Collections: The Siege of Gondor, Part I: Professionals Talk Logistics

(Note: Thanks to the effort of a kind reader, this post is now available in audio format! The playlist for the entire series may be found here.)

This is the first part of a six-part (II, III, IV, V, VI) series I expect to roll out taking a historian’s look at the Siege of Gondor in Peter Jackson’s Return of the King.  We’re going to discuss how historically plausible the sequence of events is and, in the process, talk a fair bit about how pre-gunpowder siege warfare works. As with other Collections posts, this series will come out one-per-week, on Friday, until it’s done.  This is, after all, a very long and involved sequence and there is a lot of context to work in.

Book Notes: While I am not going note every time Jackson diverges from the books, I will note significant divergences when they impact the historical review.  When I do so, I’ll place those portions in a little box like this.  I am one of those people who was a book person first – I heard these stories for the first time literally before I could read, having them read to me by my parents.  That said, discussing the book here is more than just nostalgia.  Tolkien’s deep knowledge of Anglo-Saxon literature left him with a fairly keen sense of how pre-modern battle worked.  His own experience in the First World War also leaves deep impressions on the narrative, as we’ll see.  That deeper knowledge and personal experience is often most visible in places where the film and the books part ways, and offer some of the best opportunities to illuminate key details about pre-modern siege warfare.

As an aside: welcome new readers! If you enjoy this series, check out the ‘resourceslinks at the top of the main page to see some of my other writings and if you want updates on my future projects, you can follow me on twitter (@BretDevereaux) or subscribe using this button:

And if you want to support this project, you can support me on Patreon, or you can help me by sharing this and other posts. I get all of my audience through word of mouth, so if you want to help me spread some history knowledge, tell your friends!

(Subsequent edit: Also, for readers reaching this post in 2020 or later, you will be happy (or deeply distressed) to know there is a companion series on the Battle of Helm’s Deep in eight (yes, I know, eight) parts: (I, II, III, IV, V, VI, VII. VIII). But don’t worry – you should still read this (mercifully shorter) series first!)

So, without further ado, let’s get to it.

Objectives: Strategy and Operations

Note that the orcs are five abreast and tightly packed. An army does not normally march this tightly over long distances (note, for instance, there is no space for backpacks or any of the other kit soldiers have to lug).

We’re going to start here, with the army of Minas Morgul marshaling out of the main gate.  It is an incredible scene, the seemingly endless line of orcs marching past our hidden heroes, who crouch, overawed by the spectacle of it.

That may seem a touch early to start a review of the siege, but there are two points to this, both of which are historically illuminating.  What we are watching at this stage is what is called operations – the coordinated movement of large bodies of troops to their objective.  Operations is the level of analysis between tactics (how do I fight when I get there?) and strategy (why am I fighting at all?).  And its worth asking, before proceeding any further: what is Sauron’s overall plan and does it make sense?

This map is from the LotR Project ( – go check them out.

Because we’re looking at operations, a map is going to be useful, so here is one – I suggest pulling it out into a separate tab so you can keep a sense of where things are.

In this context, the immediate operational goal of Sauron’s army is getting the army, intact, to Minas Tirith to lay siege to it; in comparison, the strategic goal of the campaign is the destruction of the Kingdom of Gondor through the capture of its capital and primary defense (Minas Tirith).

This set of objectives and the means chosen to achieve them are immediately historically plausible.  Pre-modern states – like the Kingdom of Gondor – often had a very limited administrative apparatus which was focused in a single place (it is hard to distribute your administration when the best communications technology you have is “man on horse”).  The destruction of that administrative center might very well be enough to end the war.

More broadly, it is fair to say that most pre-modern operations had this objective: to deliver the siege (hat tip to @MilHist_Lee).  While fiction tends to focus a lot on battles, an open field battle was rarely the objective of a campaign – rather a battle resulted because one army, in transit to an enemy city to begin a siege, was opposed by another army attempting to stop them. 

Let’s be clear on this point: ancient and medieval warfare was mostly about sieges. Particularly during the Middle Ages (in Europe) sieges (and raiding) were common, but large pitched battles were rare. Pitched battles were far more common in the ancient world (think Persia, Macedon, Rome, etc) – this is a product of the larger size and greater organizational capabilities of those armies. Because of that, they were much more likely to win a siege – the defenders know that, so if you cannot defeat them in battle, it was better to surrender than try to hold out (a city taken in a siege was often subject to pillaging and massacre). Of course, this only works if the army approaching you can credibly siege your city – even in the absence of a siege itself, the threat of siege was crucial.

Another short side-note here: we’ve been talking in terms of fortified cities, because that is what Minas Tirith is, but castles – which are fortified private residences – operate mostly on the same basic set of rules (the court and household of the military aristocrat whose house has been fortified is generally also the local administrative apparatus). In both cases, control of the fortified center is what allows for control of the countryside around it. Consequently, the goal of the army is not to fight a big battle, but to deliver a siege to the fortified center, castle or city, which can capture it and thus take over administration of the countryside (where most of the economic activity – farming – takes place).

The goals here (operational objectives) of Sauron’s plan here absolutely check out.  Minas Tirith contains most of Gondor’s military, and functionally all of its leadership and administration – its destruction could very well be war-ending. At the very least, control of Minas Tirith would open the rest of Gondor to raiding as well as enable Sauron to control the resource-rich Pelennor Fields. Delivering a powerful and effective siege (the operational objective) is very likely to lead to victory over Gondor and territorial control of it (the strategic objective). Now the question is Sauron’s plan to achieve that operational objective (we will talk about Gondor’s planning too – a little later in the series).

Now, as we’ve noted, operations are all about the problem of moving large armies.  Late season Game of Thrones notwithstanding, armies do not generally teleport around the world, they have to march.  That imposes all sorts of restrictions and costs on movement: where are the roads?  Mountain passes?  River Crossings?  The terrain Sauron’s army must attack over is defined (as we’ll see) by a series of transport bottlenecks that have to be negotiated in order to deliver the siege.  Then there is the issue of supplies – even orcs need to eat.

Logistics of the Army of Mordor

Looking at the logistics of moving the Army of Mordor to Minas Tirith is actually a great way to introduce some of these problems in more depth.  They say ‘amateurs talk tactics, but professionals study logistics.’  Well, pull up a chair at the Grown-Ups Table, and let’s study some logistics.

The army Sauron sends against Minas Tirith is absolutely vast – an army so vast that it cannot fit its entire force in the available frontage, so the army ends up stacking up in front of the city:

We’ll come back to almost everything here later – but for now: I’m sure those trolls are thrilled to have to push the weight of all of these armored orcs what appears to be a mile or more to the wall. Rather than them, you know, getting in the tower a bit closer to the target.
Another look at the orc army. Note that it is so huge that the entire army cannot get to the city, because the rest of the army is in the way (there is insufficient frontage for the entire army to deploy), forcing the army to stack up.

The books are vague on the total size of the orcish host (but we’ll come back to this), but interview material for the movies suggests that Peter Jackson’s CGI team assumed around 200,000 orcs.  This army has to exit Minas Morgul – apparently as a single group – and then follow the road to the crossing at Osgiliath.  Is this operational plan reasonable, from a transit perspective?

In a word: no.  It’s not hard to run the math as to why.  Looking at the image at the head of the previous section, we can see that the road the orcs are on allows them to march five abreast, meaning there are 40,000 such rows (plus additional space for trolls, etc). Giving each orc four feet of space on the march (a fairly conservative figure), that would mean the army alone stretches 30 miles down a single road.  At that length, the tail end of the army would not even be able to leave camp before the front of the army had finished marching for the day.  For comparison, an army doing a ‘forced march’ (marching at rapid speed under limited load – and often taking heat or fatigue casualties to do it) might manage 20 to 30 miles per day.  Infantry on foot is more likely to average around 10 miles per day on decent roads.

Ideally, the solution to this problem is to split the army up.  By moving in multiple columns and converging on the battlespace, you split one impossibly long column of troops into several more manageable ones.  There is a danger here – the enemy might try to overwhelm each smaller army in turn – but Faramir has had to pull his troops back out of Ithilien, so there is little risk of defeat in detail for the Army of Mordor.  The larger problem is terrain – we’ve seen Ithilien in this film and the previous one: it is heavily forested, with few roads.  What roads exist are overgrown and difficult to use.  Worse yet, the primary route through the area is not an east-west road, but the North-South route up from Near Harad to the Black Gate.  The infrastructure here to split the army effectively simply doesn’t exist.

A map from regular Earth, rather than Middle Earth. This is Napoleon’s Ulm Campaign (1805) – note how Napoleon’s armies (the blue lines) are so large they have to move in multiple columns, which converge on the Austrian army (the red box labeled “FERDINAND”). This coordinated movement is the heart of operations: how do you get your entire army all to the battlefield intact and at the same time?

This actually understates the problem, because the army of Morder also needs supplies in order to conduct the siege.  Orcs seem to be able to make do with very poor water supplies (Frodo and Sam comment on the foulness of Mordor water), so we can assume they use local water along the march, but that still leaves food.  Ithilien (the territory they are marching through), as we have seen in the film, is unpopulated – the army can expect no fresh supplies here (or in the Pelennor beyond, for reasons we’ll discuss shortly).  That is going to mean a baggage train to carry additional supplies, as well as materials for the construction of all of the fancy siege equipment (we, in fact, later see them bringing the towers pre-built – we’ll get to it).  This would lengthen the army train even more.

All of that raises a second point – from a supply perspective, can this operation work?  Here, the answer is, perhaps surprisingly, yes.  Minas Morgul is 20 leagues (around 60 miles) from Minas Tirith.  An infantryman might carry around (very roughly) 10 days or so of rations on his person, which is enough to move around 120 miles (these figures derive from K. Chase, Firearms: A Global History to 1700 (2003) – well worth a read! – but are broadly applicable to almost any army before the invention of the railroad).  The army is bound to be held up a bit along the way, so the Witch King would want to bring some wagons with additional supplies, but as a matter of supply, this works.  The problem is transit.

As a side note, the supply issue neatly explains the aggressive tactics the Witch king employs when he arrives at Minas Tirith, moving immediately for an assault rather than a siege.  Because the pack animals which pull wagons full of food eat food themselves, there is literally no amount of wagons which would enable an army of this size to sustain itself indefinitely in a long siege.  The Witch King is thus constrained by his operational plan: the raw size of his army means he must either take the city in an assault quickly enough to march most of his army back, or fail.  He proceeds with the appropriate sense of urgency.

That said, the distances here are short: 60 miles is a believable distance for an army to make an unsupported ‘lunge’ out of its logistics network.  One cannot help but notice the Stark (hah!) contrast with the multi-hundred-mile supply-free lunges in the TV version of Game of Thrones, which are far less plausible.

We’ve Had One Logistics, Yes.  But What About Second Logistics?

BOOK NOTE: In the film, Jackson has split the host of Mordor into three groups (the fleet, the Haradrim and the Orcs) each of which moves and arrives as a single unit.  As discussed above, this is insufficient to resolve the overwhelming logistics problems of such large armies.  However, the books largely resolve this issue.  While the size of the orc army in the book is never spelled out, it is clearly quite a bit smaller than 200,000 (the ever-trusty wiki suggests around half the size – putting it just a bit over the upper-bound of an army that might move as a single group in this kind of terrain).

But Tolkien notably does not have the orc army move as a single massive group.  Instead it sets out from multiple logistics bases, with the main force leaving Minas Morgul and attacking as Osgiliath, but a secondary force leaving Mordor via the Black Gate and crossing the Anduin at Cair Andros (see the map) into Anorien (RotK, 104).  Denathor had been aware of this possibility, but had concluded (correctly, it seems) that if he could not hold Osgiliath it would matter little if he held in the North and so prioritized one over the other – in the event he lost both, but it isn’t clear how that could have been avoided (RotK, 98).

Breaking up the orc army into two columns of 50,000 orcs with supplies resolves most of the problems of the previous section.  Moreover, by having those armies leave from multiple logistics bases, we may imagine an entire road network being used, instead of what we see in the film, where the orc army is bottlenecked by individual bridges and causeways.  The late arrival of the ‘Southrons’ makes perfect sense as well – their forces came north through Ithilien, but could not possibly cross at Osgiliath until the entire orc army was across.  Given that the Southrons – who have infantry, cavalry and war elephants in the books, rather than the nearly all-elephant army seen in the film – seem more mobile than the orc army (which is entirely on foot), one wonders if this column was intended as an exploitation force.  Once the orcs had taken Minas Tirith – or at least neutralized it via siege – the highly mobile Southron cavalry could scout and raid deeper into Gondor, exploiting the breach in the defenses at Osgiliath.  In that case, it makes perfect sense to put these soldiers further back in the marching order.

In terms of supplies, the books keep very careful track of the timeline of the assault.  The main body of the army departs Minas Morgul on March 9th (Frodo actually witnesses this event, placing it securely).  The army from the Black Gate heading to Caer Andros left earlier – it takes the crossing on the 9th.  The assault on Osgiliath, perhaps 20 miles away (maybe less) comes on the night of the 11th (suggesting a very plausible 10ish miles per day marching speed).   By the morning of March 12th, Faramir has been forced from Osgiliath and is defending the Rammas Echor, which is lost later that day.  The orcish army takes the Pelennor on the 13th, and reaches Minas Tirith on the 14th, taking thus two days to cross the Pelennor after taking the causeway forts (around 15 miles per day, quick but not unreasonable).  The assault begins in earnest on the night of the 14th, with the Rohirrim arriving on the morning of the 15th.

We can suppose that the Witch King did not assume his army would be destroyed in the field for supply purposes – meaning he would have to be able to feed them after the battle (you might argue that an evil wraith cares little for his army and this may be true, but angering 120,000 starving and armed orcs would be foolish). Given how aggressive the assault is, he probably also might have expected to take the city no later than the 16th or perhaps 17th (indeed, he very nearly takes it in the morning hours of the 15th).  He can perhaps count on getting some supplies from the rich farmland of the Pelennor or from the storehouses of the city, once taken, but in practice then, he needs around 20 days of operational endurance (10 days out, 10 days back).  That is, in fact, well within the possible, but it will take a significant supply train.

Indeed, the Witch King’s aggressive assault on the city makes good sense from two different perspectives: from the supply perspective, he cannot afford to settle down to a long siege with so large an army. From a larger operation perspective, he must know about the beacon system and that aid is likely en route to Gondor – better to move quickly with a massive force and overwhelm the defenders before Minas Tirith – a formidable defense position – could be reinforced.

An infantryman can carry about 10 days’ worth of food.  Adding one large wagon per each company of 30 infantrymen doubles this distance, but diminishing returns hit fast (because the wagon-driver and animals eat the food in the wagon for each day travelled).  For the Minas Morgul army – around 60,000 strong – that would mean some 2,000 wagons (and the same number again for the army out of the Black Gate).  It’s a large number, but not an impossible one.

However, there is a better option available than loading up on wagons: naval supply.  While the film only shows dedicated crossing craft in the assault on Osgiliath, the books note “floats and barges in great number” already in Osgiliath.  Naval supply, by riverboat or by ship, is far more efficient than overland supply (moving supplies by water is roughly twenty times more efficient than moving the same supplies by land in the pre-modern era).  Add in additional supplies stockpiled at Osgiliath and carried by the Umbar fleet.  By relying on those supplies, the Witch King’s own forces could travel light, with minimal wagons (for siege equipment) and thus move much faster.  Such audacious ‘lunges’ between supply networks were hallmarks of the success enjoyed by Caesar and Alexander, but they could also go brutally wrong – Marc Antony’s failure at Actium (and subsequent defeat in the Final War of the Roman Republic) is a textbook case of a risky lunge failing and it resulted in the complete loss of his army and nearly his entire fleet as well.

I find such an audacious attack plan perfectly in keeping with both Sauron’s and the Witch King’s character and motivation.  The former has endless reserves to fall back on and so can afford to lose an army in a risky gambit.  Moreover, he is rushing his offensive in an attempt to forestall Aragorn, who he suspects has taken the One Ring (or so Gandalf, correctly it seems, supposes).  Meanwhile, the Witch King is, throughout the battle, confident to the point of arrogance (perhaps because he believes himself invulnerable).  Much like Alexander, he opts to lead storming attacks himself and flies down to fight an enemy king (Théoden) directly. For such an aggressive commander, an audacious plan to make an unsupported lunge and have the fleet catch up with naval supply later seems perfectly in character. 

In conclusion: While the broad outlines of Mordor’s plan make clear operational and strategic sense, Jackson has inflated the orc army to be a bit too large.  Shrinking that army back down and splitting it up – as the books do – resolve this issue into a largely plausible operational plan.  A lot of this may be credited to Tolkien, but I think Jackson’s skill with the material must also be noted.  Film is a very compressed medium and Jackson simply does not have the time to show us Caer Andros (though he mentions it) or the orc army taking multiple routes to Osgiliath.

Jackson does, however, do an admirable job in keeping track of where the armies are and where they are going even in that very compressed filmic medium – something the genre generally has struggled with (indeed, Jackson himself struggles with it in The Battle of Five Armies, repeatedly losing track of where things are in the titular battle).

So far, this is a good beginning: the bad guys have a solid plan.  Next week, we’ll see that plan contact the enemy, and take a look at the plan Jackson gives to the Good Guys (hint: it’s not as good), and actually get around to delivering that siege.

Hey, uh, boss? Why are we hiding behind walls instead of in battle formation?

79 thoughts on “Collections: The Siege of Gondor, Part I: Professionals Talk Logistics

  1. Bret, some further proofreading corrections, in case you have the time and/or interest:

    And its worth asking -> And it’s worth asking
    leaving Minas Morgul and attacking as Osgiliath -> leaving Minas Morgul and attacking at Osgiliath
    Denathor had been aware -> Denethor had been aware
    time to show us Caer Andros -> time to show us Cair Andros

  2. But isn’t it an established fact in Tolkiens universe that Orcs can eat each other or just their slain enemies? I mean going in immediately fighting the defending army gives a lot of dead body’s ready to be cooked up for the orc Army.

    1. A friend of mine made this exact point and I did some back-of-the-envelope calculations. An 80kg orc might consist of 60kg of recoverable meat (same proportion as a deer) or 30 orc-days of rations. But you get to a tricky point where casualties that are too light don’t leave enough food, and casualties that are too heavy leave the army too weakened to fight.

      Hypothetically, let’s say 10,000 orcs encounter a force of 2,000 mounted rangers – the rangers fight a delaying action and fall back with 5% casualties, and the orcs take, say, three times as many casualties. 100 dead humans plus 300 dead orcs plus 100 dead horses is 42,000 kg of potential rations. For the surviving 9700 orcs that’s two and a bit extra days of rations – and all lying conveniently concentrated, so no need to pause on the march or scatter to forage.

      But, crucially, the orcs have to win to do this. Just as controlling the battlefield after Kursk allows you to recover and repair a lot of your tank losses, controlling the battlefield after a delaying action on the approach to Minas Tirith allows you to recover the dead and eat them. But the besieger dead in a siege might not be easily recoverable because, almost by definition, they’ll be lying within effective range of the defenders’ weapons, and the defenders are still there.

      (Also, the dead horses make up a big part of that 42,000kg. An intelligent orc commander might even go out of his way to invite attack by cavalry during an advance, because of the logistic benefit. Exchanging 300 soldiers for an extra two days of endurance is a good trade.)

  3. Thanks for the fantastic articles. I wish I had time to read them all today!

    > moving supplies by water is roughly twenty times more efficient than moving the same supplies by land in the pre-modern era

    Just a detail, but that could be misleading. I learned somewhere [1] that rivers costs about 1/5 of roads for transport, and it’s sea-based transport that costs 1/20th of roads.


    1. There is, with any sort of effort in writing and explaining things historically, some point at which some degree of simplification is necessary. Part of why I like the structure of this blog is that I can use subsequent essays to clarify and expand on points in previous essays. Thus the “Preposterous Logistics” essay (and its accompanying “How Fast Do Armies Move”) along with the Lonely Cities posts go into more detail on the human terrain and logistics of the countryside.

  4. Why did Sauron use such a large army, if it couldn’t even deploy in front of Minas Tirith? Wouldn’t his logistical problems be eased by holding back his forces?

  5. Brett, *very* belatedly here–after I had to look this up for another project–I would like to ask about your claim: The main body of the army departs Minas Morgul on March 9th (Frodo actually witnesses this event, placing it securely).

    But how does this mesh with the Tale of Years in Appendix B, where the last entry for the date of March 10 has: “Frodo passes the Cross-roads, and sees the Morgul-host set out.” ???

    1. (sorry, Bret…you’d think after following your blog for the past two years that I’d remember how to spell your name!)

      1. No, I wouldn’t have thought it of you

        Actually, I know of several oddities that have crept into the writings of other Tolkien enthusiasts because they had not realized that the newest version of LOTR–as corrected and updated by Hammond and Scull in 2004/2005–included several differences from the, by then, long-running versions on both sides of the pond.

  6. “Because of that, they were much more likely to win a siege – the defenders know that, so if you cannot defeat them in battle, it was better to surrender than try to hold out ”

    I’m surprised by the wording here. Sun Tzu said “the worst policy of all is to besiege walled cities,” and naively I would have thought that there will be very large defenders’ advantage in walled cities in most pre-modern eras. So naively I would have guessed that from a defenders’ perspective, they’re much more likely to win defending a siege than in open warfare.

    1. Sun-Tzu favored a much more mobile strategic model; rather more to the point, at the time he was writing China did not yet have much in the way of advanced siege equipment and techniques. His situation can’t really be compared to, say, 13th century Europe, where a besieged city was eventually doomed, given time and no relief army.

      1. Also, doesn’t he make a big deal out of not being constrained by traditional “locate and siege the administrative center” ideas? On top of not having the tech, he seemed (in my super-uneducated reading almost 30 years ago) to regard it as a philosophical trap that can limit your entire strategic line of thought.

Leave a Reply