This is the first part of a planned five-part series covering some of the basics of fortifications, from city walls to castles and field fortifications! We are going to discuss what fortifications were for and how their design changed in response both to different strategic and operational conditions and also to changing technology. Throughout this, we are going to stay mostly centered geographically on the broader Mediterranean world, though I will make occasionally connections to the ways in which fortification techniques were similar or different in other areas of the world. Naturally, with this chronological range, we are not going to cover everything or even most things in that period, but the aim here is by hitting the ‘high points,’ we can cover some of the major principles which determined why fortifications took the form that they did (and thus why they were assaulted in the ways that they were).
Though we’ll also discuss some of the earliest evidence for permanently fortified cities, this week we’re going to begin principally with a look at the problem of a fortified city from the perspective of the attacker as a way to set the foundation for the kinds of assaults that the designers of defensive systems are looking to defeat. In particular, we’re going to look at the evidence for the siege-assault toolkit employed by the Neo-Assyrian Empire (911-609 BC), who enjoyed a well-earned reputation as one of the most effective siege-attackers of the ancient world, matched perhaps only by the Romans and Macedonians. We’re also going to discuss the strategic concerns that motivate the attacker since, as Clausewitz tells us (drink!) the political object fundamentally dictates (or ought to dictate) the conduct of war and that includes siege warfare.
As always, if you like what you are reading here, please share it; if you really like it, you can support me on Patreon. And if you want updates whenever a new post appears, you can click below for email updates or follow me on twitter (@BretDevereaux) for updates as to new posts as well as my occasional ancient history, foreign policy or military history musings.
A View From Outside
Now the besieger’s side of the equation may seem like an odd place to start a primer on fortifications, but it actually makes a fair bit of sense, because the capabilities of a potential attacker is where most thinking about fortification begins. Siegecraft, both offensive and defensive, is a case of ‘antagonistic co-evolution,’ a form of evolution through opposition where each side of the relationship evolves new features in response to the other: neither offensive siege techniques nor fortifications evolve in isolation but rather in response to each other.
In many ways the choice of where to begin following that process of evolution is arbitrary. We could in theory start anywhere from the very distant past or only very recently, but in this case I think it makes sense to begin with the early Near Eastern iron age because of the nature of our evidence. While it is clear that siege warfare must have been an important part of not only bronze age warfare but even pre-bronze age warfare, sources for the details of its practice in that era are sparse (in part because, as we’ll see, siege warfare was a sort of job done by lower status soldiers who often didn’t figure much into artwork focused on royal self-representation and legitimacy-building).
But as we move into the iron age, the dominant power that emerges in the Near East is the (Neo-)Assyrian Empire, the rulers of which make a point of foregrounding their siegecraft as part of a broader program of discouraging revolt by stressing the fearsome abilities of the Assyrian army (which in turn had much of its strength in its professional infantry). Consequently, we have some very useful artistic depictions of the Assyrian army doing siege work and at the same time some incomplete but still very useful information about the structure of the army itself. Moreover, it is just as the Assyrian Empire’s day is coming to a close (collapse in 609) that the surviving source base begins to grow markedly more robust (particular, but not exclusively, in Greece), giving us dense descriptions of siege work (and even some manuals concerning it) in the following centuries, which we can in turn bring to the Assyrian evidence to better understand it. So this is a good place to start because it is the earliest point where we are really on firm ground in terms of understanding siegecraft in some detail. This does mean we are starting in medias res, with sophisticated states already using complex armies to assault fairly complex, sophisticated fortifications, which is worth keeping in mind.
That said, it should be noted that this is hardly beginning at the beginning. The earliest fortifications in most regions of the world were wooden and probably very simple (often just a palisade with perhaps an elevated watch-post), but by the late 8th century, well-defended sites (like walled cities) already sported sophisticated systems of stone walls and towers for defense. That caveat is in turn necessary because siegecraft didn’t evolve the same way everywhere: precisely because this is a system of antagonistic co-evolution it means that in places where either offensive or defensive methods (or technologies) took a different turn, one can end up with very different results down the line (something we’ll see especially with gunpowder).
We’ve gone over this before, but we should also cover the objectives the attacker generally has in a siege. In practice, we want to think about assaults fitting into two categories: the raid and the siege, with these as distinct kinds of attack with different objectives. The earliest fortifications were likely to have been primarily meant to defend against raids rather than sieges as very early (Mesolithic or Neolithic) warfare seems, in as best we can tell with the very limited evidence, to have been primarily focused on using raids to force enemies to vacate territory (by making it too dangerous for them to inhabit by inflicting losses).1 Raids are typically all about surprise (in part because the aim of the raid, either to steal goods or inflict casualties, can be done without any intention to stick around), so fortifications designed to resist them do not need to stop the enemy, merely slow them down long enough so that they can be detected and a response made ready. We’ll actually be discussing some of these sorts of defenses (used by armies on the march rather than permanent settlements) next time.
In contrast, the emergence of states focused on territorial control2 create a different set of strategic objectives which lead towards the siege as the offensive method of choice over the raid. States, with their need to control and administer territory (and the desire to get control of that territory with its farming population intact so that they can be forced to farm that land and then have their agricultural surplus extracted as taxes), aim to gain control of areas of agricultural production, in order to extract resources from them (both to enrich the elite and core of the state, but also to fund further military activity).
Thus, the goal in besieging a fortified settlement (be that, as would be likely in this early period, a fortified town or as later a castle) is generally to get control of the administrative center. Most of the economic activity prior to the industrial revolution is not in the city; rather the city’s value is that it is an economic and administrative hub. Controlling the city allows a state to control and extract from the countryside around the city, which is the real prize. Control here thus means setting up a stable civilian administration within the city which can in turn extract resources from the countryside; this may or may not require a permanent garrison of some sort, but it almost always requires the complete collapse of organized resistance in the city. Needless to say, setting up a stable civilian administration is not something one generally does by surprise, and so the siege has to aim for more durable control over the settlement. It also requires fairly complete control; if you control most of the town but, say, a group of defenders are still holding out in a citadel somewhere, that is going to make it very difficult to set up a stable administration which can extract resources.
Fortunately for potential defenders, a fortification system which can withstand a siege is almost always going to be sufficient to prevent a raid as well (because if you can’t beat it with months of preparatory work, you are certainly unlikely to be able to quickly and silently overcome it in just a few night hours except under extremely favorable conditions), though detection and observation are also very important in sieges. Nevertheless, we will actually see at various points fortification systems emerge from systems designed more to prevent the raid (or similar ‘surprise’ assaults) rather than the siege (which is almost never delivered by surprise), so keeping both potential attacking methods in mind – the pounce-and-flee raid and the assault-and-stay siege – is going to be important.
As we are going to see, even fairly basic fortifications are going to mean that a siege attacker must either bring a large army to the target, or plan to stay at the target for a long time, or both. In a real sense, until very recently, this is what ‘conventional’ agrarian armies were: siege delivery mechanisms. Operations in this context were mostly about resolving the difficult questions of how to get the siege (by which I mean the army that can execute the siege) to the fortified settlement (and administrative center) being targeted. Because siege-capable armies are either big or intend to stick around (or both), surprise is out of the window for these kinds of assaults, which in turn raises the possibility of being forced into a battle, either on the approach to the target or once you have laid siege to it.
It is that fact which then leads to all of the many considerations for how to win a battle, some of which we have discussed elsewhere. I do not want to get drawn off into the question of winning battles, but I do want to note here that the battle is, in this equation, a ‘second order’ concern: merely an event which enables (or prohibits) a siege. As we’ll see, sieges are quite unpleasant things, so if a defender can not have a siege by virtue of a battle, it almost always makes sense to try that (there are some exceptions, but as a rule one does not submit to a siege if there are other choices), but the key thing here is that battles are fundamentally secondary in importance to the siege: the goal of the battle is merely to enable or prevent the siege. The siege, and the capture or non-capture of the town (with its role as an administrative center for the agricultural hinterland around it) is what matters.
The Walls of Jericho
These strategic (and operational) considerations dictate some of the tactical realities of most sieges. The attacker’s army is generally going to be larger and stronger, typically a lot larger and stronger, because if the two sides were anywhere near parity with each other the defender would risk a battle rather than submit to a siege. Thus the main problem the attacker faces is access: if the attacker an get into the settlement, that will typically be sufficient to ensure victory.
The problem standing between that attacking army and access was, of course, walls (though as we will see, walls rarely stand alone as part of a defensive system). Even very early Neolithic settlements often show concerns for defense and signs of fortification. The oldest set of city walls belong to one of the oldest excavated cities (which should tell us how short the interval between the development of large population centers and the need to fortify those population centers was), Jericho in the West Bank. The site was inhabited beginning around 10,000 BC and the initial phase of construction on what appears to be a city wall reinforced with a defensive tower was c. 8000 BC. It is striking just how substantial the fortifications are, given how early they were constructed: initially the wall was a 3.6m stone perimeter wall, supported by a 8.5m tall tower, all in stone. That setup was eventually reinforced with a defensive ditch dug 2.7m deep and 8.2m wide cutting through the bedrock (that is a ditch even Roel Konijnendijk could be proud of!), by which point the main wall was enhanced to be some 1.5-2m thick and anywhere from 3.7-5.2m high. That is a serious wall and unlikely the first defensive system protecting the site; chances are there were older fortifications, perhaps in perishable materials, which do not survive. Simply put, no one starts by building a 4m by 2m stone wall reinforced by a massive stone tower and a huge ditch through the bedrock; clearly city walls is something people had already been thinking about for some time.
I want to stress just how deep into the past a site like Jericho is. At 8000 BC, Jericho’s wall and tower pre-date the earliest writing anywhere (the Kish tablet, c. 3200 BC) by c. 4,800 years. The tower of Jericho was more ancient to the Great Pyramid of Giza (c. 2600 BC), than the Great Pyramid is to us. In short, the problem of walled cities – and taking walled cities – was a very old problem, one which predated writing by thousands of years. By the time the arrival of writing allows us to see even a little more clearly, Egypt, Mesopotamia and the Levant are already filled with walled cities, often with stunningly impressive stone or brick walls. Gilgamesh (r. 2900-2700 BC) brags about the walls of Uruk in the Epic of Gilgamesh (composed c. 2100) as enclosing more than three square miles and being made of superior baked bricks (rather than inferior mudbrick); there is evidence to suggest, by the by, that the historical GIlgamesh (or Bilgames) did build Uruk’s walls and that they would have lived up to the poem’s billing. Meanwhile, in Egypt, we have artwork like the Towns Palette, which appears to commemorate the successful sieges of a number of walled towns:
So a would-be agrarian conqueror in Egypt, Mesopotamia or the Levant, from well before the Bronze Age would have already had to contest with the problem of how to seize fortified towns. Of course depictions like these make it difficult to reconstruct siege tactics (the animals on the Towns Palette likely represent armies, rather than a strategy of “use a giant bird as a siege weapon”), so we’re going to jump ahead to the (Neo)Assyrian Empire (911-609 BC; note that we are jumping ahead thousands of years).
How exactly does an attacker take one of these places?
The Assyrian Siege Playbook
The first thing I want to note here is that we are working in the age before large, bespoke siege artillery. The catapult was a Greek invention and didn’t see widespread use until the fourth century BC; before that, while there were, as we’ll see, dedicated siege engines (like towers and rams), there was not dedicated siege artillery or indeed artillery of any sort. That’s actually convenient for us; popularmodern assumptions about siegecraft tend to place artillery front and center, but prior to the development of effective gunpowder artillery (in the 15th century AD) siege artillery was mostly a secondary supporting arm, not the primary tool; in most cases it typically enabled the traditional siegecraft toolkit to work more effectively and only less often did it open truly new options for attack or defense. Consequently, looking at the system before the catapult can be helpful because it gives us a chance to think about a ‘pure’ form of the system without the temptation to import all sorts of modern assumptions about what artillery should be doing.
Sieges were long, difficult affairs and so the first step in the Assyrian siege playbook was to try to capture cities without lengthy sieges wherever possible. Armies approaching a town would seek to get it to surrender in advance, both to avoid a siege and also to make its grain markets available to supply the army (this sort of pre-surrender was clearly a factor in Assyrian campaigns, but you can see it most clearly in our much fuller narratives of the campaigns of Alexander III).
Of course a town is only going to preemptively surrender in this way if they think there is little hope of actually outlasting a siege. Information in this context is limited: large armies move slowly, to the city likely knows that a large army is coming, but they probably do not know precisely how large, or what its capabilities are, or how long its supplies might last. What they do know is its reputation and here the Assyrians put in quite a lot of effort. The panels I have been showing from the Lachish reliefs, which were mounted in the king’s palace in his capital at Ninevah (and so there to be seen by visiting dignitaries and notables) represent only part of a strategy whereby the Assyrian Empire went out of its way to communicate its effective siege prowess to its neighbors and subjects. Assyrian inscriptions brag about successful sieges, the brutal punishments meted out to cities that resisted and the impressive size and power of the Assyrian army, particularly its professional infantry corps who, as we’ll see, do most of the actual work in a siege.. This effort is actually a huge help to historians, because (for reasons that will soon become clear) it renders visible parts of the Assyrian army that we normally wouldn’t hear much about in the sources.
Assuming that didn’t work, Assyrian armies would aim to breach the city’s defenses. Of course the major obstacles here are walls, which leave three real options: you can go over walls, under walls, or through walls. The through option was provided by battering rams. While in film and TV battering rams are generally depicted as only ever being deployed against gates, in ancient warfare they were often deployed against the wall itself (in part because gate houses tended to be very well defended precisely because the gate represented a vulnerable breach-point). The Romans, for instance, regarded the time before a ram had reached the walls of a settlement (murum aries attigit; “the ram touched the walls”) as the last point at which surrender could be offered (Caes. B.G. 2.32); that the ram is being deployed against the walls and not the gate is explicit in the phrasing.
That said, looking at Assyrian artwork, it is hard not to notice that rams are rarely shown battering the curtain wall (that is, the actual wall blocking ingress into the city; we’ll define this term in more detail later) itself nor gates but rather engaging towers. Most defensive towers, after all, are hollow (because you have internal stairways and fighting positions inside the tower) and so easier to knock over by percussive force. Collapsing a tower won’t create a breach, but as we’ll see a lot of what the attackers are doing here is trying to suppress the enemy’s ability to fire missiles (arrows, slings, javelins, rocks, etc.) from the walls; high projecting towers are ideal firing positions and removing a few might make an approach to the walls substantially safer. A ram could, given enough time, punch through a wall – apply hammer blows to a stone wall enough times and it will come down (and a ram has the advantage of being able to reliably strike the same point with precision, unlike a catapult or early cannon, as well as striking much faster than both, though with far less energy than a cannon), so ramming through a wall was possible, assuming you had the labor and coordination to both keep at it around the clock and guard the ram against missiles and sally attacks from the city.
You might also tunnel under the walls. We do have depictions of the Assyrians digging tunnels under enemy walls (though I don’t have any to hand), although it seems to be a less common motif. Generally such ‘mining’ or ‘sapping’ efforts aimed not to use the tunnel to enter the city (since the small tunnel would be a terrible bottleneck for the attackers) but rather to undermine the walls or towers by digging under them and then collapsing the tunnel. Of course that means the tunnel needs to run deeper than the foundation of the walls, which were often very deep. And once again, beginning a tunnel, perhaps 100 meters or more beyond the wall and then digging your tunnel down several meters and then over 100 meters demanded tremendous coordination of labor and also a not-at-all-trivial level of engineering know-how.
But the primary way the Assyrians would breach a fortification is escalade or going over the walls. Now in the movies, attacks by escalade are all about ladders or siege towers, with the latter rolling up to the wall and dropping big drawbridges that disgorge tons of attacking troops. Both of these things did happen, but their prominence is out of proportion to their use. Ladders absolutely could succeed against unprepared or overwhelmed defenders, but against a decent defense, they offered little chance of success when used alone: a soldier climbing a ladder is just far too vulnerable.
Siege towers could be used to deliver troops to the top of walls (a siege tower was used this way in the 1099 Siege of Jerusalem, although you will note in conjunction with many other siege engines), but their more common use was as elevated firing platforms. By having a siege tower as tall or taller than the towers on the walls, the attackers could use missile weapons (arrows, slings, javelins, etc) to fire down on defenders on the walls; the angles would leave those defenders without any kind of cover, allowing the attacker to quickly suppress and clear the wall of defenders. What seems to have been the standard Assyrian siege-tower was actually a combination engine which incorporated a ram on its lower level (to collapse the enemy tower) and a firing position for archers on the top (to suppress the walls). Such a tower doesn’t appear able to deliver troops to the top of the wall, but it is expertly designed to suppress the defenders along a long stretch of wall.
In Assyrian artwork, these towers are generally in turn supported by more missile fire from the ground. The Assyrians seem to have generally deployed these soldiers in pairs, with a shield-and-spear troops (called sab ariti, ‘shield bearers’ or nas asmare ‘spear men’) to cover the archers who could then fire from safety behind the shields they carried. Archers used like this must have been firing in a fairly high arc and thus have been less accurate and effective than archers firing from the towers, but one can easily imagine how large numbers of archers deployed like this firing at and over the wall could create a ‘beaten zone’ of fire which would force defenders to seek cover, which in turn would enable the other elements of the siege package to do their work near the wall in relative safety.
So if ladders are too dangerous and siege towers are mostly about suppression, how do you get over the wall? Well the answer is a mole or in more common English, a ramp: you build an earthwork ramp up to the wall which you can then attack over directly onto the wall and into the city. Building a ramp like this is a difficult and complex task: the attacker needs both to be able to organize the massive amounts of labor required to move that much earth and the engineering know-how to construct it (which often means wooden support-beams which need to then be protected from fire) and provide enough security to enable the soldiers hauling earth to do so close to the wall.
For the attacker, the coordination problem here is intense. There is a tremendous amount of construction that needs to get done here (also including constructing your own fortified camps and siege positions to keep you safe from sudden raids by the defenders or the appearance of a relieving army); almost none of these siege devices can be transported practically with the army (they’re too big and heavy), so they all have to be built on-site. That means first coordinating the construction of a fortified camp at the same time you are getting other soldiers working on gathering supplies for siege engines. Then it means building the towers at the same time as you have hundreds of soldiers working around the clock beginning construction on the mole (out of arrow-shot), while also foraging the countryside for supplies. Then, as the mole gets closer, you need to deploy the towers, direct arrow-fire from the towers and the ground to suppress the defenders all while conducting what would be a major and difficult engineering project (a wide, earthwork ramp many meters tall) even if people weren’t shooting at you.
Armies that could master that coordination – the Assyrians, Macedonians and Romans all had well-earned reputations for being very good at this – could in turn make their assault more likely to succeed by simultaneously presenting the defenders with many threats at once: rams, towers, ramps, archers, mining operations, ladders at the ready to threaten night attacks. Remember, the defender’s army is almost by definition smaller than the attacker’s force, so the attacker can potentially win by forcing the defender to spread their troops thin. And indeed, you can see in the illustrations here, especially the Lachish reliefs, how the Assyrians seem to do exactly this; we ought to be on our guard for artistic license here, but other ancient descriptions of sieges also show commanders deploying multiple vectors of attack at once to try to overwhelm defenders or confuse them as to the direct of the main assault (for example, this is the core of Scipio Africanus’ strategy in his successful storming of Carthago Nova in Spain in 209).
But what I want to stress here, because we’ll come back to it, is how this version of the attacker’s playbook assumes not merely a large army, but a very well coordinated army. There’s actually another wrinkle here with social status: your elites – the sort of military aristocrats who like to ride into battle on horses or chariots – are unlikely to want to spend their days digging ditches and making ramps, so your army needs to not only be large, but it needs to have large numbers of soldiers who you can order around. For the Assyrians, Macedonians and Romans, this was done by professional or semi-professional infantry, who had the experience to be able to master the complex tasks of siege-work while also not being of such a position that they might feel this kind of labor was ‘beneath them.’ We’ll come back a little later to how this ‘playbook’ changes for armies that do not have that kind of coordination.
The Food Problem
Of course the final option to win a siege was by blockade (note that this is the technical term for this tactic, even when performed on land): simply prevent anyone entering or leaving the city until starvation forces it to surrender. In theory this is the less risky method but there are complications to it which explain why the most effective ancient armies only rarely starved out enemy fortifications. In depictions of this method – I am thinking in particular of strategy games – one you lay siege to an enemy city, a timer starts which, if it runs to zero, results in the surrender of the city.
But in actual practice, there is almost always not only an invisible timer to the defenders, who have a fixed supply of food inside of their city, but also a timer for the attackers. Armies, after all, eat a lot of food and, as we’ve discussed, moving large amounts of supplies long distance overland is simply not feasible in most cases before the advent of railroads. To siege a large city, the attacking army must be large, but a large army requires more food. Now if the attackers are lucky, they can forage much of that food from the surrounding countryside, but since they are not moving, the available food will deplete rapidly (all the more so if the defenders had time to move the food and the farmers inside of the walls in time). Of course in a complex operational environment, the attacker may also be concerned that enemy armies allied to the city under attack might arrive at any time to disrupt the siege, either by forcing a battle or else by essentially laying siege to the siege, something we’ll talk about a bit more next time.
Consequently, while people generally imagine sieges as situations where the attacker merely needs to outlast the defender’s food supplies, it is often the reverse case: in a siege, the defender does not need to defeat the attacker, but merely outlast them. This is going to be important because fortifications are going to be built with this in mind: the goal is not necessarily to ‘hard stop’ the attacker, much less kill all of their troops, but merely to delay them, to draw out the siege process (particularly the preparations necessary for an assault) so long that a siege becomes impractical to sustain such that the attacking army retreats due to want of supply or a changing operational environment or, even better, doesn’t besiege the city at all, determining that the likely cost of a long and drawn out assault outweighs the benefits of campaigning elsewhere or not attacking at all.
And that’s where we’re going to pause for this week. Next time, we’ll jump forward and look at the evolution of Roman fortifications, from single-use fortified marching camps designed to resist raids to the stone-walled fortresses of the late empire.
- For more on this kind of warfare and its ‘pounce-and-flee’ methods, read A. Gat, War in Human Civilization (2006) and L. Keeley, War Before Civilization (1996). For how this kind of raiding could lead to changes in territorial control and resource exploitation, check out this talk by W. Lee
- Note that all elements of that are generally necessary here. Agriculture here is a necessary, but insufficient requirement. You can, for instance, still have raiding-oriented pounce-and-flee warfare with agriculture; the ‘cutting off’ way of war in Native North America provides a clear example of this, see W. Lee, “The Military Revolution of Native North America: Firearms, Forts and Politics” in Empires and Indigenes (2011). State organization also matters, but even then, situations where state-on-state warfare is not primarily about land (such as warfare in 15th/16th century West Africa, where the goal was typically to gain control of labor instead of land), the raid may end up being more important than the siege.