Overhead shooting in the late Roman line

[Justin Swanton]

This one has always puzzled me. The late Roman legion relied heavily on missile fire. Part of its repertoire was to have the back ranks of its infantry line armed with bows. These troops would shoot over the heads of the front ranks into the enemy and decimate/disorganise them whilst the front ranks fought them hand to hand.

In my youth I did archery as a hobby, and I can say straight out that there is no way you can stand at the back of a body of friendly troops and hope to shoot another body of enemy troops just behind them without turning your buddies into kebabs. Also the idea of shooting straight up so the arrows fall on an enemy about ten yards away just does not work. A slight gust of wind and you are the one who gets the point.

Any ideas on how back rank shooting could have been made to work?

[aligern]

Two ways Justin,
Firstly the back ranks shoot at a defined angle to the horizontal. Their officers start them at 45% and then drop them to 30 and then 20 as the opponent advances. It's a simple enough drill to shout position one, position two etc. shooting at the different positions will result in a different beaten zone for each so the arrow fall will track with the enemy as he comes on.
Secondly the front ranks , armed with verutum and dart could kneel whilst the archers shoot at an enemy200 yards away, rising up when he reaches 50 yards so that they could use their darts.

We know that Romans, Byzantines and Arabs used back rank archers. I think it pretty likely that goths do the same.  From Maurice's description of Byzantines I would expect a combination of the two ways I suggest above with the front ranks leaning forward to facilitate shooting.

Roy

[Justin Swanton]

OK, that makes sense. Back rank shooting took place whilst the enemy was still a distance away. Presumably the archers put their bows away and took hold of melee weapons once the lines met.

My 42 pound recurve bow was inclined up at about 20 degrees when shooting a target 100 yards away. You would be shooting at 45 degrees from 200 yards. It would be better for the front ranks to be kneeling for anything less than the extreme ranges unless the ranks of archers are a few yards away from the melee troops.

[aligern]

Of course the front ranks of a Late Roman legion are missile armed so they need the last few yards of an enemy advance to throw their javelins. I imagine that arrows are less effective against opponents charging in with shields when shot at a flat trajectory so the formation isn't losing that much firepower as the archers cease and the scutati take over.  Of course, if cavalry are charging the unit the archers have a target all the way through.  One wonders too whether , as in the old WRG 6th ration days the legion might be deployed on a rise or hill so that the back ranks were naturally elevated?
Roy

[Mark G]

it raises one interesting question for me.

In the accounts of Hoplites fighting Sparabara, are their any references to the bow fire ceasing at a certain range - or conversly, the front rank spearmen getting out of the way to allow a flatter trajectory?

i.e., are we looking at a mortar style minimum range for mixed spear and bow units?

[aligern]

Yes, good point Mark. Once the wall of shields is raised I cannot see how the front rankers could drop it and get it back up in any order. Your minimum distance idea is a good one. That would help explain how the Greeks changed the rules of the game by running forward at marathon.
Roy

[Patrick Waterson]

Ammianus' account of Argentoratum suggests that missile use continued during hand-to-hand fighting.

"Then a cloud of hissing javelins flew hither and thither, the dust arose with steady motion on both sides and hid the view, so that weapon struck blindly on weapon and body against body. [44] But the savages, thrown into disorder by their violence and anger, flamed up like fire, and hacked with repeated strokes of their swords at the close-jointed array of shields, which protected our men like a tortoise-formation." - XVI.12.44

"But the Alamanni, who enter eagerly into wars, made all the greater effort, as if to destroy utterly everything in their way by a kind of fit of rage. Yet darts and javelins did not cease to fly, with showers of iron-tipped arrows, although at close quarters also blade clashed on blade and breastplates were cleft with the sword; the wounded too, before all their blood was shed, rose up to some more conspicuous deed of daring." - idem, 46

The repeated waves of Germans imply considerable depth to their dispositions, as does the consideration that they outnumbered the Romans 3:1 on a similar frontage.  Most of the missiles were probably travelling from the Roman rear ranks to lines of Germans not yet in melee - and Ammianus' observations that the Germans were fighting uphill and there was a clear slope all the way to the riverbank suggest that the Roman missilemen might even have been able to see their targets, at least before the dust became too prevalent, and could thereafter probably adjust their aim by the kind of noise (howls of pain or nothing) their volleys were generating when they landed.

[Erpingham]

Ammianus' account of Argentoratum suggests that missile use continued during hand-to-hand fighting.

It does go back to the issue of what hand-to-hand fighting was like.  If it wasn't continuous, but the lines eased apart at intervals, the archers could restart harrassing shots without hitting their own side.  If it was continuous, as you say, the implications are of deep formations, or multiple lines, which could be engaged by the rear ranks while the front ranks have at it.  Also, we should watch the ammunition supply issue.  I don't know how this worked in the late Roman army but I'd guess the officers would not want their men shooting when there wasn't much chance of having a meaningful effect, keeping arrows back for critical times.

[Mark G]

and frankly, I would not want to be in any front line that allowed men behind me to try and throw over my head.

Mind you, my head is a lot closer to the stars than most of those guys were.

there are some diagrams here

http://garyb.0catch.com/line3_pilum/pilum_volley.html

on Pila thorwing which are pertinant.  the example from the OP on the problem with a bow attempting to hit a target at close range without a flat trajectory are also relevant.

it does seem to me that the problems do suggest that this reference to missiles while fighting can be understood better by accepting a non continuous melee model, and having the missiles occurring at a time when the front lines are apart (but still 'fighting'), rather than as having things whirling overhead while you are raising arms and chopping down with axes.

I can see a second row man throwing an axe or dart over my shoulder with a bit of team training - just as him issue a mighty blow while I did the blocking for him could work with training. 

But any further than that, and the guy throwing is not going to see the target anyway, and there are all those variables with backwards space and the need for a run up, the loss of velocity from the need for a high angle, the strength of helmets and everything - it seems like a nuissance weapon  for men not in the front line, not a killing weapon at best the further back it comes - and the greater the risk to your own men from a mis timed or wind gusted missile.

[Patrick Waterson]

It may also be worth considering terrain.  Argentoratum was fought on a continuous slope, and I suspect that, given the keenness of Germans to get into a fight, Late Roman armies would where possible deploy on a slope and let the enemy come to them. ("They came on in the same old style and were beaten off in the same old style.")  This would also incidentally give the rear ranks (those who loose missiles overhead) a decent view of the enemy rather than a blind loosing-by-numbers situation and a lower 'safe' trajectory.

Another advantage of an up-slope, on-slope deployment is that when the attacker churns uphill he kicks up such dust as is around, but while this impedes his vision the defender's missile-using troops are above the dust and can still pick their targets.

All this suggests that the Late Roman army's infantry was tactically and organisationally optimised to fight on slopes.  Could this have been a conscious evolution of doctrine?   Vegetius says this:

"Good generals are acutely aware that victory depends much on the nature of the field of battle. When you intend therefore to engage, endeavor to draw the chief advantage from your situation. The highest ground is reckoned the best. Weapons thrown from a height strike with greater force; and the party above their antagonists can repulse and bear them down with greater impetuosity, while they who struggle with the ascent have both the ground and the enemy to contend with. There is, however, this difference with regard to place: if you depend on your foot against the enemy's horse, you must choose a rough, unequal and mountainous situation. But if, on the contrary, you expect your cavalry to act with advantage against the enemy's infantry, your ground must indeed be higher, but plain and open, without any obstructions of woods or morasses." - Epitomai Rei Militaris III.12

On level ground, it would still be necessary for someone to 'call the numbers' for shooting, most of which would be volleys by order.  The old questions about how to get arrows to land where desired would still apply.

Concerning pilum-throwing, the current estimate of c.20 yards may be much too modest.  Remembering Caesar's account of Pharsalus:

"Between the two armies, there was an interval sufficient for the onset: but Pompey had given his troops orders to keep their ground, that Caesar's army might have all that way to run. This he is said to have done by the advice of C. Triarius, that the enemy's ranks might be broken and themselves put out of breath, by having so far to run; of which disorder he hoped to make an advantage. He was besides of opinion, that our javelins would have less effect, by the troops continuing in their post, than if they sprung forward at the very time they were launched; and as the soldiers would have twice as far to run as usual, they must be weary and breathless by the time they came up with the first line." - Caesar, De Bello Civili, III.92.

A run-up of this nature would have significantly increased the range of the pilum.  However the limited range from a standing start may help to explain why in the Late Legion back-rankers used lighter javelins or bows: this would make sense if the later legion were expected habitually to fight as a static entity holding its ground, as did the Primani at Argentoratum.

Patrick
[Edit: 'gorund' to 'ground' - in case anyone wondered how a soldier could hold his 'gorund']

[Justin Swanton]

One thing does come to mind: target archery involves the use of a sight, which is raised or lowered on the bow depending on the distance of the target. Hunting archery does not use a sight and the draw technique is different. A target archer pulls the string to the nose, mouth and chin, and aims by lining up the string and sight pin on the target. A hunter pulls the string back to the ear whilst his face looks forward. He must have a feel for the distance; he doesn't have a mechanical way of calculating it.

That being the case,  a skilled hunter-archer can place an arrow at any distance he chooses, whether he can see the target or not. He notionally conceives the target as being 200 yards away and holds his bow at the correct elevation accordingly. Since arrows will have a high trajectory at extreme ranges, an archer can target enemy troops in melee with his own side provided he is a good distance away and able to lob his arrows in at steep angles. It would require training, but it could be done.

[Patrick Waterson]

Thanks, Justin: as a matter of fact, this post seems to pretty much answer the original question!

[aligern]

Picking up on Patrick's quoted section on Pharsalus
1) Clearly both sides expect to run towards each other releasing javelins.
2) it is highly likely that at a certain distance pila are released mutually. Of course, if the enemy were Celts standing still the run up would start that much closer to the enemy.
3) after release one would expect the lines to run into one another , but in good order, because we are told that if the legionaries had to run double distance they would be disordered.
4) as Mark G and I have said pretty clearly in Slingshot this is a tightly controlled process. The Pompeians bet is that, once commenced, the Caesarians are on autopilot and will have to keep running, but they are not, their  centurions halt them and  recommence the procedure.
5) I don't think that this is an argument for throwing at a greater distance because here we have two bodies that are supposed o close mutually and rapidly. As the point about running further causing confusion shows, it is important to avoid running too far and thus releasing early at a longer range has no benefit at all. The a in is most likely a disciplined release at a trot at 15 or so yards and then a disciplined move to contact at the same pace. The legion could not thrown and stop.
There are other instances that Patrick can possibly magic up here where closing fast precludes pl
Ilum use, one I think against Ariovistus and one against the Etruscans by. T.he Etruscans?come to
mind

Roy

[Justin Taylor]

I think I agree with a couple of posters, overhead shooting is allowed in TDIC because it happened. However I restrict it to stationary troops only.

Yes Romans throw pila but just before combat is joined, not like with bows or even javelins as a missile weapon. The other thing I have done in the latest version of the rules is to allow all troops with missile weapons to shoot during a charge. I figure that if the Romans can throw pila then why should the Gauls not be allowed to throw javelins (note that pila are more effective than javelins)?

Close order troops are only allowed to throw javelins at short range because unlike skirmishers they are assumed not to have many javelins per man, to throw.

I can see arrows coming down from a high trajectory as being more effective as they could then avoid hitting shields. Likewise overhead fire is not aiming particular targets but rather creating a beaten zone of fire and hitting men just comes down to them being unlucky or not.

[Patrick Waterson]

5) I don't think that this is an argument for throwing at a greater distance because here we have two bodies that are supposed to close mutually and rapidly. As the point about running further causing confusion shows, it is important to avoid running too far and thus releasing early at a longer range has no benefit at all. The aim is most likely a disciplined release at a trot at 15 or so yards and then a disciplined move to contact at the same pace. The legion could not throw and stop.

The point is that both sides have a good run-up prior to throwing as part of the expected standard procedure.  We know it is a run-up rather than a walk-up (or trot-up) because continuation would result in disorder (continuing a disciplined trot would not).  A run-up adds to the range of a hand-hurled missile weapon.  I do not think the re-enactor estimates of c.20 yards for pilum range are based on anything other than throwing from a static position.  Ergo, methinks pilum range was greater than this in practice.

Timing may also have been a consideration: closing at a 3 mph walk covers about 1.5 yards per second.  Upping this to a 6 mph trot covers 3 yards per second.  A good 9 mph run closes at 4.5 yards per second.  A pilum would take perhaps one second to fly ten yards (please feel free to replace this with a more accurate figure if known).  Its flight is a low parabola, up a bit then level a bit then down a bit to land point-first - each phase taking about one third of the flight time and distance.  Two running lines mutually close at 9 yards per second.  Releasing pila at 15 yards from the opponent means they will be half-way through their trajectory when the lines close, and will land 7-8 yards behind an 8-deep foe.

One would prefer to release pila at the point where they could fly for three seconds and make contact at the end of that time.  Hence one needs to throw when the enemy are at least three seconds away, which means three seconds of pilum flight time and distance plus enemy closure distance.  Hence pila need to be launched when a mutually closing enemy is 30+13.5 yards away - 43.5 yards distant.  The pila will travel 30 yards and the foe 13.5 yards in the intervening three seconds during which one's own men will have time to draw swords and put up shields against what they are about to receive (the discrepancy of loss at Munda mentioned in Spanish War 31 indicates a possible lack of appreciation of this latter point by Pompey's recruits).  In practice, assuming a launch at 45 yards' distance against a mutually closing target is probably sufficient, which suggests beginning the run-up at about 80-85 yards from the foe.

Loosing pila against a stationary opponent would be done at about 30 yards - there is no need to add in a vector for closing.  Begin the run-up from about 50 yards away, loose at 30 yards and gravity does the rest.  Again, 15 yards could be running into trajectory problems unless the opponent is deployed very deep or one is happy to settle for a weak throw.

That at any rate is my 'take' on that particular subject.

[Edit: adjusted the closing vector because once pila leave the hand the forward velocity of the thrower no longer counts!]