This is the third part of of a six part series taking a military historian’s look at the Siege of Gondor in Peter Jackson’s adaptation of Return of the King. Part One and Part Two can be found here and here. Last time, we discussed Gondor’s execution of a strategy of defense in depth, designed to delay and weaken the advance of the army of Mordor. This time, we’re looking at the assault on the city itself: how accurate and plausible a portrayal of pre-gunpowder assault techniques is this sequence? And what is the historical basis of this siege sequence itself?
For the sake of clarity, we’re going to break this sequence down by the type of assault and response, rather than strictly by scene order (because there is a lot going on at once here). In order, we’re going to look at the approach, catapults, then siege towers, and then finally the battering rams (I won’t spend much time on the aerial ringwraith assault – it will not surprise you to learn that there are no historical precedents for being attacked by immortal wraiths riding flying reptiles). The book-notes in these sections will be brief, because I want to discuss the book’s take on the siege in a separate, final section. So, without further ado:
The Approach (or, the Pelennor Steppe)
The orc army arrives, formed up outside of Minas Tirith. There are a few oddities with this. The first is the terrain: the army is able to form up like this because they appear to be forming up for battle in a large, relatively featureless plain of grass. I have a full post on this trope of cities placed in empty plains brewing, but for now, it’s worth noting that this is not what the land outside of a large pre-modern city would look like. There is a river nearby and this is valuable farmland which could be irrigated and tilled. The countryside outside of Minas Tirith should be broken up by dozens of villages, hundreds of small farms and even the occasional small town (there should probably also be trees somewhere for firewood).
Pre-modern societies do not generally build cities out on Steppe grasslands (…unless you are the Mongols – Chinggis Khan built his capital, Karakorum out on the Steppe). Long distance transport before the industrial revolution is extremely expensive (land transport more so than river or sea). so building a large city somewhere where most of the food and water it consumes would have to be imported at long distance is simply not something pre-modern societies can generally do.
Book Note: These terrain problems do not exist in the books. Tolkien describes the Pelennor thus, “the townlands were rich, with wide tilth and many orchards and homesteads there were with oast and garner, fold and byre, and many rills rippling through the green from the highlands down to the Anduin.” (Rotk 23). The evacuation of these people (along with their food stores) is witnessed by Pippin (Rotk 38). This is a smart move by Gondor – it limits the non-combatants in the city who need to be fed or defended and at the same time denies an approaching army local supply. All those scenes in the film of cowering women and children are film inventions, since – again – Denethor is smarter in the book.
You may readily imagine that an army navigating built-up farmland will look very different from the neat formations of the army of Mordor here. Units will have to follow roads, or else navigate hedgerows and fences dividing fields. Villages, small copses of trees and the like will disrupt formations. And we certainly expect an army like this to loot, pillage and burn as they go. I think in this case, Jackson missed an opportunity to show a real human cost to this war: even if the assault on the city had failed at its outset, simply moving an army over the Pelennor like this would have caused tremendous economic and human damage.
The second oddity is the formation. Forming an army up in neat rectangles like this is difficult. It takes time, planning and effort. Some orc had to sit down, calculate the size of each unit and then tetris them all together. Which is strange, because I am not sure exactly what this formation was intended to accomplish. It doesn’t completely envelop the city, so it serves to advertise very clearly the intended point of contact (something most armies would want to conceal for as long as possible). And apart from the orcs in the front, all of these troops are formed up in range of enemy weapons with nothing to do.
This pre-assault period should actually be very busy. The paths the siege towers will take must be cleared and leveled (those towers have very little clearance and even a slight grade will tip them over – they need a path made for them). Earthwork cover for the approach on the gate should be set up, along with obstructions to prevent the army within the city from advancing out of it at an inopportune moment. In assaulting a fortified city with a large army, the spade is often the most important weapon. Even simply building a ramp right up the side of the enemy walls to enter the city was a common and successful tactic, if the assaulting army had enough labor to do it quickly enough.
Book Note: These issues are avoided in the book. We are directly told the orc army engages the city wall at all points (RotK 104-5, 111) and that many of the orcs are engaged in digging earthworks or setting up siege machines (RotK 105). The goal is to spread out Gondor’s defensive forces, weakening resistance at the gate, where the main blow (via Grond) was always going to fall.
Nor are these formations effective battle formations. Some of the lines look to be dozens of ranks deep and densely packed. That both prevents these blocks from moving around and through each other (a key component, for instance, of Roman battlefield tactics) or of these men moving on their own. If an open battle breaks out, only a small portion of this army can fight effectively – most of the orcs will be trapped with buddies in front and behind (of course lines of melee infantry were often quite deep, but not this deep – the standard depth for Romans was 6, Greek hoplites 8, Macedonian pikemen 16). This is simply not a good way to organize an army for a siege or a battle – and it’s also a difficult way to organize an army, so you are not likely to have done it by accident either.
But this style of assembly does have a historical precedent – just not a military one. Jackson is mimicking a very famous sequence from Triumph des Willens (Triumph of the Will), a Nazi propaganda film. The scene, which shows the Nazis gathered for a political rally, was calculated to impress on the viewer the great and united strength of the Nazi party (paradoxically, Hitler’s party was, at that moment, fairly weak and divided – this is why he wanted the propaganda film in the first place – check out this brief treatment of the film’s deception and continuing relevance). Since then, this has become a standard visual trope for ‘large, powerful army of bad guys’ (and sometimes very awkwardly for assembled ‘good guys’ as in the medal-ceremony at the end of Star Wars: A New Hope) even though it isn’t a military formation per se (and in fact, the Nazis doing it in Triumph were not soldiers of the German army, the Wehrmacht, but members of the party – very few actual soldiers appear in Triumph of the Will).
The next phase is an exchange of missiles by way of catapults. Catapults in film and video games are an often misunderstood and misused weapon. Historically, there were different kinds of catapults for different purposes, with some designed to batter enemy fortifications, while others were designed to fire over fortifications to strike the inside of a city or castle. Except for Roman double-arm torsion artillery (don’t worry, I’ll explain those terms in a moment), catapults were almost never used as field artillery: catapults targeted castles and cities or else engaged in ‘counter-battery’ fire from castles and cities. Mercifully, Jackson never shows catapults used in field battles, only in sieges.
So what kind of catapults are these? Let’s start with the easiest one to answer: Gondor is using very standard designed late-medieval counterweight trebuchets. These are designed like a giant lever, with the firing energy stored by lifting (with muscle power) a heavy counterweight (typically a box filled with stone) into the air, with the projectile placed on the far end of the other arm. When released, the counterweight falls, causing the firing arm to fling upwards, sending the projectile down-range. The setup is great for range, but the high firing arc (which we see in the film) reduces accuracy.
In terms of design, we get a few really wonderful (but brief and hard to screencap) shots of the catapult mechanisms working and it looks to me like they built one and filmed it in actual operation, which is fantastic. The counterweight box is fixed on a hinge so that it can drop straight down (rather than in an arc), to maximize acceleration, a standard medieval design feature. The sling attacked to the firing arm is also historical – it lengthens the sweep of the projectile (giving it more time to accelerate, absorbing kinetic energy from the firing process) and also gives the crew an easy way to adjust the firing arc.
Having catapults fire from the walls in a counter-battery function (trying to destroy enemy siege works or their own artillery) is historical. Catapults (in this case, torsion, rather than counterweight) were first incorporated into city defenses in the Hellenistic period (323 BC – c. 31 BC), waxing and waning with use over time depending on the availability of the technical know-how to build and operate them.
What is strange here is the mounting of the trebuchets, which are placed up on high towers on top of the walls. First, this is unnecessary – the high firing arc of a trebuchet, combined with the very long firing arm should allow them to be fired from the safety of a position behind the walls. The second issue is more conjectural – I have been told, but have never seen in print – that the firing action of a large trebuchet (particularly the force on the base of the machine, which will tend to ‘rock’ if not well secured) will tend to damage a stone tower. That said, there is artwork (see above) that depicts trebuchets firing from inside castles, so the positioning is not entirely crazy.
What is crazy is the size of the projectiles they are throwing. Because a trebuchet is a giant lever powered by gravity, it can never fire anything heavier than its own counter-weight. But the massive chunks of masonry being show are huge (see below) – that block alone must be 20′ by 20′ and at least 6′ thick (2,400 cubic feet). Ignoring how the masonry kept it together through the force of firing, such a block of limestone would weigh around 150 tons. For comparison, Warwolf, an uncommonly large and expensive largest trebuchet built in Europe (built by King Edward I of England in 1303) hurled a 140kg (300lbs) projectile. Heavier projectiles did exist, but the heaviest report I am aware of is 240kg (560lbs).
A broader question is why – having set up these trebuchets at such great expense – Gondor did not apparently provide for any ammunition. Large irregular blocks of rubble have terrible aerodynamic characteristics, which is going to result in inaccuracy and excess drag, further reducing range. While medieval catapult ammunition was often not well standardized (by contrast, Hellenistic and Roman catapult ammunition was of precise caliber and weight – ammo depots have been recovered archaeologically, with the stone balls, neatly organized by weight and caliber, still there), they could manage to carve a round ball of roughly predictable size and weight.
A final oddity is the range they are being fired at. On the upside, the trebuchets are never shown striking too deep into the orc army, but on the other hand, the orcs do seem pretty far away and many of the trebuchets are deep inside the city. In practice, reconstructions of large trebuchets made on medieval model tend to cap out around 400m or so of range depending on the weight of projectile thrown. Also, the Gondor trebuchets should soundly outrange the orcish ones. Not only are they (as we’ll see) much better designed, they can also throw much lighter projectiles, because they’re being used in counter-battery and anti-personnel functions, rather than trying to bash apart stone buildings.
The design of the orc catapults, on the other hand…oof. This is not a great design. On first viewing, I thought these were torsion catapults (like the Roman onager – a late Roman single-armed torsion siege engine), because in the wide shots where the catapult is presumably pure CGI, the firing arm snaps very quickly forward when fired. Counterweight catapults do not ‘snap’ like this, because the counterweight can only accelerate as fast as the constant acceleration due to gravity (9.76 m/s). Nevertheless, in the close shots, it is clear that these are counterweight catapults, with the large stone counterweights clearly visible on the far end of the arm. So what’s wrong?
First, let’s get the nit-picky stuff out of the way. The ratio of the firing arm to the weight arm is off. Historical trebuchets have high ratios – from 1:3 to 1:5 – because the longer the firing arm and the shorter the counterweight arm, the more leverage the weapon can apply. Here, the firing arm seems quite short. The counterweight itself is odd – I have seen historical drawings of siege engines with such a simple counterweight, but not often. More typically, the counterweight is a box attached to the arm by an axle so it can rotate freely as the weapon is fired. The axle allows the counterweight to fall straight down as it rotates relative to the arm, which maximizes the speed of the lever’s rotation by allowing the weight to fall straight downward – fixed counterweight trebuchets are substantially less efficient.
There’s also no sling and instead they use a simple metal bucket. This isn’t impossible, per se, but it is far less efficient than the slings you see Gondor using. The sling allows for more kinetic energy to be imparted to the projectile by allowing it to accelerate over a longer space. it also allows the operators to alter the intended trajectory on the fly by adjusting the length of the sling.
A much bigger problem is the stopper-bar on the top. While torsion catapults, like the Roman onager often feature such stop bars, trebuchets generally do not. The reason is simple: they don’t need one. Once the missile is discharged (at the top of the arc), almost all of the weight on the lever is at the bottom (where the counterweight is), so gravity does the work of bringing the lever to a stop (much like a pendulum slowing down). The force of the firing arm slamming into the stop-bar is going to be considerable, and that limits how big a catapult you can make before you have constructed a self-breaking catapult. Moreover, rather than letting gravity, friction and the missile take all of the energy of firing, a stop bar ensures that this energy will be translated through the A-frames into the…wheels.
At least in Europe, trebuchets appear to have never had wheels (at least, I have never seen a period illustration of a wheeled trebuchet; I have seen images of wheeled trebuchets from China). Instead, they were carried to the siege disassembled and then reconstructed on site (Warwolf, for instance, occupied from 40 wagons on its way to Sterling Castle). Wheels are a pretty big liability – a fair chunk of the rotational energy from firing ends up in the frame as the counterweight’s kinetic energy transitions from vertical (falling) to horizontal (having now reached the bottom, the arm begins pulling it towards the back of the frame). Consequently, trebuchets are generally built with relatively long, durable and fixed frames, without wheels, which can keep the platform stable during firing. It’s hard to tell (the scenes go by fast), but I don’t see any effort to really fix these weapons into the ground. The ‘kick’ from such a weapon could be dangerous – it was how the Roman onager got it’s name – it kicked like a wild ass (an onager in Latin) when firing as the arm hit the stop-bar.
Ok, so these aren’t good catapults. But are they doing appropriate catapult things? Largely, yes. We see them fire three kinds of projectiles – solid stone, severed heads and rocks covered in burning material (probably pitch). Those are all perfectly historical projectiles. Critically, they are not fired at the walls but over them. Trebuchets could be used to collapse castle walls (the aforementioned Warwolf knocked a hole in the curtain wall of Sterling Castle after repeated firings), but they were also often used to strike within a castle. Late period castles featured walls robust enough to render them almost immune to trebuchets (Tolkien tells us explicitly that the outer wall of Minas Tirith was so immune), but striking at more vulnerable towers, gate-houses and especially vulnerable buildings inside a castle remained a valuable tactic.
Demoralization attacks (using dead bodies or rotting animals) were common, as were attempts to set wood-and-thatch buildings within the city or castle on fire (note: using such munitions in a field battle is nonsense – looking at you Gladiator). Also note the attempt to de-crenelate the wall. Crenelation refers to zig-zag pattern stonework on the tops of walls to provide cover for defenders. Even when a wall might resist catapult fire, the crenelation was much more fragile. Roman armies often used catapults (particularly their more accurate torsion catapults) this way – de-crenelating the wall and essentially suppressing the defenders (usually while the Romans built a ramp right up the wall). That said, the orcish catapults have no way to control range or firing arc – I have a hard time believing they would be accurate enough to target anything smaller than the entire city.
Book Note: Tolkien stresses the moral effects of the battle much more strongly than Jackson does. The missile exchange achieves two key things: it furthers the demoralization of the defenders, and it sets parts of the lower city ablaze, drawing men away from the walls and eventually cutting off portions of the outer defenses (RotK, 107-8)
Finally, I should note, there are a lot of catapults here. For a medieval setting, that would be inappropriate – medieval sieges rarely involved more than a handful of siege machines of any type. But whereas the battle between Rohan and Sarumon was fought on a roughly medieval scale (a few thousand against roughly ten thousand), this battle is fought like a great clash of empires from the ancient world (which, in fact, it is). The Macedonians or Romans often brought large numbers of catapults to sieges – Vegetius notes that a legion ought to bring 65 catapults on campaign (55 ballista and 10 onagers; Veg. 2.25), which would have been in a range of sizes and calibers for a variety of purposes.
As the missile exchange proceeds, the orc army advances a number of siege towers to assault the walls. Siege towers were a standard part of the siege assault toolkit of the ancient and medieval world, but while they absolutely could deliver soldiers to the walls, this was often a secondary purpose for the tower. The major distinction here has to do with how much labor is available for an attacking army – the siege techniques available to an attacking army change radically as the size and organization of the army improves. And that’s a topic worth talking about.
For ancient siege armies, like the Macedonians, Persians, Assyrians or Romans – who had very large, well organized armies – the purpose of siege towers was not so much to deliver an assault on the walls as to give an elevated firing position for archers and (torsion catapult) artillery to fire down on the defenders to suppress (as in ‘suppression fire’) defenders on the walls. The primary assault was more likely to be delivered by sapping, or earthwork ramp (you can see the latter on the Lachish relief image near the top of the post – and indeed on the far left is a siege tower being rolled up the ramp to offer missile support. It has no ‘getting over the wall’ function, but is a pure firing platform).
For this kind of siege-craft, army size matters, but so does organization and technical know-how. While building an earthwork ramp sounds relatively simple, doing so with any speed turns out to be quite complex. Work crews must be coordinated and supervised, the ramp itself must be engineered and planned. Some men must be working (likely out of armor and without weapons) and other men must be guarding the works, while still other men man siege engines to suppress defenders. That kind of organization and engineering is something that, generally speaking, professional or semi-professional troops excel at, but other kinds of armies (medieval retinue-of-retinue armies, tribal levies) tend to struggle with.
In the Middle Ages, few armies could muster this kind of mass-organized labor to move the many tons of earth required to build ramps straight up the enemy fortress. Consequently, towers became access points (as with the two siege towers at the Siege of Jerusalem, 1099) – because a small group of men, given time, could at least build a wooden tower or two. Even if the tower was primarily for delivering troops to the walls, being able to suppress defenders was important: despite what we see, men charging over ramps or bridges from the tower to the wall would have been extremely vulnerable and all sorts of things could damage the tower as well.
That the orc army is thus using the towers exclusively to access the wall is strange. This army is clearly very large and very well organized (remember the nice neat blocks of men?). In the film, this is especially strange since these towers seem to have been constructed at great expense and are a key part of the assault (whereas in the books they are explicitly only a distraction). I think this problem arises because the organization and professionalism of the orcs – shown visually but never in dialogue – is actually a product of Jackson borrowing film tropes from modern propaganda films, rather than an intentional part of the story. Nevertheless, an additional level above the ramp, filled with archers, would greatly improve the effectiveness of these towers.
So what about the design of the towers? The underlying wood frame looks to be fairly reasonable (although I struggle to imagine it supporting the weight). What is less reasonable is the layered iron cladding over the front of the towers. At first I thought this might be leather (wet leather was used to face siege towers to resist fire), but it is clearly iron. They hold their shape when towers are smashed by trebuchet fire. You can see some of the plates are rust colored, and when the front ramp comes down, you get a clear look at what is intended to be iron plates. The plates are thick too – at least an inch or so, some appear thicker.
Now the concept of a metal-plated siege tower is not, in and of itself, absurd. Famously, Demetrius Poliorcetes, in his failed siege of Rhodes (305 BC) employed a metal plated siege tower called the Helepolis (which later become a generic term in Greek for siege tower). This tower was of the ‘suppression’ variety, loaded with multiple levels of catapults (both stone and arrow-throwing varieties), but nevertheless, to resist fire and counter-battery fire, it was metal plated. The tower was truly massive and required a crew of 200 just to man the catapults.
The issue here is the thickness of the metal plate. These towers look to be about 60 feet tall, and perhaps 20 feet wide (trying to estimate this conservatively – but it is very hard to do from film). In that case, just covering the front (not including the wrap around on the sides) would require 75 cubic feet of iron, which doesn’t sound too bad until you remember that iron is very dense – 75ft cubed of iron is 26,960kg (59,436lbs; 29.7 tons); we may safely double this for the wrap-around on the sides giving a total iron weight of c. 53,920kg (118,873lbs; 59.4 tons). Now the Helepolis, we are told by Vitruvius (a Roman engineer) weighed more – some 360,000 Roman pounds (270k lbs modern; Vitr. De Arch 10.16.4), but the Helepolis was much larger (45m high and 22m wide- note that is meters, so that’s 147ft high and 72ft wide) with crucially a much wider wheel-base than these towers have. Even then, Vitruvius notes that it was defeated precisely because its great weight made it unable to navigate wet terrain (Vitr. De Arch 10.16.7). These iron-clad towers thus have something like 44% of the weight of Helepolis in iron alone (we haven’t factored in the wood, or orcs), but with only about 17% of the wheelbase (around 900sq ft (30ft by 30ft) compared to 5,184sq ft (72ft x 72ft). This bodes poorly for stability in motion or the ability to move over unimproved terrain. At the very least, these things need a road built for them.
The other problem with the thickness of these plates is their production. Even in the early days of the industrial revolution, producing iron plates of this kind of thickness was a difficult task. Casting was not an option – cast iron is brittle – so they would have to be forged, but no pre-modern forge is likely to be large enough to accommodate plates of this size and thickness. Imagine trying to produce these plates by hand with a hammer and a charcoal forge.
The failure of the siege towers actually illustrates neatly why larger and more sophisticated armies tended to use earthwork ramps. Against a large defending force which can afford to have men stationed several ranks deep along the entire wall, the orcs disembarking the towers face fearful casualties almost immediately and don’t appear to have been able to establish a meaningful foothold on the wall (note there is no real orc presence in the city proper until the gate comes down) – it simply isn’t possible to move enough orcs up those ladders fast enough and they are too vulnerable while doing so. By contrast, an army moving up and earthwork ramp can do so with shields in front of them, and under movable cover, arriving in fighting formation.
All of this is meant as distraction for the effort to breach the gate. The orcs initially try a small, uncovered battering ram, it seems incredible that – for all of the effort that went into protecting the siege towers – neither of the battering rams we see (this small one or Grond) are covered. Fortress designers knew that the gates were the most vulnerable part of any defensive circuit and they designed accordingly – being in front of the gate was the most dangerous position in an entire assault. Most gates were either recessed into the wall (as this one is) or surrounded by towers, or both.
The archers of Gondor fire awkwardly over the edge of the wall down at the attackers, but they needn’t have bothered – at this height and range, dropping heavy rocks would be exactly as murderous (castles were often designed with this in mind, with ‘murder holes’ (a hole in the roof over a gate) and machiolations (small projections out from the wall with a slot to drop things on attackers) for the purpose. ‘Boiling oil’ is often pop-culturally referenced here, and such a thing was used, but more often boiled substances were cheaper, easier to get and easier to keep at the right temperature – water and sand were both used heavily. Hilariously, the orc commander berates his orcs for failing to take the wall, despite having given them a ram which is obviously too small and completely unprotected. Here is an orc truly prepared to command an infantry brigade.
Grond is more impressive and I have to say, apart from two quibbles, I am actually quite impressed by what Jackson’s team have done here. The book notes that the Grond is moved by huge monstrous beasts and worked by mountain trolls. Fantasy writers often really fail to consider how warfare might adapt to the existence of such creatures and instead feature troops wielding weapons designed to defeat humans against monstrous foes to no effect. Real armies are more dynamic than this. The closest thing to a war monster is a war elephant, yet the ancient world developed specific anti-elephant tactics, which in the end proved so effective in the Mediterranean that the war elephant effectively vanished from warfare there by the first century AD (we might talk about anti-elephant tactics at some point, if I can find the right pop-culture reference point).
Here, Jackson realizes visually the possibility to construct a truly massive ram, using the great strength and size of the monsters he has to move and deploy it. My quibbles: first, as noted, Grond is uncovered – it seems likely the trolls would simply become victims of the density of arrow-fire from the walls just like the orcs did. Second – the harnessing of the giant beasts pulling the ram isn’t well designed.
If you look at it, the two pairs of pulling animals (on the right and left) have their yokes chained together, meaning there is no way to get them out from in front of the ram, without completely unyoking them. And then the ram must be pushed the final distance (since you have to remove them from the yokes with enough space to get them out from in front of the siege engine). Funnily enough – here’s a thing I didn’t expect to say – the 1980 cartoon by Rankin and Bass did this better, positioning the creatures to the sides so that they could pull the ram nearly flush with the gate themselves:
The Somme of Gondor
This entire section is a book note, but since I need to put actual block-quotes in it, I’m going to save the quote-box for them (alas, I have not yet mastered nested quote-boxes). But I want to bring attention to how differently framed the emphasis of the siege is in Tolkien’s Return of the King, and some of the influences there. I want to begin by pulling out Tolkien’s description of the beginning of the siege:
Busy as ants hurrying orcs were digging, digging lines of deep trenches in a huge ring, just out of bowshot from the walls; and as the trenches were made each was filled with fire, though how it was kindled or fed, by art of devilry, none could see. All day the labour went forward, while the men of Minas Tirith looked on, unable to hinder it. And as each length of trench was completed, they could see great wains approaching; and soon yet more companies of the enemy were swiftly setting up, each behind the cover of the trench, great engines for the casting of missiles (RotK, 104-5)
First, I should note that Tolkien, unlike so many other fantasy writers, has remembered what trenches are for in pre-modern sieges. They are not there to hide in from missile fire, but rather as obstructions, to prevent the enemy in the city from rushing out to try to destroy the siegeworks. Sallies to light fire to rams or towers, or wooden supports for rams, were key parts of siege warfare – a well executed sally could waste weeks or months of labor and possibly cause the failure of a siege. That the orcs fill their trenches with some sort of evil fire drives home their purpose as obstacles. Note also that the catapults are unloaded from wains (wagons), and assembled on site, which is historically accurate.
But even more so, I have always been struck by how this passage – the creeping setup of the artillery, the knowledge of an impending assault, the steady assembly of trenchworks too far away to hinder, evokes the Western Front of World War 1. Tolkien was in the war, going over the top with his men at the Somme in October, 1916, although he was frequently ill with trench maladies and thus saw less combat than he might have. It probably saved his life: as Tolkien himself noted later, he lost all but one of his best friends in the war. The horror of the experience stuck with him and informed his writing (interestingly, while Tolkien frequently objected to the idea that Lord of the Rings was a Cold War or World War II analogy, he was open about how it was informed by his experience in the First World War).
Elsewhere in the siege, he gives a gripping sense of the creeping dread of knowing an assault is coming and being able to do nothing about it:
‘Nay,’ they said, ‘not even if the Nameless One himself should come, not even he could enter here while we yet live.’ But some answered: ‘While we yet live? How long? He has a weapon that has brought low many strong places since the world began. Hunger. The roads are cut. Rohan will not come.”
Then among the greater casts [of catapults] there fell another hail, less ruinous but more horrible. All about the streets and lanes behind the Gate it tumbled down, small round shot that did not burn. But when men ran to learn what it might be, they cried aloud or wept. For the enemy was flinging into the city all the heads of those who had fallen fighting at Osgiliath, or on the Rammas, or in the fields…in vain men shook their fists at the pitiless foes that swarmed before the Gate. Curses they heeded not, nor understood the tongues of western men, crying with harsh voices like beasts and carrion-birds. But soon there were few left in Minas Tirith who had the heart to stand up and defy the hosts of Mordor. For yet another weapon, swifter than hunger, the Lord of the Dark Tower had: dread and despair. (RotK, 105-6)
Indeed, Tolkien frames the entire siege as being about morale and despair. He is, after all, explicit at the beginning – with the voice of the narrator – to declare that Minas Tirith is so well fortified that its citadel could never fall if any could hold weapons to defend it (RotK, 25). But Tolkien himself had been to war and knew there were many things other than bullets that might slay a man. Seeing the body of a fallen friend, shredded by artillery fire – it was often not possible for the bodies of the fallen to be removed from the No Man’s Land between the trenches on the Western Front – might render a soldier as ineffective as any wound.
In the film, the defense of Minas Tirith collapses because Grond breaches the gate and then mountain trolls push through the defenders. Mordor triumphs by raw physical might smashing aside the defense. But in the books, the defense hinges not on military strength – of this, Gondor has enough – but on courage in despair. Battles and sieges are rarely won because one side kills the entire enemy army – or even most of it. They are usually won because the will of one side to resist is broken.
This morale-focus is actually keenly correct on Tolkien’s part. Far more cities and castles were taken by surrender, or else by betrayal, than were ever taken by storm. The example of the Siege of Antioch (1098) – where a single guard opened the gate to the Crusaders after a grueling siege – was typical. Likewise, I’ve mentioned Warwolf at the Siege of Sterling Castle (1304) – in the end it wasn’t Warwolf’s firing power that defeated the castle, but its imposing bulk – the defenders began surrender negotiations upon seeing the size of the machine. The more overwhelming the might of the enemy, the deeper the despair of the defenders, the most difficult it became to maintain the solidarity in defense that was required to successfully resist a siege.
Pop culture and video games often give the impression, however, that battles are won by killing all of the enemies. Opponents in war-games (save for dry, strategy titles) rarely flee, or cower, or become combat ineffective out of terror or despair. By the time Grond advances:
On they came, reckless of their loses as they approached, still bunched and herded, within the range of bowmen on the wall. But indeed there were too few now left there to do them great damage, though the light of the fires showed up many a mark for archers of such skill as Gondor had once boasted. Then perceiving that the valour of the City was already beaten down, the hidden Captain put forth his strength… (RotK, 107-8).
In the end, the terror and despair is so great that when the gate is finally breached, only Gandalf stands to defend it, upon Shadowfax, “who alone among the free horses of the earth endured the terror, unmoving, steadfast as a graven image in Rath Dinen” (RotK 113; I will never quite forgive Jackson for having Shadowfax quail and throw Gandalf in the movie adaption of this moment – or for having Gandalf’s staff break – a signal of defeat – in the film during a confrontation that, in the book, Gandalf clearly wins).
This sensitivity to the psychology of soldiers, largely absent from Peter Jackson’s work (where the Ringwraiths do not inflict creeping dread, but sudden terror like a crowd-control spell in a video-game), I think must be a product of Tolkien’s war experience. It is not something I have encountered in Anglo-Saxon epic (admittedly, I have not Tolkien’s deep knowledge of it), the other great well he draws on for his battle narratives. Return of the King, specifically, among the Lord of the Rings, deserves a place in the pantheon of First World War literature – not only for this scene, of course (Frodo’s efforts to live with trauma after the scourging of the shire rate, to me at least, as one of the greatest and saddest passages in all of English literature) – alongside works like All Quiet and A Farewell to Arms, yet it seems it is rarely appreciated as such.
I actually suspect Peter Jackson understood this. He has an abiding interest in the First World War – if you haven’t seen his They Shall Not Grow Old, fix that – and his decision to keep Frodo’s woundedness, even as the scourging of the shire was dropped, speaks to his understanding of how crucial it was to the story. Nevertheless, where Jackson’s Siege of Gondor is about might, Tolkien’s Siege is about courage and despair – the question is not ‘can the men of Gondor resist’ but will they?
I’ve sounded rather harsh on Peter Jackson’s adaptation here, nitpicking here and there, but on the whole, I actually think the siege sequence is a cinematic triumph. Jackson takes some liberties in order to deploy filmic language (mostly paralleling the orcs with Nazis), but while some of the designs are outlandish, the overall sequence is plausible. Both sides of the assault are, admittedly, taken up to eleven compared to historical precedent – the number of siege engines, the size of the armies, the scale of the fortifications are all mythic, rather than historical, in scale. Nevertheless, while Jackson owes something here to the books, but there is an element to the success of this sequence that is all his own.
What I particularly like is how it contrasts with Helm’s Deep from Two Towers. Helm’s Deep is the longer sequence – and yes, one of these days, I’ll discuss it too. But in filling in the visual details, Jackson captures something on screen that was in the books about the real difference between Sarumon’s army and Mordor’s army – the army of Mordor is far larger, but also far more prepared and better organized. Jackson uses his CGI spectacle (a lot of Helm’s Deep was practical effects – and weeks of night shoots) to communicate very strongly that the army of Mordor is not only quantitatively bigger than the army of Sarumon, but also qualitatively superior. Viewing them side by side immediately puts the lie to Sarumon’s vain and arrogant notion that he could manipulate Sauron by joining him.
Next time, the Rohirrim arrive and we look at probably the most famous cavalry charge in all of cinema.