This article is part 3 of a series on Fantasy Fortifications by Toni Šušnjar.
The design of a fortification depends on its purpose and on the threats it is expected to face. A fortification facing only infantry-held weapons, one facing mechanical artillery, and one facing gunpowder artillery will all significantly differ in design characteristics. Some characteristics however will be the same – geography will always provide advantage (or disadvantage) in defending a fort or a city, and thus location has to be carefully chosen. In some cases, location may be good enough to allow the defender to skimp on certain design features – as seen with e.g. Klis fortress, where northern wall is waist-tall at best, thanks to its position on an inaccessible cliff (clissa). In other cases, disadvantageous terrain may have to be compensated with by massive man-made features.
In order to cope with development of artillery, design of fortifications changed with time. First fortifications, which only had to deal with handheld weapons, were simple wooden palisades. These were later supplemented with earthen ramparts
As siege weapons developed, fortifications grew both in height and thickness. Up until the development of gunpowder weapons, walls were taller than their thickness. Even thickest walls were two or more times of their thickness in height – inner walls of Constantinople were 12 meters tall and 4,5 – 6 meters thick. Those of Avilla were 12 meters tall and 3 meters thick; of York, 4 by 1,8 meters and Ravenna, 9 meters tall and 2,4 meters thick.
With development of gunpowder weapons, walls grew thicker and thicker, but also lower. In 14th century, cannons produced more smoke and noise than damage to the walls, but by 15th century fortifications had to be adapted to their presence. This included not only thicker walls, but also reversion to earthwork construction within the stone shell – or sometimes without it. During siege of Constantinople in 1453, hastily constructed earthworks built to replace smashed portions of the wall proved more resilient than the original fortifications. Salses castle built in 1497 is of traditional stone construction, but its stone walls are thick, much lower than in earlier castles, and sunk into the terrain. Sisak fortress in Croatia, built in late 16th century is also a traditional castle, not a Renaissance trace italienne fortress, but its construction was dictated by financial and spatial considerations. Small fortress required smaller garrison, and walls were supposed to be reinforced by earthen ramparts. By 17th century, fortifications had also developed slope and were significantly reduced in height. Between stone facade and earthen ramparts which formed main mass of the wall, such walls were typically much thicker than they were tall. Earth works also served to absorb the impact of projectiles hitting the face of the wall.
Fortifications always utilize geography. Cities, especially large cities, have to be near communications lines: rivers or the sea, as roads are rather limited in carrying capacity up until development of railway and later the internal combustion engine. As a result, it is ideal if a city can be built on a peninsula (e.g. Constantinople, Dubrovnik, Zadar) or on an island (e.g. Trogir, Lindau). Doing this significantly limits possible attack avenues of the enemy. It is however rarely possible, and need for a water source as well as significant arable land in any case constrains where a city may be built (lack of arable land may be less significant if reliable seaborne supply is available). A mountain-top city is thus impossible, and underground city such as Moria is only possible if it has significant mountain valleys available, or else can somehow grow food underground. Consequently, most cities – especially large ones – will be either at sea shore or in river valleys, and always near a water source unless aqueducts can be built. This also holds true for castles, as they will have civilian population around them, while forts outside feudal context may or may not have population living nearby.
Cities may be built on hills, but such cities will typically still have one side at least that is easily accessible for logistics purposes (e.g. Jerusalem). That is where any attack will be concentrated, and where defences will be greatest. This is also true for cities built on a peninsula – Constantinople’s Theodosian walls are much more massive than any seaward walls. Weakest part of any maritime city’s defences is typically the harbour, as it will require large gates to move through supplies and wares. Walls are also often much lower and weaker in harbours (again, Constantinople and Dubrovnik as examples), especially in cities which have significant local naval presence. As such, entrance to harbour was oftentimes (and possibly in majority of cases) defended with a chain (Constantinople and Dubrovnik, again).
Castles are usually much easier to situate at a defensive position, such as a hilltop (e.g. Klis fortress in Croatia). However, depending on purpose of the castle, that may not be practical: a castle which is intended to oversee and protect a trade route must be capable of quickly sending reinforcements to the said route.
Immediate surroundings of the castle or (less often) the city will usually be bare. Further away, but not much further, will be farms and villages. Bare lands form the fire line or a killing zone, where the attacker cannot hide from missile fire from the walls. The keep within the castle often has doors situated above the ground, with wooden access steps: these are burned if the enemy has reached the keep. Within the keep, under the ground, are water cisterns, cellar, and waste pit.
Adapting to Flying Units
In fantasy, there are often flying units where there would have been none in real life: pegasi, hippogriffs, wyverns, dragons and so on. Against a fire-breathing dragon, the best defence would likely be another dragon, or enclosed towers and battlements covered in fireproof materials, depending on potency of the fire.
When it comes to other flying units, defences may be active or passive. Active defences would be scorpions, ballistae and other precision arrow-shooting artillery. Passive defences would include steep reinforced roofs, to stand up to dropped projectiles. Crenelations would also be covered, and chains could be stretched between towers, walls and poles in order to prevent fliers from landing inside the fortress. If flying units are combined with contact explosive projectiles, however, the only defence may be in moving underground; constructing thick roof is technically possible, but would be very expensive for a pre-industrial society. The only alternative is deploying one’s own flying beasties.
Adapting to Magic
Adaptations to magic depend on type and potency of the same. In low-magic settings such as Lord of the Rings and A Song of Ice and Fire, essentially no adaptation is necessary (unless dragons start breeding too much). Things change when mages start throwing fireballs around. If such fireballs are thrown much like artillery, with limited range and elevation, then no adaptation may be necessary. If however they can be summoned from the sky, or else can achieve high elevations, castle will have to have wooden hoardings and turret roofs, covered with some non-flammable material such as thin iron or fresh animal skins. Structures inside the castle may be covered with the same to prevent them from catching fire, assuming logistical capacity for such endeavour exists.
Design of a fort will depend on the terrain. In the open terrain, such as plains, a fort or a castle will often be of a square or maybe circular ground plan. Such castles will also very often – if larger in size – be of a concentric design. If smaller, it may resemble a Roman castrum, with a single wall and rectangular plan (though it may also be round). If a castle is built on significant terrain features, it will adapt to them. For example, a castle built on a cliff or an island may not be concentric at all, but rather have only one level of defences towards inaccessible sides, but layered defence towards the side which is exposed to the attack (for example, see Theodosian walls in Constantinople or Klis fortress in Croatia).
Concentric or layered design – as seen in e.g. Krak des Chevaliers, and already mentioned examples – will typically have inner wall overlooking the outer wall. This allows the attacker to be brought under missile attack from both lines of the walls. Outside the outer wall there will be a moat, often filled with water, and fields outside the moat will be cleared of any growth or structures in order to prevent the enemy from sneaking up on the castle.
Towers can be jutting out of the wall, or in line with it; they may also be square, hexagonal, octagonal or round. Generally, more points is the better (up to infinity, i.e. circle), because shape that is closer to the circle is more resilient to being undermined. It is however more difficult to build. Likewise, towers that are jutting out of the wall are superior, as they remove blind spots the attacker might use. Extending them onto the wall itself can be used to compartmentalize the wall, requiring the attacker to take control of one section at the time. Towers may be part of the wall itself, but superior choice is for them to be independent structures that merely lean against the wall as this enhances integrity of the main wall.
Smaller towers were used by lookouts, and could be built as part of the larger tower.
Walls are the primary defense of a castle. Concentric castles had two lines of curtain walls, one within another. Curtain wall was thick enough (1,8 – 2 meters) to stop most projectiles.
Talus is a sloping face at the base of a fortified wall (easily seen in Krak des Chevaliers). It makes wall more resilient against traditional siege equipment, and also makes it more difficult – or even impossible – for ladders and siege towers to reach the top of the wall. Most walls however did not have this feature, relying instead on moats.
Machicolations – overhanging holes – were built on top of the wall, allowing defenders to attack enemies right up the castle walls without exposing themselves.
A moat is a ditch surrounding the castle (or city, as in Constantinople) and filled with water. A moat helps prevent access to the walls – getting siege tower across is hard to impossible – but its main purpose is to prevent the attacker from undermining the walls. Tunneling was dangerous to the castle or city, as it could at best provide access to the enemy, but usual purpose was to collapse a section of the walls. A tunnel under the moat however would collapse and fill with water – assuming that it didn’t end up in the moat in the first place.
Water in the moat would be rather disgusting, as it was stagnant and filled with waste produced by the castle (including toilets). As it prevented the access, moat also made it easier for archers and crossbowmen to pick off attacking troops.
Main gate may be the most vulnerable point in castle walls, but it is also the main death trap in the castle. As early as Roman times, gatehouses would consist of an outer gate, courtyard and inner gate. Anyone who penetrated the outer gate would come under the attack from the terraces surrounding the courtyard before the inner gate. In medieval castles, the courtyard was often roofed, so that defenders could literally stand above the attackers and pour boiling water, excretions and other fun stuff straight onto attackers’ heads through murder holes (holes in the ceiling). Arrow loops could and often would be built into the sides, creating a three-way crossfire.
To go over the moat, a drawbridge would be used. This could vary in design – the classical descending drawbridge is a late medieval design, and earlier designs could range from an unsecured long piece of wood to horizontally rotating bridge. Behind the drawbridge – and possibly the first gate – would often be the portcullis (iron grille) which could be lowered, either to reinforce the gate or to trap the enemies which had succeeded in breaching the outer gate.
Barbican was an extension which jutted out of the gatehouse, and added another layer of portcullises and murder-holes onto the preexisting gatehouse.
Stairwells inside the castle – particularly those inside the towers – often curved very narrowly in clockwise direction. This ensured that attackers – most of whom would be right-handed – would have their right hand against interior curve of the wall, making it difficult for them to swing their swords.
Often included were hidden postern gates and secret passages, allowing defenders significant mobility advantage over any attackers unfamiliar with castle’s layout. Postern gates were also included in cities, sometimes as sally ports and sometimes – in case of cities with more than one circuit of walls – for easy communication between the inner and the outer wall circuit.
If it is the city that is fortified and not the castle, and sometimes even a castle, it will typically have a harbour – either on the sea or on the river. A harbour is an obvious weak spot in defences, as it has to handle significant amounts of traffic. Because of this, entrance to the harbour will be designed so that it can be blocked by a chain held up by floating barrels.
Any castle or a fort must have water supply not dependent on going outside. An ideal solution is if it can be built over an existing well. If that is not possible, it will have a large number of wells, and possibly cisterns as well, both of which will be filled by rainfall.
- Povijest Ljudskog Roda – Srednjovjekovni Dvorci
- Kraljevski husar
About the Author:
Toni Šušnjar is an amateur historian and fantasy enthusiast with particular interest in Ancient and Medieval history as well as Medieval and High Fantasy. He also writes Military Fantasy blog. You can follow him on the Military Fantasy Facebook page.