Flying Castles

Castle in the Sky
Castle in the Sky (Photo credit: Wikipedia)

Who can resist a stone confection floating in the sky? The concept of a structure in the clouds is ancient, Jack in the Beanstalk being a more recent variation of an old theme. “Building castles in the sky” is a hoary expression for daydreaming which goes back, according to some to Don Quixote’s battle against the windmill/giant. In my own fantasy experience, it has long been a staple of my D&D campaigns. Many decades ago I even created a floor plan of an elaborate flying castle, one I’m using as a major place in my current project. My favorite Studio Ghibli film is Castle in the Sky.

Typically, flying castles are depicted much as a terrestrial castle but with many more towers and turrets. There’s nothing wrong with that, I love the imagery myself, but would a flying castle really look like that? Here I’ll poke a bit into flying castles, not to denigrate the common image but to toss out some other conceptions.

One staple of flying castles are battlements: crenellations, machiolations, turrets, barbettes, and all sort of related castle bits. Question is, do these make sense on a flying castle? These are all intended to provide a point defense against attack by ground forces, often ones at the foot of a wall. For starters, most flying castles seem to have a sheer cliff at the foot of the wall, thus providing no place for ground troops to stand and, therefore no reason for machiolations for dropping stones on to them. But step back a bit, who is going to attack a flying castle on foot anyway? You certainly can’t walk to one. To attack a flying castle, one needs to fly and flying over the wall at the castle is a bit of a trivial step having flown all the way to the castle.

That raises the next point- why have walls at all? There are similar problems with conventional towers (usually open at the top) and even crenellations: they offer protection in only one direction, outward facing, and not from above or behind.

My current book has flying mounts and floating castles. But how to establish that without losing the reader?
Note the battered (sloped) walls: a flying fortress that moves might be more stable with sloping walls.

So what might a flying castle look like? One might still have walls to contain and compartmentalize attackers, once they have landed. But since the attackers probably didn’t fly in with siege towers or even scaling ladders, the walls need not be very high or very thick.  In this case, internal walls would be just as important as peripheral walls: the place might look like a honeycomb. And instead of crenellations and hordings built to face an attack from the ground and outside the wall, one could imagine crenellations that arched over the defender. Perhaps this might be a tube with arrow slits in sides and top, although that would severely limit visibility for those inside. But it could also be crenallations that had a front and top, and were set on both sides of the wall (like interdigitated but untouching fingers of two hands). Now your defender can duck behind something that provides front and top protection when he needs it but can also poke out to see around. If he needs protection from an attack from the back, he could use the crenellations on the opposite side of the wall. Of course, you could dispense with walls entirely and imagine instead a castle composed more as a series of connected blockhouses.

What to do about roofs? A flying attacker can land on a roof and either attack from that vantage point with missile weapons or break through the tiles (or what-ever serves for a covering) to make a breach. From there they can either gain entrance to the castle or just set it on fire. How to defend the roof? You could imagine spikes to keep flying creatures off the roof or some other magical defense (oil of slipperiness for instance 🙂 ). You could also imagine roof-line sally ports for defenders to shoo off attackers and multiple towers of the same height sited so that defenders in one tower can attach someone on the roof of the other.

One likely defense for such castles, although a defense with surprisingly high upkeep: netting. String nets between your walls and over the tops of your towers and it won’t be easy for a flyer to land unless he can remove the netting (dragon fire might prove handy for that.) Conventional netting, though, would break down fairly quickly high up in the air with its bright sunlight and plentiful moisture. Perhaps more exotic materials could be used or maybe the nets are only run out when attacked (although stringing out lots of netting could take a very large and skilled crew: it took many hundreds of sailors to run out the awnings in the Imperial Roman Coliseum). Or maybe the netting is actually some form of magic. In D&D terms, low-level wizards with wands of web or some other stringy magical force, not necessarily sticky.

But in the end, why have any above constructs at all? Maybe the castle is built into the floating rock itself. Now you don’t have roofs, you just have arrow slit galleries, sally ports and the like. In the end, if you have a flying mount, you can probably carry a catapult stone high above the flying fortress and just drop it on the castle. It’s pretty hard to build a roof that could resist such an attack.

Or maybe the defense isn’t the stonework. Fortifications are all passive defenses. There are active defenses. Like a modern aircraft carrier, perhaps the real protection for the castle are the aerial creatures it carries. A few fire-breathing dragons might make stonework moot. If not dragons, griffin riders or other aerial mounts could serve. Such a castle would have to have a place for their aerial defenders: hangers or caves, large sorti ports, places to land, barracks for riders and craftsmen to maintain harnesses. There might even be huge cattle pens for food for the mounts if they could not feed off the land below. Start adding in structures for aerial mounts and  your flying fortress begins to look very different from a simple Middle Ages fort tossed into the sky.

Of course, there are reasons for having more conventional castles, even in the air. Whimsy or awe might cause a castle builder to dress his flying fortress with unnecessary defenses. These may serve no other function than to please the builder’s fancy or frighten a neighboring lord. Beyond that, a castle that is now in the sky may have been built on the ground and later lofted into the sky (I do that myself in my current story where the flying castles are very much conventional in design.) Finally, there aren’t that many readers who are going to quibble with your flying castle if it has stonework defenses that are unnecessary or useless for aerial combat. Sometimes a flying castle is just a cool flying castle.

The Medieval Fortress: Castle Encyclopedia for Writers

Cover of "The Medieval Fortress : Castles...
Cover via Amazon: The Medieval Fortress

Want a book that explains the parts of a fortress with copious line drawings, provides a very wide range of castle pictures and floor-plans, and perhaps most importantly for the writer, lots of info on the scale, purpose and history of different castles? The Medieval Fortress: Castles, Forts and Walled Cities of the Middle Ages by the J. E. Kaufmann and H. W. Kaufmann may be the book for you.

There are many fine books on castles and depending on what you are looking for, this may or may not be your best choice. Osprey’s gorgeously illustrated Norman Stone Castles (1) The British Isles 1066-1216, for instance, is a lot more readable albeit focused on a particular time and place. Life in a Medieval Castle by Joseph and France Gies is a wonderfully readable introduction to castles with attention to the people who really made such a place function (cooks, stable-hands, and so on). There are many full color books with better photographs and floor-plans.

But for a survey of castles in all their myriad forms, for floor-plans of the famous and obscure, for a sense of the scale of castles, not just the outer dimensions you can read from any floor-plan but thickness of walls, arrangements of battlements, gatehouses and drawbridges, this book is hard to beat. I’m not really sure who the authors. I suspect they are “just amateur” historians with a love for castles. I put that in quotes because there is nothing wrong with dedicated amateur historians. They typically write better books than the professionals who get too hung up on impressing other academics at the expense of dreadfully dull prose and historical interpretations that are feeble and caveat’d to uselessness.

The illustrations are all black & white, which is fine for the line art (very nicely done), and not so fine for the castle photographs. The book is divided into an overview of castles and their components followed by a survey of historical castles by region. Even as a history nut with a special thing for the middle ages and a love of castles, this book showed me castles I never knew of, castle features I never considered, and provided dimensions I was eager to have. Where it really shines are the drawings and explanatory text of castle components, from the 5 different wells and cisterns, to the four diagrams of various machicouli (machicolations) that clearly show size and how they work, the eight different ways to make a drawbridge, to more arrow-loop styles than you could imagine. Looking for an architectural flourish or wondering what that odd bit of a castle is for? Then this encyclopedic book is for you.

But it isn’t for everyone. The text is a bit on the dry side and the descriptions of the castles are cursory. I checked the Amazon reviews expecting to find they were mediocre, prepared to plug a book that while not perfect has its merits. Instead, I found almost all 5 star reviews. Curiously, the edition listed there was published in 2004 and my copy is 2001 with, from what I can tell, identical text but a different cover. I have a suspicion that most of the reviews are ‘plants’ by friends of the authors. So, don’t expect a 5-star book, but buy it or check it out of the library if you want to understand castles and all their shapes and pieces.

Here are a few tidbits from the book that might be of interest if you are crafting your own castle. Measurements in the book are in meters but I’ve converted back to English.


A section of Rome's Servian Wall at Termini ra...

One thing that usually irritates me in these books is a lack of generally useful dimensions. How tall and thick was the average wall? This book gives you both generalities and specifics. The typical Roman wall was 4 times as tall as it was wide and normally at least 9 feet tall. Thus a 40 foot high wall might be 10 feet thick, with walls tending to get thicker in the late Middle Ages (probably as a reaction to better siege weapons including cannons). Some specific wall dimensions: Rome’s Servian walls were 13 feet thick and 21 feet tall (stubbier than typical but perhaps to provide a road atop the wall around the city). Actual wall thickness varied with need (thinner where the fortification was less vulnerable, for instance). Also, unless very thin, they were general faced in cut stone with rubble fill, not cut stone all the way through.


Keeps varied a bit in size, with a typical one 65 to 100 feet tall. Exceptional ones were as high as 120 feet with walls usually 5 to 7 feet thick but as much as 13 feet thick or even thicker. Moats could be 65 across and 35 feet deep: we’re not talking about a little ol’ ditch here. In many cases, the moats were stone faced and where possible, served as the quarry for the castle.


As an example of the treatment you can expect from this book, the discussion of towers not only shows illustrations of a range of towers, square, circular, half circle, octagon and more, it explains why the different forms evolved (mostly has to do with protection against sapping and fields of fire). Ever wonder what an architect did if he was concerned about the enemy seizing a tower and using it against the defenders? This book shows you with diagrams of open-backed towers that provided no protection towards the inside of the keep.


Ever wonder how the towers and keeps were capped? In modern fanciful reconstructions, you see soaring, conical peaks, were those really used? Yes (as you can also tell from Medieval illustrations like the books of hours) but this book will give you several cross-sections, of not just the roof but the stone-vaulted intermediate floors.