. before they met enemy forces, this seems to invalidate the idea that the first line charges (with gaps) and then the second line charges into those gaps, so that closing up only occurs on contact (assuming I have understood this correctly).

Thinking about the possible operation of the suggested 'gap system', keeping cohort-sized gaps between cohorts while advancing is going to be tricky unless someone has been out and set up a row of pegs for the troops to orient their movement upon.  Furthermore, if the principle of attacking in quincunx is sound, why not have the first line quincunx its centuries without the second line needing to be involved?  Centuries would be more manageable than cohorts and the smaller intervals easier to maintain.

There is plenty of time in those Tacitan battle descriptions for the two line system to operate and be commented upon. Like you say there is no such comment. Are we then to assume that the deployment is one thick line.

I would rather not assume anything.  Tacitus' reticence on this point is one reason why so much of our understanding (if we can call it that) of the Early Imperial Roman army's method of operation depends upon Vegetius and surmise.  Even Josephus neglects to fill in details of deployment for battle, perhaps because of a dearth of pitched battles in Judaea, perhaps because he did not know, having never seen the Roman army deployed for battle.

The frontage question is no problem, it simply means that all cohorts deploy to the same frontage and that what varies is depth.

Indeed.

Interesting that  4 up three back deployment allows the three to fill in the three spaces left when the four are deployed.

Although Afranius' 5-5-0 deployment would not fit this pattern at all.

I don't know that you can argue that there are no gaps because they are not mentioned when my whole contention is that the second line does not perform line relief because the sources do not mention it!

What is sauce for the goose is sauce for the gander! 

There is a further problem with your reconstruction, Patrick. That is that the second line of your 4:3:3 has only three cohorts. If they are to perform line relief they will suddenly be facing an opponent who has driven back one Roman line with only 60% of the troops that line one had. If they are deployed in some way as standard cohorts the opponent will be round the flanks of the three.

We seemed to be in agreement earlier that the lines were of the same frontage but a four-cohort first line would be deployed in greater depth.  Assuming, of course, that all cohorts were the same size.

When we see the third line used it is to go to where there is trouble and the Romans are being pushed back. This is not line relief, it is adding more men to a melee .

Not at Pharsalus: there it is simply taking over from the first two lines and giving the weary Pompeians the cold steel.  And they do not like it.  Against Ariovistus, committing the third line allowed Caesar to bring the battle on that wing back under control, and although he does not specify whether this was achieved by line relief or by crowding tired troops with fresh ones Adrianople-style, I know which I would do in the circumstances.

Again, none of your examples fail to fit with there being two lines for manoeuvre and just the one (lines one and two combined) for the fighting with line three as a reserve that moves to either wing.

Although to see this system in our sources seems to require a leap of faith beyond the descriptions we are actually given.  Absence of evidence may not be evidence of absence, but a patchwork combination of first and second lines acting as a single formation in a quincunx pattern should leave clues like successive charges, e.g. at Pharsalus, where a charge to contact is described, but the impression Caesar gives is of one coherent line charging, casting missiles and closing to melee together.

[Erpingham]

May I propose this definition of recoil:

One of the engaged lines steadily gives ground whilst maintaining its structure and cohesion. This occurs when the troops of the line are outfought by their adversaries. In this process individual soldiers here and there back away from superior opponents. The soldiers behind them back up to maintain the necessary fighting space between one rank and the next, and the soldiers alongside back up to keep a continuous line facing the enemy and prevent their flanks becoming exposed.

To return to the original topic for a moment, this does seem to be were some of the evidence leads.  But what is its significance?  Justin started the thread with comments on the push backs in DBx.  It seems to me this process would be gradual enough for units either side of the unit falling back to conform and maintain a line, without exposing themselves piecemeal to flank attacks.

[Patrick Waterson]

This is where we tread the border between recoil and rout: at Delium in 424 BC, the Athenian right routed its opponents except for the Thespians, who stood their ground and were promptly surrounded and wiped out (the over-excited Athenians suffering a bit of 'friendly spear' into the bargain).  Much the same happened to the Spartan left at First Mantinea in 418 BC.  Some actions - notably with Roman armies, which seem to have been custom-designed for controllable retirements - show recoil happening slowly and not necessarily progressively, allowing a general with ability and/or reserves the chance to do something about it.  The Battle of Ilipa in 206 BC demonstrates the Spanish wings of a Carthaginian army, which by that date seems to have been using the Roman system, 'recoiling' in similar fashion.  Some combats, however, especially hoplite actions, suggest a brief period of recoil leading to sudden collapse.

I would therefore suggest that the controllability of recoil, or more accurately the duration of non-collapse recoil, is related to the type of troops involved, both in organisation and training and in quality.

[aligern]

To an extent the answer to the question depends, as was said, upon the difference between  a push back and a longer fall back. I suspect that the WRG rules are modelling a push back where the consequence is to give a minor advantage to the winner of each phase. This advantage can be overturned, but if it goes on through the phases of fighting then one side will become the winner.
That looks a fair way of representing a combat where units fight for around 15 minutes and then need a rest and mutually back off. There is then the possibility that the side that is doing least well steps back further and that when combat resumes the winners advance but the losers cannot motivate themselves to advance or to advance as far.
This is not a new suggestion, I think it goes back to Keegan or even beyond and is based upon an understanding of human physiology, that is that hand to hand combat is so intense that it cannot be maintained for longer than around 15 minutes, perhaps 30 minutes. I suppose it depends upon how energetically they go at it. Once this physiological 'truth' is accepted the reconstruction proceeds with a logic of its own. The soldiers must either win within this period or they must stop by mutual consent  and if they stop then separation is logical. I suppose the opponents could slow down to give a more sustainable rate of action or perhaps men behind the front rank could filter through to take the place  of fighters, though descriptions of Gauls and Germans being tightly packed does not suggest that would be easy to accomplish.
We may well be in one of those areas where the process of combat is so well understood by the writer and his audience that it is not something to excite comment unless there is a specific reason for the comment which could be literary style, some unusual occurrence or perhaps in Caesar's account of the pila pinning Helvetian shields together a knowing wink at the traditional effectiveness of Roman weaponry.
If the advance, fight, pause and fall back sequence is perfectly normal then we will not find much specific evidence if it occurring.
Roy

[Patrick Waterson]

We may well be in one of those areas where the process of combat is so well understood by the writer and his audience that it is not something to excite comment unless there is a specific reason for the comment which could be literary style, some unusual occurrence or perhaps in Caesar's account of the pila pinning Helvetian shields together a knowing wink at the traditional effectiveness of Roman weaponry.

I do feel Roy is right about this: even in our own living memory, accounts of combat rarely detail the tactical systems or the procedures taught, concentrating instead on significant occurrences - especially those which do not go to plan - and results.

I suspect that the WRG rules are modelling a push back where the consequence is to give a minor advantage to the winner of each phase. This advantage can be overturned, but if it goes on through the phases of fighting then one side will become the winner.

This makes sense to me.  WRG 6th rules base collapse through recoil on how good or bad a unit's quality was, with good quality units able to take continuous recoil for longer.  Veterans ('B' class) can stand two successive recoils but go on the third, while 'D' class melt away as soon as they are pushed back, making them not the troop quality of choice.  WRG 7th dispense with this, so recoil can go on for quite some time before a unit breaks through becoming exhausted and/or taking double its opponent's losses and three casualties per figure.

I think it goes back to Keegan or even beyond and is based upon an understanding of human physiology, that is that hand to hand combat is so intense that it cannot be maintained for longer than around 15 minutes, perhaps 30 minutes. I suppose it depends upon how energetically they go at it.

I shall eschew the customary quotes of Josephus etc. indicating that Roman soldiers at least were up to much longer periods of continuous combat.  If one thinks of hand-to-hand combat as the kind of energy-wasting flourishes seen on films, then one can imagine people thinking it cannot go on for more than a few minutes.  Battlefield veterans would however tend to know what worked and use it with a minimum of fuss and energy - Spartans in particular seemed able to outlast anyone else in the hoplite era.  Mutual pauses are attested, particularly in mediaeval battles (the occasional notable getting himself shot while resting with his helmet off), but we also have numerous accounts which give the impression of a battle fought out continuously from start to finish over several hours, e.g. Cannae, where Hannibal's men appear to maintain continuous pressure on the Romans and do not allow them a pause to recover.

Hence, in addition to the WRG 6th recoil tolerance by troop quality, I would propose recoil tolerance by troop type, some being more breakable than others (does anyone know of a scythed chariot unit which survived even a single recoil?).

[Jim Webster]

 Seriously I don't think a scythed chariot which actually hit an enemy unit could recoil. It had three options, it veered off, conquered, or died

Jim

[Erpingham]

even in our own living memory, accounts of combat rarely detail the tactical systems or the procedures taught, concentrating instead on significant occurrences - especially those which do not go to plan - and results.

In fact, it is sometimes only through reporting what went wrong that we get a glimpse of what right looked like.  The example given of men being shot while taking a breather is an example of this

This makes sense to me.  WRG 6th rules base collapse through recoil on how good or bad a unit's quality was, with good quality units able to take continuous recoil for longer. 

A simple mechanism.  I can't remember whether there were conditions for a push back (e.g. casualties per figure) or whether it was just killing more than your opponent.

If one thinks of hand-to-hand combat as the kind of energy-wasting flourishes seen on films, then one can imagine people thinking it cannot go on for more than a few minutes.  Battlefield veterans would however tend to know what worked and use it with a minimum of fuss and energy - Spartans in particular seemed able to outlast anyone else in the hoplite era. 

While agreeing about the Hollywood style (although the Romans had a literary trope that this was exactly what barbarians did) the idea of full on combat lasting more than 15-30mins does stretch the limits for human endurance, as Roy says.  Better trained and battle-fit would fight intensely for longer, of course.  But I can't see battles lasting for hours unless there was some kind of variation in intensity.  The Roman military system, as you say, really rated battlefield endurance high on its list of success factors.  Not only were individuals trained to fight longer but the entire tactical system was designed to give the army operational endurance.  It is almost like the football pundits going on about the fitter team having to absorb the attacks in the first 45 minutes and rely on their superior legs to overrun the opponent in the second half

[Patrick Waterson]

I can't remember whether there were conditions for a push back (e.g. casualties per figure) or whether it was just killing more than your opponent.

A bit of both: in WRG 6th, if one side killed even one more than the other side the latter was pushed back ('recoiled') provided the loser took at least one casualty per figure.  Large units in deep columns could get their front ranks salami-ed (although figures were removed from the back ranks, which favoured mixed units with the rubbish at the back) without having to budge, so if the player happened to be one of those masochists who fielded large quantities of 'D' class, deploying them in great depth could hold up almost anything for long enough for his battlewinners to do something useful elsewhere.

Putting less able troops in very deep formations seems to have been a staple of Achaemenid Persian armies, and judging by clues scattered here and there to have been representative of the large armies attested in the Biblical period.  The Greeks occasionally put good troops in very deep formations (e.g. the Thebans 50 deep at Leuctra and 2nd Mantinea) and these seemed always to steamroller the opposition.  They also seem to have done so by keeping up continuous and unremitting pressure until the shallower foe could no longer sustain the push.  Curiously enough, hoplites fighting against deep Theban columns seem rarely if ever to have been broken by them: at Delium the 8-deep Athenian left was pushed back by the 25-deep Theban right while the presumably 8-deep Theban allied left was being routed by the 8-deep Athenian right.  At Leuctra, Xenophon tells us the Spartans were not broken but were forced back all the way to their camp (with the body of their dead king), whereupon the Thebans disengaged and went back to erect a trophy, etc.  The Spartans considered redeploying for another go but their troops were just too exhausted so they called it a day.  (Conversely, Plutarch's Life of Agesilaus suggests the Spartans may have broken, as Agesilaus is on record as carefully avoiding enforcing the punishment for cowardice in action.)

But I can't see battles lasting for hours unless there was some kind of variation in intensity.

This I think would depend upon the combatants: if your weaponry is light (shortsword), your training intensive and your protection familiar and not over-encumbering, the actual physical effort would be no more than an afternoon of logging (which incidentally seems to be the most calorie-intensive sustained activity available to humans, at 7,000 calories/day) and probably considerably less.  Classical societies drew their military manpower principally form farmers - men who were accustomed to hard manual labour all day, practically every day, and for whom soldiering may even have represented a welcome change of pace.  Plutarch's Life of Marius describes Marius' and Catullus' fight against the Cimbri, a battle of some duration in which not one Roman was even observed to sweat, so good was their training and conditioning.

On the vexed question of intervals or 'rest breaks' in battle, I think this depends upon military systems: classical accounts have plenty of references to tribal troops in general and Gauls in particular being reduced to gasping helplessness while their trained civilised opponents are still fresh; it is tempting to conclude that at least some tribal societies grew up accustomed to what we may call 'non-continuous combat', with an enforced lull after, say, 15-30 minutes of fighting, and found themselves at a severe disadvantage when opposed by armies who had brought continuous fighting to a fine art.  On the whole it would seem that some battle styles, particularly those of civilised armies, were continuous and some, including those of several 'barbarian' cultures, were not.

[Erpingham]

Let us assume for the sake of argument that the old WRG recoil mechanism has value in representing Justin's defined battlefield recoil.  A unit bested in melee, if suffering sufficient casualties, falls back a distance.  Depending on its morale/quality rating (not its current morale state), and assuming its opponent presses its advantage,  it can endure this for a number of moves before giving way in rout.  A number of questions can be asked  e.g. :
*Should the winner best the loser by a margin or should a simple better score be sufficient?
*WRG went for 5% casualties as the loss level that could precipitate recoil.  Is this an appropriate level?
*Is basing a break on number of continuous push backs and initial quality better than forcing a morale test each time a unit recoils (i.e. using current morale, not base morale as the key criterion)
I'm sure there are others.

[Justin Swanton]

On the vexed question of intervals or 'rest breaks' in battle, I think this depends upon military systems: classical accounts have plenty of references to tribal troops in general and Gauls in particular being reduced to gasping helplessness while their trained civilised opponents are still fresh; it is tempting to conclude that at least some tribal societies grew up accustomed to what we may call 'non-continuous combat', with an enforced lull after, say, 15-30 minutes of fighting, and found themselves at a severe disadvantage when opposed by armies who had brought continuous fighting to a fine art.  On the whole it would seem that some battle styles, particularly those of civilised armies, were continuous and some, including those of several 'barbarian' cultures, were not.

Tribal armies that went hell for leather most likely were used to very short battles: one side would swing and hack away longer than the other, which broke first. Battle over in 15 minutes. Given their out-and-out combat style, the idea of a sustained fight, even with pauses, was probably foreign to them.

[Patrick Waterson]

Let us assume for the sake of argument that the old WRG recoil mechanism has value in representing Justin's defined battlefield recoil.  A unit bested in melee, if suffering sufficient casualties, falls back a distance.  Depending on its morale/quality rating (not its current morale state), and assuming its opponent presses its advantage,  it can endure this for a number of moves before giving way in rout.  A number of questions can be asked  e.g. :
*Should the winner best the loser by a margin or should a simple better score be sufficient?
*WRG went for 5% casualties as the loss level that could precipitate recoil.  Is this an appropriate level?
*Is basing a break on number of continuous push backs and initial quality better than forcing a morale test each time a unit recoils (i.e. using current morale, not base morale as the key criterion)
I'm sure there are others.

WRG evaluated everything in terms of 'casualties', which if one tries to pin it down seems to have been a bit more nebulous and wide-ranging than simple body count.  The 'casualties' essentially represented a decline in unit effectiveness, or rather the proximity to a step down in effectiveness.  To trigger recoil, a unit had to receive more of an effectiveness decline than its opponent, and this had to be sufficient to shift the unit's stability (i.e. push it back).  Now we come down to numbers.

The 5% across the board criterion looks like a simplification: some very poor units might run when the first man falls, or even before contact (the Persian left at Cunaxa being a case in point) - this is more probably the territory of the Reaction Test of WRG 6th, a morale and disciplinary threshold to be overcome before combat could be joined.  Once joined, in WRG rules the likelihood of an exactly drawn round of combat between well-matched opponents is quite good, and any imbalance (one side inflicts one more casualty than the other) would seem to represent, or be intended to represent, the accrual of sufficient advantage to start forcing a decision.  If this interpretation is correct, then the system is translating sufficient advantage to confer superiority into a 1-casualty lead, so the threshold is acceptable.  One may note in passing that swings in the random factors of both sides might lead to a back-and-forth engagement, although once one side is recoiling the other receives a "following-up" bonus, making it harder for the trend to be reversed.

Using a number of push-backs related to troop quality as the basic criterion for reaching the rout threshold allows a better-quality unit more chances to attempt to reverse the trend (e.g. by adding a general, having another unit interpenetrate and take over or just pinning hopes on the random die roll getting better).  It seems to work for the system: better units have more chances to reverse the trend, and can generally hold on for longer as one might expect from their generally superior cohesion, while less good ones crack soon after they start crumbling.

What strikes me about the way WRG handles this aspect of combat is the way the elements are integrated: each aspect is part of a system and the way they relate gives an effect not unlike the yielding of one side followed by its recovery when joined by a general (or otherwise helped out) - or its collapse if the rot is not stopped - that we see in numerous source accounts.  The individual aspects may look arbitrary and artificial (can the demise of one man more than the opposition make such a difference? Is it really appropriate to have the Theban Sacred Band pushed back through taking a mere 15 casualties?) but it seems to hang together well enough as a whole.

If in quest of deeper realism, we would have to keep in mind the way the various elements of our new system interact.  For example, fresh troops are rarely pushed back except by overwhelmingly superior opposition: in anything resembling an equal fight it seems to be tired troops, or troops who have been fighting for a while, who customarily 'recoil'.  We could thus build in a mechanism which says that it takes a 2:1 'casualty ratio' to force back fresh troops but only simple superiority to force back tired troops.  We then have to look at what this does to the rest of our rules, and if there are no problems then we may have moved one step closer to representing reality.

Then we have to try it on a variety of army types - some may bring unexpected anomalies.  If on the other hand the tabletop troops start behaving like the accounts of their real-life counterparts, we probably have a winner.