What happened when Cavalry charged Cavalry?

[Andreas Johansson]

Re Horse sizes, IIRC Duncan turned up some figures on Nisaen horses when we were discussing Persian logistics which put them in the 350-400kg range.  There is plenty of information about the height of ancient horses on line but not a lot on weight, probably because horses of the same height can vary quite a bit in weight depending on breed.

Horse Mart has a table here showing approximate healthy weights for horses of different heights and breeds. As you'd guess, weight differs a lot by breed - a draught horse weighs about 50% more than a thoroughbred of the same height - but it seems reasonable to assume most (healthy) ancient horses were within the range of modern ones.

[Nick Harbud]

What appears to be established is that a decent-sized horse can burst through an infantry line 8 ranks deep (a typical depth) unless the infantry do something about it. My guess is that they originally deployed in great depth before later switching to ranks bunched up together, othismos-like, which turned them into a stable mass a horse couldn't knock over.

...that and the pointy sticks probably put off the horses somewhat.

[Patrick Waterson]

However we look at it, a determined horse and rider seem unlikely to be stopped by less than eight ranks of infantry, and then only if the infantry can provide some form of coherent grouping to reduce the impact and/or deterrence to avoid the impact taking place.

Arrian's lineup against the Alans is indicative: it has three ranks of kontoi, which provide deterrence both long and sharp (I am now entirely convinced that kontoi in this context cannot be understood as a term standing in for pila) and supporting missilemen behind.

All of this suggests that cavalry were, or could be, quite effective as an anti-infantry shock force unless the infantry were effectively trained, and preferably suitably equipped, to counter them.

Getting back to the purpose of the thread, namely the interaction of cavalry, we have a similar consideration except that the opposing troops are going to be of approximately equal mass and better able to manoeuvre to avoid a full-on collision.

So what are the aims and intentions of cavalry charging cavalry?

The obvious intent is to break the enemy formation or, on an individual level, to insert one's weapon into the enemy rider where it will do most good (or at least be most effective).  Usually this means a neck or body thrust or cut, but Alexander taught his troops to strike for the face, which if not an instant kill was a probable unhorsing.  One notes that in the Alexander Mosaic he appears to have ignored his own advice and run his opponent through the midriff.

At the Granicus, when Alexander himself was attacked, the one opponent (Rhoesaces) who landed a blow did so to his helmet, indicating the use of a slash rather than a cut, and Alexander responded with a thrust into the chest through the breastplate.  On the single occasion where he had room for a run-up, he used it to hit his opponent in the face with a xyston, knocking him to the ground.

Macedonian and Achaemenid cavalry, and probably others, appear to have placed a premium on formation-keeping while in melee.  Spithridates attempted to strike Alexander as the later was despatching Rhoesaces, but was anticipated and disarmed (literally, in Arrian's account) by Black Cleitus, who was presumably next-in-wedge behind and to the left of Alexander.

In order to keep formation in melee, contact would have to be disciplined rather than uninhibited.  This need not rule out contact at speed, rather it emphasises the need to keep station even at speed.  We do not however read of horses smacking together against each other and sitting down on their haunches from the impact, which suggests restraint in contact speed or care in aiming to contact the opposing line between two opposing horses, distributing any closure force into semi-cushioned angular impacts.  However I think the determining and limiting factor for contact speed was the need to be able to use weapons.  A full gallop is perhaps ruled out, although Roman cavalry apparently closed at speeds which made the points of their cuspides wobble, but speed of closure would seem to be considerably more than a controlled trot.

[Dangun]

Firstly a comment on terminology.
This is not charging into contact...

Crassus' Gallic cavalry at Carrhae charged straight into Parthian cataphracts and attempted to seize their spears and grapple them from their horses.

You cannot "seize their spears" or "grapple" if you are actually charging. This is melee, not charging.
The example of Cannae, illustrates this.

If we don't have sources telling us that horse ploughed into each other, might we consider that it would be a stupendously silly thing to do. Quite apart from the high probability of crippling your horse, there would be a near 0 chance that the rider stays in the saddle if two charging horses actually collided. If you think that a horse can knock over 8 infantry men, consider that the potential acceleration for the riders of TWO HORSES colliding would be as much as twice as great as that suffered by the first infantrymen knocked over.

[Patrick Waterson]

You cannot "seize their spears" or "grapple" if you are actually charging. This is melee, not charging.

But you can as soon as the charge results in contact.  The cavalry are not frozen in a state of perpetual charge;  upon contact the charge transitions to the melee.  I would also suggest that cavalry do not end up close enough to grapple without at least one side charging the other.

... the potential acceleration for the riders of TWO HORSES colliding would be as much as twice as great as that suffered by the first infantrymen knocked over.

This is a common mistake (I blame the way physics and arithmetic are tauight in schools).  The acceleration, or more accurately momentum, of each rider is simply that imparted by his forward velocity, which does not double when he is suddenly halted.  If he is moving at, say, 15 mph prior to contact, he is still moving at 15 mph at the moment of contact, and 15 mph less the cushioning effect of his saddle, horse's neck etc. after contact.

Regarding colliding horses: I would suggest that the natural aim of a cavalryman is to try and place himself (and hence his mount) to deliver a stroke against an oncoming opponent, which is best achieved if he can get that opponent at a slight angle as opposed to straight ahead.  Hence when two formations of cavalry meet, each cavalryman will most probably be aiming to insert his horse between two horses in the enemy front rank, meanwhile attempting to insert his weapon into the enemy rider appearing ahead right or ahead left (the said rider will also be attempting to avoid the said weapon).  The result would be a coming together (if one may use the term) of both cavalry formations in a sort of saw-tooth pattern.

We might ask: what of the horses?  Newton's 'action and reaction are equal and opposite'* suggests that if two horses actually meet head-on (as opposed to the partial side-collisions the above would involve) then both will bouce back and be left sitting on their haunches.  If their riders have the type of saddle the Romans used, they will be left in position inelegantly leaning forwards; if not, they might well be departing for firmer ground at that point (and reaching it shortly afterwards).  But the partial impacts of offset horses are likely to be mutually absorbed with less fuss and consequences than we might expect.

*Seemingly inapplicable to missile weaponry, which appears to have its own rules.

Others have pointed out that when impact becomes inevitable, cavalry will tend to slow down somewhat in the final stages before impact.  Whether this was universally true I know not; indeed, I suspect that cavalry trained to accelerate as soon as they see opponents slow down would have a significant advantage in melee.  Be that as it may, there do seem to be numerous 'soft' factors which shade down the apparent simple physics of such collisions.

[RichT]

I'm no physicist but I'm sure this is wrong. I don't think acceleration is the relevant measure - the deceleration of horse A might depend only on the velocity of horse A at the point of impact (and granting for now that horses come to a dead stop on impact). But the energy or force applied depends on the combined velocity or momentum of horse A and horse B. Consider the difference in force applied and therefore damage caused between a horse heading north at 1 m/s stopping dead on encountering a stationary cannonball (damage to horse = negligible), and the same horse encountering a cannonball heading south at 300 m/s (damage to horse = severe). The velocity of both objects makes a big difference in an impact.

[Erpingham]

This needs more physics than I can conjour up but part of the issue is to decide what measures we think are significant.  The momentums of the two horses largely cancel out, because their velocities are in opposite directions (momentum has a vector, like velocity).  The energy of the collision, however, doesn't have a vector and the two kinetic energies are combined.  That energy will be converted into various things - heat, sound, deformation of/damage to horses and residual kinetic energy of the collided mass.

[Andreas Johansson]

I'm no physicist but I'm sure this is wrong. I don't think acceleration is the relevant measure - the deceleration of horse A might depend only on the velocity of horse A at the point of impact (and granting for now that horses come to a dead stop on impact). But the energy or force applied depends on the combined velocity or momentum of horse A and horse B.

I've lost track of quite where this discussion is going, but in terms of physics, acceleration is directly proportional to force. Newton's second law: force equals mass times acceleration

Kinetic energy is proportional to mass times speed squared. Because of the square, it does not relate simply to momentum (mass times velocity)

Force is what actually breaks bones and whatnot: having the same kinetic energy dissipated is going to have very different practical effects depending on how rapidly it is dissipated, which determines the peak forces you're subjected to.

(Consider slowing down from a run to stationary normally versus by running into a brick wall. You're shedding the same amount of kinetic energy in each case, but in the latter case you do so much more rapidly, so forces are much greater.)

[Patrick Waterson]

I think we were getting hung up on the amount of forward kinetic energy a rider would have if his mount were to encounter another in full career.

Andreas has covered the essentials.  A couple of other points are worth mentioning.

The force of two horses colliding (if we can get them to collide head-on rather than offset) gives the horses a bump but not the riders.  The riders simply have to put up with a 15-to-0 deceleration which is usually handled by their saddle; failing that, the horse's neck. (If the horse puts its head down at the moment of stopping, the rider could well take a tumble unless held by a Roman-style saddle.)

The energy of the collision, however, doesn't have a vector and the two kinetic energies are combined.

This is true of the horses - assuming a straight-line impact - but not of the riders.  The rider is not an integral part of the horse, and indeed articulates freely at the point of the saddle.  The riders only have to worry if both go sailing forward off their saddles into each other.

But would they go sailing forward off their saddles?  When TE Lawrence witnessed a charge of Arab camelry against Turkish cavalry, the greater mass of the camel bowled over the Turkish horse and rider together; the Turk did not immediately become separated from his mount.  I do not think we can simply apply the physics of a Newton's Cradle to cavalry impact.

[Erpingham]

Perhaps we should stop the mechanics bit, as it isn't advancing understanding?  I'm also losing track of what we set out to decide. 

[Patrick Waterson]

I think the point at issue was what happened when man on mount collided with man on mount as the result of mutual high-speed closure.  (Actually it was whether the rider should be counted as going at 30 mph if two mounts met, each going at 15 mph.)  This has implications concerning whether a mutual full-on charge would result in shattered or merely rebuffed horses.

Classroom physics would seem to be a dead end; can anyone think of any contemporary accounts which might shed light on the matter?

Do you remember in Malory there were occasions when two knights' mounts collided full on?  I seem to recall that the result left both horses sitting back on their haunches as opposed to being blobs of jelly and bone chips.  A few saddles may have come apart in addition, and I cannot remember exactly where the riders ended up.  My assumption would be that Malory is reflecting and portraying the kind of thing his readers would expect from events in his own time; would this be a reasonable assumption?

[Dangun]

can anyone think of any contemporary accounts which might shed light on the matter?

The absence of accounts is striking.

But you can as soon as the charge results in contact.  The cavalry are not frozen in a state of perpetual charge;  upon contact the charge transitions to the melee.  I would also suggest that cavalry do not end up close enough to grapple without at least one side charging the other.

So are you saying they charge, but stop short of the horse impacting another horse, and then start grappling?
What was the point of charging? Or do you mean you interpenetrate each other, for a complete loss of both side's cohesion. I really don't follow what is being suggested. This "charge" word is  ill defined.

[Patrick Waterson]

So are you saying they charge, but stop short of the horse impacting another horse, and then start grappling?
What was the point of charging? Or do you mean you interpenetrate each other, for a complete loss of both side's cohesion. I really don't follow what is being suggested. This "charge" word is  ill defined.

What I am attempting to convey, however indifferently expressed, is that at Carrhae we have an example of two cavalry formations making intimate contact.

It is physically impossible to grapple an opposing rider without the horses having already inserted themselves into the opposing formation.  I would not imagine the Parthians obligingly held open gaps for the Gauls to walk their horses into before everyone began close quarters activity by mutual consent, which leaves us with one other option: they charged into contact.  Any grappling would take place after horses had impacted and/or avoided each other.

My understanding is that cavalrymen would aim to insert themselves into the gaps between opponents' horses.  If there are no gaps wide enough, then bumping equine shoulders with two opposing horses is preferable to bumping heads and/or chests with one.  It is also very difficult for cavalrymen to use their weapons against an opponent directly ahead - the horse's neck and head are in the way.  Hence one needs the target slightly offset to one side.

At Carrhae, the amount of interpenetration would be limited, perhaps only the first rank in each formation.  The formations themselves would be perhaps ten deep, so there would be no loss of cohesion from this cause.

Precisely defining 'charge' can be something of a challenge; the concept is essentially non-stop closure intended to result in melee and at greater speed than one's customary advance.  In short, closure with acceleration and impact.  The intention is usually that the impact be with one's weapon against an opponent, preferably in a suitable spot for disabling same.

Quote from: Patrick Waterson on November 23, 2018, 08:25:36 PM
    can anyone think of any contemporary accounts which might shed light on the matter?

The absence of accounts is striking.

Is it, though?  We lack accounts of most tactical procedures (hence the extensive discussions about same on these forums), so the lack of specific information on man-to-man (or horse-to-horse) encounters is not particularly noteworthy.

[Jim Webster]

My understanding is that cavalrymen would aim to insert themselves into the gaps between opponents' horses.

I would suspect that there were always gaps wide enough, the unit couldn't maneuver if the horses were literally shoulder to shoulder. Also the riders legs would be trapped
There has to be some space and other horses would push into that space, probably at not much more than a walk by the time they got there (Because the actual pushing would slow them down a lot even if they did try it faster, the horses would act as a brake on each other)

[Erpingham]

This has implications concerning whether a mutual full-on charge would result in shattered or merely rebuffed horses.

There are numerous images of horse v. horse collisions at speed on you tube.  They are not nice, as horses end up crippled or dead in several.  There is one good one in which two American quarter horses collide shoulder to shoulder at speed.  One horse and rider are sent sprawling.  The other horse stays on its feet but the rider goes over its neck.  Thankfully, both horses and riders survive this one.  So, from this small sample, we can probably say "shattering" isn't inevitable but it certainly happens.  This is also what you'd expect from any exposure to horse racing - not all falls and collisions result in death or serious injury but a proportion do.  The shoulder-to-shoulder collision is interesting in that you might expect a horse to avoid a head-to-head impact by angling its neck, with the impact being on the neck and shoulder, as in this case.

As to actual jousting practice, I've not read the many detailed accounts that exist to check them for collisions.  I would note though that the barrier or list between the two riders was introduced to reduce collisions and injuries to (very expensive) horses in the 15th century.