Today we’re going to talk about what a pre-modern battlefield might look like after the battle is over. Obvious content warning, since this post is going to talk about (and show pictures of) some very ugly things.
Popular media has a particular image for the post-battle battlefield that shows up in film, TV and video-games in the aftermath of a medieval or ancient (or comparable fantasy) battle. We can outline the trope:
The ground is invariably muddy and soaked. It seems to have always rained just the night before, but more to the point, something has killed all of the grass, trees, and shrubs. We are often shown make-shift wooden obstacles – little palisades, stakes, abatis – scattered around, more or less at random, unless they are part of a large, elaborate system of field fortifications. Some – or all – of these will be on fire. Frequently, these obstructions are attached to larger networks of trenches, braced with wood (which may also be on fire).
The bodies on the field, typically still in full kit, are scattered around more or less evenly, or sometimes in little clumped groupings (if they haven’t been gathered for burial). Many objects will either still be inexplicably on fire (despite the rain implied by all the mud) or be smoking from some previous fire.
Also: why has no one bothered to try to take all of these expensive weapons and armor? That is valuable stuff.
Also a good example of the “random bits of broken wooden stuff” method of depicting battlefields.
You can see battlefields like this most frequently in video-games, like the Witcher 3: Wild Hunt (in Velen) or Dragon Age: Origins (at Ostergar) or Dragon Age: Inquisition (the Exalted Plains) or in TES V: Skyrim (especially in the siege outside of Whiterun). The basic visual shorthand also shows up in movies, such as after the big final battle in Netflix’s Outlaw King or in the big battle in S1E1 of The Last Kingdom. Game of Thrones even gets into it.
Historically speaking, the Battle of Loudoun Hill was fought on a narrow front, surrounded by marshland, so we might give this a bit of a pass.
Is this trope accurate? Yes and no, it turns out. It has a ring of truth to it, but not for the Middle Ages, or indeed any point in the pre-modern period. Instead, this depiction is the product of a very different battlefield environment and very different technologies.
Verdun in Velen
Let’s start with the mud. Why is it always muddy? Now, astute readers of medieval history will not that some battles – Agincourt famously – turned on the recent rains turning fields into difficult-to-navigate mud. But, of course, concealed in that fact is that the French were unprepared to fight in that terrain because it was unusual. It doesn’t rain every day, and even when it does rain, unless it rains very hard, usually the ground-cover (read: grass) is sufficient to avoid turning the place into giant mud-pits.
I’ve seen it suggested (mostly by well-meaning fans) that maybe it is the marching of soldiers that churns the ground into mud. And obviously we’ve all seen paths in the ground where the repeated, long-term marching of feet has killed the grass. But note that words: repeated and long-term. In practice, we don’t need to wonder if a day or two of battlefield maneuvers is enough to kill the grass like this, because we know: there was extensive photography of the battlefield at Gettysburg and the grass is just fine. One day’s trampling was not enough.
(I’m going to repeat my content warning here one more time: I am going to show historical photographs of real historical battlefields that include real actual casualties. That may be disturbing, and no one will fault you for not subjecting yourself to it.)
(I should note that assembly, camp and logistics areas might well see enough traffic – horses, heavy wagons, etc – to clear the grass, but you do not generally camp or store supplies in between you and the enemy. While in the ACW, such sites might have been within reach of artillery, in a pre-modern battlefield, they would usually be well out of range of the battle itself).
So why is is always muddy? The real answer is high explosive shells (particularly, but not exclusively, penetrating high explosive shells). Heavy artillery shells in the First World War were made to penetrate into the ground and then explode, sending up a rain of loose dirt – the idea was to be able to destroy or at least bury trenches and deep bunkers. The explosions were so powerful that they uprooted trees and grass, leaving behind the ‘blasted moonscape’ so common in pictures of the Western Front. All that remained were the deep craters which collected water and turned into often fatal mud-traps (Peter Jackson’s They Shall Not Grow Old (2018), includes a horrific description of one man, unable to assist, having to watch another man sucked under the mud to his death in such a crater).
This kind of terrain – with so much of the ground-cover blasted away – would turn into mud-soaked pits the moment it rained – particularly where water collected in the shell-holes.
That also explains why these post-battle scenes often lack any kind of local terrain features. Powerful explosive shells could annihilate terrain features like forests, roads, hedgerows, fences, fields – even hills and entire villages – with extended bombardments. And without any ground-cover left, almost any rain at all will then reduce the local terrain into a mud-soaked bog, especially if the local soil drains poorly (as it did so famously in Flanders).
The problem with depicting medieval, or even early modern battlefields this way is, of course, that these armies do not possess any weapons which can deliver this kind of destruction. Even as late as the American Civil War, field artillery – even massed field artillery – was simply not that powerful (although some heavy naval and siege guns were beginning to come close). Post-battle photography of Gettysburg – even in the approaches to Cemetery Ridge and around the Wheat Field – areas of fierce fighting – shows not only trees and ground-cover, but even fences and buildings largely intact.
Field artillery firing solid shot from 6 to 20lbs to is simply not strong enough to tear apart the terrain in the way that we often see in popular depictions of historical or fantasy battlefields; as pictured above, the guns doing that in WWI were often firing 1,000+ pound shells, 100 times the weight of shot of a normal ACW cannon (lighter artillery, like the famed French 75 (Matériel de 75mm Mle 1897) still fired lighter shells – the French 75 fired a c. 12lbs shell – but still had far more explosive power due to improvements in explosives; that said, the French 75, a capable field gun, was famously too light for ideal use in the trenches). Massed musketry won’t do it either and so massed arrow or crossbow fire, catapults or whatever else certainly won’t.
(This, as a side note, may go some distance to explaining why First World War commanders were so unprepared for the challenges the new terrain they were creating in turn inflicted on them. Doctrine said that the solution to well-entrenched infantry was to mass artillery against them – blast them out of position. It had never been the case before that such massed artillery would render the ground itself impassible, because the artillery had never before been powerful enough to do so.)
Field Fortifications
One of these days, we’ll have a longer discussion about the use and abuse of field-fortifications in depictions of pre-gunpowder battlefields. But to put it very quickly: armies before the advent of gunpowder did often dig ditches and trenches, but not to shelter in them. Pre-gunpowder armies simply cannot create the volume of long-range direct-fire to make trench-works make sense outside of siege conditions (and even then, it was more common to build a shelter or an earthwork wall to provide cover).
Instead, these ditches are designed to foul advancing infantry and (especially) horses. A fall into a trench can easily disable a horse, and an infantry with a spear is at a severe disadvantage if he has to fight his way out. Digging trenches or ditches could thus impede or control enemy movement on the battlefield (something, I might note, Outlaw King gets right). Such obstructions could be used, for instance, to channel enemies away from your vulnerable flanks, or into difficult terrain. But if the ditches are just there to foul up the enemy advance, there is no real need to put in lots of wood bracing, walkways or complex structures.
At most, you fill them with sharp things – stakes and the like – and call it a day.
But the larger problem in many of these depictions is the amount of field fortification. Many of these scenes, with vast, complex and apparently interconnected networks of trenchworks (the Exalted Plains from Dragon Age comes readily to mind), imply armies sitting in starring matches for months. Often these sorts of stalemates are explicitly presented as part of the narrative (Velen and the Exalted Plains both have that going on).
The thing is, as we discussed with the loot train, pre-modern armies generally do not do this because they cannot do this. As I tell my students, an army is like a shark, if it does not swim, it will sink. Pre-modern (specifically pre-railroad) armies must generally keep moving in order to continue to supply themselves.
(Of course, long sieges do happen, but they can happen precisely because the enemy is bottled up and thus not interfering with either lines of supply or foraging actions. Doing this around an active enemy field army is almost impossible; check out P. Erdkamp’s Hunger and the Sword (1998) for a more in-depth discussion of how even the presence of an enemy army can shut down foraging operations).
For an army that is blocked in the field by an opposing force, the supply clock starts ticking immediately and it ticks down fast. Most pre-modern armies march with perhaps just a couple of weeks of supplies (again, see the loot train post to see why the consumption/transport math makes this situation functionally unavoidable). Any commander will thus feel pressured to force a battle before they run out of food and supplies (for an example of this in action, look at the Battle of Philippi, where Antony had to (twice!) force a battle because his logistics did not allow him to hold position, App. BCiv 4.16.122; Antony finds himself in this situation a lot, since he had the same problem prior to Actium).
And also why is everything on fire? It should go without saying that armies armed primarily with spears or bows (or even muskets and cannon) have few means of setting the battlefield alight. it is true that incendiary weapons – pots, arrows and javelins – existed, but these were almost exclusively for sieges or naval battles. Human targets – the primary sort of battlefield target – burn poorly and pre-gunpowder incendiary weapons were almost always impractical on an open battlefield (obligatory Greek fire caveat; we’ll talk about that another day).
The War to Film All Wars
So, if none of these tropes are appropriate to a pre-modern battlefield – and especially out of place in a pre-gunpowder battlefield – where are they coming from?
Flanders, it turns out.
(Obviously, I don’t mean just Flanders, but rather World War One generally, but since I am writing in English, and we are generally thinking about fiction produced in English (if not always by native English speakers), the experience of the BEF in Flanders tends to dominate the memory of the war.)
Everything about these depictions – even the scatter-pattern of fallen bodies – has its origins in the trench warfare of the Western Front in World War One (1914-1918, although I’d like to imagine I do not need to provide dates for that one). The mud and blasted moonscape is a product of high explosive shells. The field fortifications and extensive trench-works were a product of increasing firepower (although you can see a dim image of this horrible future in the union and confederate works around Richmond in 1864/5). The scattered fires are the product of modern munitions.
That particular image of war – armies in continuous proximity to each other, using lots of very heavy high explosive ordnance – has had a huge influence, particularly, on film. WWI was the first war to be extensively filmed (not that extensively, but still far more than ever before) and as film evolved as a visual form in the 1920s and early 1930s, WWI was still the war which informed the visual medium. The war was also documented heavily in photographs and literature; its influence is profound. This is particularly true in the fantasy genre; as I’ve noted, you simply cannot read the Siege of Gondor without recalling the Western Front (where Tolkien served).
(I should also note that I think these tropes are especially strong in video-games because they fit game mechanics which – for ludic reasons – need battles that are endlessly on-going, so that it doesn’t matter when the player engages with them. Armies sit, forever next to each other, waiting for you – the hero – to show up and decide the battle. That touches on one of the real limitations of video-games as an art-form when it comes to depicting conflict: the demand that the game be fun often means that games avoid using the real power of an interactive medium to really delve deeply into the effects of conflict (although there are exceptions, e.g. This War of Mine), compounded by the fact that greater immersion and player agency makes the potential impact of showing those effects much more intense – and thus less desirable for a developer who is looking, quite literally, to make war ‘fun.’)
Conclusion: What Ought We See?
So what should the post-battle battlefield for a pre-modern (or fantasy) conflict look like? Well, first off, it would be unusual for both armies (or indeed, either army) to be in the vicinity for very long. Multi-day battles are very uncommon in pre-gunpowder warfare (somewhat more common, I am to understand, in China, but still not very common). If the two armies are still there, fortified and setting up to Round 2, this should be treated as incredibly strange and indicative of a truly massive conflict (or perhaps a battle that – with one side having lost and retreated into a fortress – has become a siege).
Second, the field should be just that: still a field. Crops and grass might be trampled, but without heavy shells, the local terrain shouldn’t change much. Most armies aimed to fight in open, flat places with grass, not mud-filled bogs, so chances are the battle took place on farmland or pasture. If the land is agricultural, the time of year is going to determine what the fields look like; fallow or recently plowed fields might indeed be mud-pits if it rained recently, but even then the divisions between fields and other terrain features (like trees, roads, hedgerows) will still be there.
If things are on fire, it should be because someone has deliberately set them on fire, which normally means not a battle, but agricultural devastation – one of these armies is burning crops and villages. That might be an invader looking to deal economic damage, or a defender engaged in scorched earth defense.
Assuming gunpowder was not a major factor in the battle, the bodies will likely be concentrated into a line, or a series of lines, piled perhaps several deep, slicing lengthwise across the battlefield. This is the line of contact where the formations met (I am occasionally asked by students, “what happens to XYZ fighting system when they break into a confused melee” – TV style – and the answer is broadly, “they don’t do that”). It won’t be a wall of bodies, so much as a long thin pile, because bodies do not stack up like that.
Likely in one direction, the bodies should train away from the line of contact, as the losing force broke and fled, and some of its fighters were chased down. If cavalry was engaged, there are likely to be a lot of dead or dying horses. If arrow or other ranged fire was a significant factor, the bodies may be more spread out, but they won’t be randomly strewn, because large masses of soldiers don’t randomly move around the battlefield.
Finally, unless we are brought to this place very soon after the battle, chances are that the field has been looted. Broken equipment might be lying around, but just about anything of value of functionality – which I should note includes almost anything made of metal, even if it is broken – will either have been recovered by the victorious army or looted by the local population. You will note, in the pictures above of civil war battlefields, even before the bodies have been moved, there is not a musket, sabre, bayonet or ammunition pouch to be seen.
I actually think that such an accurate depiction could be truly visually powerful. The normal narrative purpose of these spaces is to impress on the viewer how terrible and destructive the conflict is. And certainly, the Western-Front-style moonscape does this in its way, but at the same time, it has become rote and familiar and so loses a lot of its impact outside of its own setting (that is, modern conflicts). Whereas the sharp juxtaposition between the savagery of war and the seemingly uncaring calm of nature provides an appropriate shock all its own.
Next week: We’re going to talk about making Greek or Roman style polytheism work for you.
Why haven’t we found the remains of Cannae, Gaugamela, etc.?
My intuition says that it’s because battlefields are really small relative to the Italian countryside; it’s like finding a needle in a haystack, once the haystack has been scattered over several acres and probably covered in a layer of dirt.
We have. We know exactly where Cannae was fought; there’s a monument and a museum there and everything. Not much in the way of active excavation IIRC but there’s not much I would expect we could find at the site; the weapons and cultures of the people involved and the overall tempo of the battle were well recorded.
It occurs to me that many of the stories that depict some of the historically inaccurate battlefield conditions you discuss do contain mass scale artillery akin to WW1 in the form of magic, dragon fire, various ahistorical alchemy weapons, and the like. Some of those trees and fields are burning because a dragon just flew over and lit the all on fire, and some of those mud filled craters were blasted by a wizard. It’s fun to think about how that might interact with the more historical armaments to change the nature of the battlefield.
I considered addressing this, but ended up taking it out for length.
The question is yields and tactics, for me. Basically, in most settings, a wizard’s fireball isn’t anywhere as destructive as a heavy caliber high explosive shell c. 1918. A really good run-down of the basics of First World War barrages is here: https://angrystaffofficer.com/2016/07/01/anatomy-of-a-world-war-i-artillery-barrage/
Let me put this into a bit of perspective by comparing the venerable Fireball to modern explosives:
So, a D&D 5.0 fireball (no meta-magic) has a lethal radius (if we assume nothing can just tank the 8d6 damage) of 20ft (6m). That’s about the same as the posted lethal radius of a m67 hand-grenade, which has as its filling 6.5oz of Composition B explosive. For comparison, a British 12′ howitzer shell – fairly standard heavy WWI artillery – drops with 83lbs, 3oz (37.96kg) of Amatol explosive as filling; Comp B is a bit higher energy than WWI Amatol, but not a huge amount.
As a result artillery shell probably has in excess of 150 times the explosive power of the grenade or the fireball (it has 200 times the raw mass of explosive filling; Amatol’s relative effectiveness is 1.1 to Comp B’s 1.33). A universe in which wizards or dragons can produce that kind of destruction is going to have tactics which look nothing like historical pre-modern tactics, because if you drop something like our 12′ shell into something like an Anglo-Saxon shield wall or a Roman legion, you’re going to get 100% casualties. The entire formation is just going to be gone, because that shell has a lethal radius the size of a football field (I’m estimating, I don’t have lethality figures handy, but it should be the right order of magnitude).
Even a modern 155mm shell has a lethal radius of 50m – 10 times the standard ‘wizard’s fireball’ and that’s a much smaller amount of explosive (around 7kg) than the heavy artillery that produces this sort of muddy-moonscape effect.
I keep debating if I should get into talking about how wildly inconsistent the destructiveness of some of these things can be (GOT is a big one here – if dragon-fire is hot enough to collapse castles and hits with enough force to shatter the walls of the Red Keep, then *everyone* in the Lannister line should have died from the heat and overpressure of Dany’s loot-train strafe), but for the most part, the implied magnitudes are just a lot too low to produce moonscape, and the heats aren’t high enough to cause things to hit their ignition points beyond the blast radius.
This is the type of hard hitting historical wizard’s fireball related analysis that really puts your blog on the map.
D&D’s fireball is smaller than artillery, but on the other hand you get other works discussing terrain being destroyed. From Malazon Book of the Fallen:
Raest spread his arms wide and unleashed his Warren. His flesh split as power flowed into him. His arms shed skin like ash. He both felt and heard hills crack all around him, the snapping of stone, the sundering of crags. To all sides the horizons blurred as dust curtained skyward.
From The Wheel of Time:
Beneath the feet of the Shaido nearest the wagons the ground suddenly erupted in fountains of flame and dirt, hurling men in every direction. While bodies still hung in the air, more gouts of flame roared from the ground, and more, in an expanding ring all the way around the wagons, pursuing the Shaido for fifty paces, a hundred, two hundred.
The latter seems to me to be intentionally trying for a magical version of 20th century artillery against unprotected soldiers.
As an aside, Malazan Book of the Fallen has a truly fascinating/nonsensical logistics situation in which one group fields an army hundreds of thousands strong that travels with no food whatsoever, primarily subsisting by eating their slain opponents as well as each other.
For Wheel of Time in particular, the book is often informed by the author’s experiences in Vietnam. Blasted moonscape is definitely a feature of that war.
One in particular that I find interesting is where the protagonist’s issues with killing women come from. The author apparently at one point killed a female Viet Cong soldier, and it kind of messed him up. He was pretty messed up in general by that war, though. Here’s a quote:
> I have, or used to have, a photo of a young man sitting on a log eating C-rations with a pair of chopsticks. There are three dead NVA laid out in a line just beside him. He didn’t kill them. He didn’t choose to sit there because of the bodies. It was just the most convenient place to sit. The bodies don’t bother him. He doesn’t care. They’re just part of the landscape. The young man is glancing at the camera, and you know in one look that you aren’t going to take this guy home to meet your parents. Back in the world, you wouldn’t want him in your neighborhood, because he is cold, cold, cold. I strangled that SOB, drove a stake through his heart, and buried him face down under a crossroad outside Saigon before coming home, because I knew that guy wasn’t made to survive in a civilian environment. I think he’s gone. All of him. I hope so.
I feel it’s probably more appropriate to compare a fireball to incendiary munitions like napalm rather than to, say, a hand grenade or a howitzer shell. It could certainly produce the charred grass and trees you often see in fictional battlefields.
It’s still the closest thing to artillery that exists in most fantasy settings. The only heavier “artillery” spells I can think of off the top of my head are in Mother of Learning, where they A. require multiple war-mages (or one dragon) to cast and B. wreck individual buildings rather than neighborhoods.
I mean D&D also has meteor swarm… but it is such a high level spell that it is a lot like the Paris guns in WWI: much too rare to have an effect on the tactics of the war overall.
Depends on how many high-level casters there are. D&D is never consistent even within its own settings.
> if dragon-fire is hot enough to collapse castles and hits with enough force to shatter the walls of the Red Keep
The collapse of the Red Keep’s walls through dragonfire would merit a few new movietropes.
I know I’m a little late to the party here, but something did occur to me here. In some fantasy settings – notably The Witcher and Dragonage, you might not get the moonscape effect because even profligate use of fire magic won’t churn the soil, but might have a profound effect on both the physical design of fixed fortifications and the prevalence of field fortifications. There are spells used fairly commonly on Thedas (Dragonage) that can turn a hefty chunk of a battlefield into a sustained firestorm, so things like trench-works with dugout shelters might end up being the equivalent of a roman marching camp just to prevent nighttime ambush by wizards dry-roasting everyone in your army.
There was an article in Dragon Magazine back in 2nd edition days about how castle and fortress construction might be different in a D&D world. Some of the features I remember:
1) Long sections of fortifications, e.g. ramparts, machicolations, hoardings, might be divided into sections of 20 ft or less, closed off by doors with small openings in them so lines of defending archers can still communicate and coordinate, but an area of effect spell like fireball or lightning bolt can’t take out all of them at once.
2) The tops of castle towers, instead of being open to the sky might be domed, to defend against flying creatures. In effect, the top of the tower is like a modern pill box, with an opening to allow archers or ballista-type engines to operate without being bombarded from above.
3) The interior doors and doorways of castles might be lined with fur-and-felt, creating an airtight seal that keeps out creatures in gaseous form.
4) (Not from that Dragon article, but the 1st ed. Manual of the Planes) Mortar could be mixed with gorgon’s blood in some alchemical process, creating stonework that cannot be crossed by beings in the Astral or Ethereal Planes.
The flame spells wouldn’t create moonscapes regardless of kill radius- they are incendiary not explosive after all. They will however ignite fires that will likely survive after the battle.
“Likely in one direction, the bodies should train away from the line of contact, as the losing force broke and fled, and some of its fighters were chased down.”
There’s a fantastic mid-16th century sketch map on the aftermath of the Battle of Pinkie Cleugh drawn by William Patten, who fought in the battle: https://archive.org/details/tudortracts00polliala/page/119. You can see where the majority of men tried to flee around the head of a small stream beyond the Esk, while others fled to the ocean or tried to get across the bridge. Judging by the position of the dead, some of those who fled towards the bridge later attempted to flee back towards the headwaters of the stream after the bridge became too crowded or were killed heading towards the bridge to avoid the cavalry pursuing the majority of the Scots. After the stream, the split in destinations becomes more obvious as those who crossed the bridge head exclusively for Leith while those who went around the stream headed for either Edinburgh or the Holy Rood Abbey.
The map *is* somewhat impressionistic, since Patten notes that most fled towards the marsh around Dalkeith, but it’s nonetheless a great demonstration of what a defeated army would look like. Patten’s description of the pursuit is also something any budding author or director should read, since it describes the varied actions of the fleeing men and also how they occasionally attempted to fight back against the pursuit: https://archive.org/details/tudortracts00polliala/page/124.
A few questions:
You mentioned that multi-day field battles are rare in pre-gunpowder western history, which seems to imply that they are not completely unheard of. What are some examples of these battles?
How often was offensive or defensive economic warfare practiced? Was it more common in classical or medieval times? Especially in the context of feudalism scorched-earth warfare seems like it would be a last resort.
I’ve heard it claimed both in historical fiction and in history that casualties during shield-wall clashes were extremely low, and almost all the damage was dealt after one side broke. Piles of bodies at the point of contact seems to suggest otherwise. To what degree is it true that most cultivates came during a rout?
My last question is less related to the article. Did field battles almost always involve infantry combat? In the High Medieval period, for example, it’s my understanding that the trained soldiers were all or almost all mounted. Were there ever large clashes between two groups of cavalry, and if so what did this look like?
Ok, some quick answers:
1) Multi-day battles. Extremely rare. Thermopylae runs over three days (due to limited frontage), while the Battle of Phillipi is fought over two non-consecutive days, because the armies were unusually large (such that the first day was essentially two loosely connected battles, with each side winning one of them). Those are the ones that come to mind, I’m sure there are others; in some cases the issue is drawing the line between a battle and an active siege.
2) Some level of economic warfare is essentially a part of basically all warfare, everywhere. The line between foraging and pillaging is a matter of perspective, so pretty much all armies engage in some kind of economic warfare merely to eat.
Medieval warfare was often very economically focused – in a context where getting an enemy out of a castle or fortified town, it was often better to focus on disrupting agricultural revenue by raiding farmers and to force a favorable solution that way. The fact is, as I’ve discussed elsewhere, medieval armies were too small to do permanent economic damage most of the time – but they could interrupt the revenues that the land produced much more easily.
3) Casualties in clashes of shield-walls were low as a proportion of total army size (the usual guesstimate is c. 10% for the loser, 2-5% for the winner), but there’s a lot of variance in there. Roman legions were, for instance, a lot more lethal at contact than the hoplite phalanx (or so our sources say). But the issue with bodies is with geographic distribution, not total magnitude. If a 10,000 man army loses 10% of its strength, but all of those loses occur in a space 10ft wide and 1 mile long (roughly along the line of contact, plus men run down in the first few meters of flight), you are going to have a whole lot of bodies in a very small space. See Keegan, Face of Battle (1976) for more on this point.
4) Always? No historian ever says always. But even in the high medieval period, there’s a lot of infantry on the field. Our sources often essentially ignore them in favor of what the military aristocrats are doing (because the military aristocrats (or their cleric family members) writing the sources want to talk about themselves), but they are there.
There are large clashes between cavalry. Again, The Face of Battle is a good starting point on discussions of how the mechanics of that work out. Ibn al-Athir’s account of the Battle of Hattin might also be interesting to you on this point. I may talk more about cavalry-on-cavalry dynamics in the future.
I’d suggest against Keegan for the mechanics of pre-modern combat. He’s got some weird ideas, like the dismounted French men-at-arms essentially couching their lances and charging the English or that horses would never charge into an infantry formation, that distort the picture. I’d like to suggest Clifford J Rogers’ “Soldiers’ Lives Through History: The Middle Ages”, but it’s rather expensive and not the most common book in libraries. J.F. Verbruggen is otherwise probably the best English language author with regards to cavalry and infantry combat, and he can be supplemented by Thucydides 6.64-6.71.2 (which is not only a textbook battle but which also demonstrates the small percentage of men killed during actual shield wall fighting) for the infantry. Ardant du Picq is also worth a read, given that his model of pre-modern warfare has essentially become the standard in spite of Keegan’s best efforts.
Keegan absolutely has his limits (not the least of which is the fact that he really ought to have credited du Picq more clearly). Here I was suggesting it for two specific points: cavalry on cavalry clashes and the distribution of bodies on the field. While he has his problems elsewhere – you note several – on those points, I think he is still useful to the beginner lay-reader.
As for cavalry ‘impacts’ against infantry formations, I feel its important to note that the question is still actively debated by specialists and Keegan’s view that most horses – not all, perhaps, but most – will not rush onto a hedge of spears, that view has its defenders. For me, it’s always seemed a touch beside the point: you may be able to get a horse to charge into a solid block of infantry, but if they don’t break, you wouldn’t want to – it won’t end well for you or for the horse. Which just gets back to the core question of cohesion and morale (which in turn circles back to Ardant du Picq, so there’s that).
After giving the relevant chapters by Keegan a reread (it’s been a while since I read him cover to cover), I seem to have only remembered part of his discussion on cavalry vs infantry. While he does say that horses will not charge a body of men, he also points out that they can be trained to and, if the rider has the will, absolutely will. His cavalry vs cavalry observations, while at odds with my own understanding a few years ago, also seem to me now entirely reasonable and explain a lot about how medieval cavalry could be both described as packed so tight that the wind couldn’t blow between their lances and yet still manage to use their lances in combat.
I remain, however, unimpressed with his infantry vs infantry section for Agincourt. While there is the occasional good insight (especially regarding the importance of having room behind you in a fight, something that both movie makers and academics frequently ignore), overall it’s much inferior and explains less of the infantry vs infantry dynamic than Ardant du Picq does.
Regarding the role of the rider in the cavalry vs infantry fight, I completely agree that the will of the rider is more important than whether or not the horse wants to charge home. It’s a point that most 19th century cavalry officers make time and time again, and I don’t often see it brought up in discussions of whether or not a horse will charge home. In this too Keegan seems to be in agreement, but a lot of people, like me, seem to have read/remembered only one half of his paragraph on horses being unwilling to charge infantry.
“But the issue with bodies is with geographic distribution, not total magnitude. If a 10,000 man army loses 10% of its strength, but all of those loses occur in a space 10ft wide and 1 mile long (roughly along the line of contact, plus men run down in the first few meters of flight), you are going to have a whole lot of bodies in a very small space.”
I wouldn’t say that all or even most of the casualties would be along the line of contact. Mostly as long as both armies kept their formations, the casualties were not that high. Once one side broke and its soldiers began to run and scatter the light infantry and cavalry of the winning side started the pursuit and more or less one-sided killing of the fleeing enemies. Such pursuit could often last for several hours and extend on the area of several miles. After the battle with Ariovistus Caesar’s legionaries chased the fleeing Germans all the way to the Rhine (ca. 50 miles; Caes. B.G. I.53), after the battle of Idistaviso the Romans pursued their enemies until nightfall and over some 10 miles (Tac. Ann. II.18), Antony’s cavalry once chased the Parthians over ca. 150 stadia (Plut. Ant. 39 – but this was, because they were pursuing cavalry and even then were able to inflict only minimal casualties). See also the pursuit of the losing side after Pharsalus (Caes. B.C. III.95-97) or Philippi (App. B.C. IV.129) to see other examples of what extent such operations could have.
So yes, there would be many corpses along the line of contact, but then probably also many corpses scattered on the loser’s side of the battlefield and then probably many smaller groups of corpses all over the area in the larger vicinity (several miles) of the battlefield.
Indeed, you will note that I say exactly this:
Likely in one direction, the bodies should train away from the line of contact, as the losing force broke and fled, and some of its fighters were chased down. If cavalry was engaged, there are likely to be a lot of dead or dying horses. If arrow or other ranged fire was a significant factor, the bodies may be more spread out, but they won’t be randomly strewn, because large masses of soldiers don’t randomly move around the battlefield.
“You mentioned that multi-day field battles are rare in pre-gunpowder western history, which seems to imply that they are not completely unheard of. What are some examples of these battles?”
As well as the ones listed, Bannockburn happened over two days; Day 1, 23 June 1314, saw initial clashes between the English cavalry and the Scots, after which the English withdrew into the Carse and made camp; the battle was renewed on Day 2 with an attack by Scots infantry on the English main body.
This may be one of the times in which Rome: Total War is accurate!
They don’t change terrain features as a result of battles, probably because it’s a game from 2004.
The dynamics of the battle, given that soldiers fall dead where they are killed, determines the patterns of casualties, so you will see clusters of dead where a formation fought, and then bodies trailing away as the formation broke.
They don’t model wounded, for obvious reasons.
Reference image: https://steamcommunity.com/sharedfiles/filedetails/?id=973514883
I came to the comments to say this. People have commented for a long time that you can read the battlefield by the corpse distribution in a Total War game. It’s like a fingerprint – an easy rout, a brutal infantry slugfest, an archery duel, an ambush, and a force that got flanked and surrounded all look totally different. And while thre games aren’t always realistic, this part seems to be.
Just one note on the Witcher (3, but 2 as well): You only ever come to the battlefield after the fight is over. There is a large Nilfgaardian camp to the Southeast. In in-game geography, it’s probably a couple days worth of travel away.
I’d be interesting if you could do an analysis of the Nilfgaardian campaigns in the Witcher novels, though there might simply not be enough information there.
Another fantastic example that supports many of your observations – third battle of Panipat
http://www.bl.uk/onlinegallery/onlineex/apac/other/largeimage55447.html
Belated proofreading corrections for this post:
history will not that some battles -> history will note that some battles
note that words -> note those words
sitting in starring matches for months -> sitting in staring matches for months
battles that are endlessly on-going -> battles that are endlessly ongoing
bodies should train away from -> bodies should trail away from
I know this is later but I’ve been reading through this blog and its great. I wanted to mention about the Velen battlefields from Witcher 3. Velen itself is mostly a bog on the banks of the major river delta dividing two countries. Both Redania and Nilfgaard recognize that it is an awful place to fight specifically because it is so wet and muddy. The only reason they are fighting there is because the two bridges in the area cross the river into Southern Redania, so the Redanians have established a foothold in Velen itself and Nilfgaard is investing in the region to prepare for a push over the river. Nobody wants to be there but they’re stuck there until the battlefront of the eastern campaign resolves which could either open Redania to attack from the east or put Nilfgaard on the defensive, depending on who wins the fight in neighboring Kaedwen.
I think it’s worth noting that this includes most (though obviously not all) fantasy mages.
Take D&D as an example, since the limits of its spells are well-established (and its magic is towards the high-power end of the fantasy spectrum, especially at high levels). Fireball is possibly the spell closest to modern artillery, along with derivatives like delayed blast fireball and meteor swarm.
The radius of their area of effect is just 20 feet, while artillery shells can reliably kill people 150 feet from their point of detonation, and injure people a football field away. Not only that, but their “explosion” only heats stuff up; it’ll set dry ground cover on fire, but it doesn’t make craters (which you’d think would at least qualify as difficult terrain).
And, of course, the range of a fireball is 150 feet in 5e and 600-1200 feet in 3.5 (depending on caster level), but artillery can reach several miles. You don’t even need line of effect; parabola of effect is more than sufficient.
There are works where attack magic causes damage comparable to artillery, but it’s rare.
The only example I can think of off the top of my head is Mother of Learning, but even then the (aptly-named) artillery spells, which require several mages working together, seem closer to a mortar than a towed gun; in the big battle from chapter 4, they’re used within a major city and destroy specific buildings without wrecking the entire neighborhood. Moreover, a dragon-archmage using them in a similar battle near the end of the series doesn’t do any better; proper artillery strikes would apparently drain too much energy for even the most powerful magicians in the setting to use them.
I’ve found the description “windrows” in US Civil War commentaries on battlefields, meaning that the dead are in line and often facing in the same direction. The original meaning comes from stalks of wheat or rye blown down by a high wind.
As an exception to the linear form, Keegan also mentions the pattern left by the English battalions on the ridge at Waterloo, which stayed in square under heavy attack by cavalry and horse artillery in sequential phases (they were horribly vulnerable to even light cannon while in square but couldn’t form line because of the cavalry threat) – thick clumps of dead and dying men were left when the battalions moved on.
I’ll note that burying the dead was considered a Christian duty after any battle, and it’s what the victors generally did the day after (the commoner infantry, at least, the warrior aristocrats not so much). I’d expect a medieval battlefield to be cleaned up the day after it is fought, of human bodies at least, although there may well be fresh mass graves.
Religious impetus was one thing; another was that cleaning up the bodies also let the victors strip them of anything valuable. Plus, bodies decompose rapidly even in the mild summers of the Little Ice Age, so if you want the cleanup to be profitable (and the commoner infantry certainly do), it needs to be done right away. Note that at Visby, the victors actually pursued the defeated, and didn’t get back to the original field until the bodies were nice and ripe. Which is why Visby is one of the few battles we can locate with a high degree of confidence; they dug mass graves and chucked the dead in, armour and all.
I have a question about the mud thing – from my experience at music festivals and amateur local football matches, it doesn’t necessarily take that much rain, or that long, before the ground becomes very muddy indeed. Not like WW1 levels of mud, but still slippery and difficult to traverse. So if pre-industrial armies rarely fought in these conditions, was there some other factor?
Where I live (a place that is not especially rainy or arid) it turns out that less than ~15% of the days see any rain or snow at all (and lots of the days that did saw like 0.1mm of rain total). This, combined with campaign season being a season that has less than average rainfall, mean you can already expect a random battle day to have only a 10% chance of rain. And if rain is expected, you can also wait a day or two for it to dry again before you have to fight in the rain, so you get an even lower chance.
Thus it makes sense that it is not especially prepared for. Sure, experienced fighters might have done it a few times before, but infantry levies for example probably haven’t done enough battles yet to have seen another rainy battle.
This got me thinking about Tolkien – especially the big first age battles in the Silmarillion more than what we see in the Lord of the Rings: Morgoth is described as using his own personal volcanoes, dragons who are hundreds of feet long and radiate heat such that grass burns away if they’re just standing near it (so I don’t want to know what their fire was like, and also poisoning streams and rivers. (These forces are also described as being hopelessly overwhelming to medieval-style defenses.) Now, Tolkien was obviously, consciously influenced by his own experiences in World War I, but he was also a medievalist (even though his expertise was in literature and languages rather than military history), so probably knew something of how medieval battles worked? Because at least in the Silmarillion, it seems these things would give reason to imagine the battlefields as completely destroyed moonscapes – the Dagor Bragollach with its artificial volcanic eruptions is described as turning a green plain into a lifeless dustbowl in no time, and the Nirnaeth Arnoediad takes place on this same lain (so it already looks like that before that battle). (I’m less certain about the battles in the Lord of the Rings – Sauron didn’t have dragons and balrogs under his control after all, and his personal volcanos were in Mordor itself, which means the battlefield would have been a blasted moonscape at the siege of Barad-dûr in the War of the Last Alliance, but not at the battle of Pelennor, for example.) Not sure where I’m going with this, this is just what I kept thinking about when I was reading this.
I suppose that’s why at Gettysburg they called it “The Peach Orchard” instead of “The Nothing.”
First Witcher pic – that particular battlefield is explicitly full of corpse-eating monsters attacking anyone who comes near. Same deal for the second Witcher pic, though that one DOES have a scavenger team, who task Geralt with driving off the monsters.
Also – yeah, the scabbard mod for TW3 really helps with immersion.
A few years late, but I just had the following thought: You compare a pre-industrial army to a shark, in that it must keep moving to live, but a better comparison to me seems to be a swarm of locusts: Creatures that can support themselves from the land normally, but then concentrate to a density that cannot be sustained without continuous movement while leaving a wake of destruction.