[Analemma_]

I think the bigger point from the article is that by the time people are fighting hand-to-hand in the tower stairwells, the defenders have already well and truly lost: comeback from such a state was probably impossible (and if the walls were breached the defenders would almost certainly have surrendered rather than fought to the last man). So it wouldn't have really made sense to design things for this possibility.

[ramblenode]

That's not a good counter-explanation.

A castle staircase takes a lot of time and effort to build. Choosing to build the staircase in one direction or the other has negligible cost. If there is even a slight or possible advantage to one direction then it would make sense to build it that way.

Defense in depth.

[prewett]

Any defenders defending a tower are obviously above the ground floor, which is where access to food and water is. So why bother fighting up the stairway, when you can just block all the downstairs exits? The castle's defenses are the walls; if attackers are in a position to go up the stairs then the castle's defenses have failed, and the only defense left is the manpower of the defenders. So instead of being "besieged" up their towers, the only realistic strategy the defenders have is to come down from the towers and join the melee. Or just surrender, because the attackers have an army and the castles only had dozens of defenders (if that). What tactical situation do you have in mind where the success of the attack depends on success in a staircase battle?

[ramblenode]

Defense in depth isn's about justifying in advance how every measure will win the battle; it's about giving yourself as many small, incremental advantages as possible so that the odds steadily tick up in your favor. Battles are famously difficult to predict so every advantage is sought, and even small advantages can have multiplier effects.

[wongarsu]

Every day the attackers besiege you is one they have to defend against potential counterattacks from your allies. And even if your castle falls you might buy your empire time to raise a bigger army and rally more allies in order to win the next bigger war. Delaying enemies could be an important function of castles.

[Tao3300]

It's a really good explanation. Castle sieges were big events, so historically we know the outcomes. Nearly 100% of the time, the garrison has already surrendered if it's this bad. Medieval sieges come in three major flavors: ones where you sneak in, ones where you bombard the fortification, and ones where you don't let anything in or out and you wait until they give up.

[ramblenode]

If you examine a military you will find volumes of plans for incredibly unlikely situations. Once you have addressed all the likely and significant threats, you don't just stop planning--at least not any good military.

Saying that castle sieges didn't tend to involve stairway fights doesn't imply that stairways wouldn't have had defensive measures built in. That is post-hoc rationalization.

[221qqwe]

> That is post-hoc rationalization.

You're asking to a disprove a purely(?) speculative claim, though.

[the_af]

> If there is even a slight or possible advantage to one direction then it would make sense to build it that way

If it's true that the battle at this point is lost for the defenders -- and known history indicates this is so -- then why would the builders choose directions based on this extremely unlikely scenario, instead of on just about any other consideration (aesthetic, practical, or even random)?

[lazide]

Why does the USMC issue bayonets still?

[qwytw]

Because it's a multi-purpose utility knife that can also optionally be mounted on a rifle?

Having a knife might be useful in various situations even outside combat as far as I know.

[the_af]

I've honestly no idea.

I'll take a stab at guessing (mind you, this is blind guessing, happy to be corrected!): the USMC still issue bayonets because of both tradition, which is important to the military, and also because they are actually useful in close quarters battle, which still occurs on occasion, such as in urban warfare and house-to-house combat clearing, etc. The likelihood of having to use a bayonet/knife in modern CQB is probably significantly higher than the likelihood of medieval defenders recovering from an enemy army that has stormed their castle.

[the_af]

Someone apparently disagreed with me but provided no additional explanation. Oh well, such is life... on HN!

[CydeWeys]

An interstate takes a lot of time and effort to build. And yet which side you drive on doesn't matter; it simply needs to be consistent with all the other roads you're connected to. There are plenty of countries that, through historical happenstance, drive on the opposite side of the road, and it's fine.

So in other words, just because the staircases take a lot of time and effort to build, simply means that having the staircase itself is important, not necessarily that its chirality is important. It has to have a chirality but it may well not matter which one, just like roads.

[ramblenode]

Your example misses the point.

If there were evidence that driving on the right side or left side of the road slightly reduces car accidents and a country with previously no roads or cars began planning to automotize the country, then, all things considered, it would make sense to have people drive in the lane with a slightly reduced fatality rate.

If there are two choices where one presents a slight advantage but no additional cost then a rational actor will go with that choice.

[qwytw]

The event has to occur relatively frequently for that slight advantage to become statistically noticeable. Direct assaults on castles with hand-to-hand combat occurring in stairwells were extremely rare as far as we know.

[notahacker]

Yeah.

For perspective, Norman keeps were often built with a large internal cross wall, so even if troops made it through the stair door and swarmed into the room they'd still have to fight their way into the other half of the floor. By the stage these expensive and space consuming walls were defensively relevant, defenders would have already lost outer walls, viable long-term food and water supplies and much of the garrison defending it... and any real chance of holding out. But an invading army would still lose more men storming it; so it functioned as a deterrent.

I've heard this "it's a myth" argument before, but 70% of staircases is quite a large proportion of staircases spiralling in a particular direction which would offer the defender a marginal advantage to be pure coincidence. Particularly when the ratio of clockwise to anticlockwise staircases in Norman castles was about 20:1; it was later generations of castle of builders who added many more anticlockwise stairwells, in an era when individual tower defence was less importance, and builders may have simply forgotten or come to doubt arguments about the defensive advantages of clockwise spirals (the blog's arguments for why spiral staircase defence is rubbish work here of course!). Contemporary cathedrals which were not at all defensible tended to build clockwise and anticlockwise spiral staircases as matching pairs, so it wasn't like there was some other sort of massive aversion to stairs in a particular direction.

[gowld]

Where's the contemporary research to determine which direction has advantage?

[Cass]

That's a bizarre line of thought to me. If you build an expensive structure for fortification, you don't usually get to the interior design and then go "Oh fuck it, this extra safety measure wouldn't cost anything, but if they've got this far we might as well surrender, so let's not bother."

Going by that logic, the president's bunker under the pentagon would've been built without a lock. After all, people don't usually have to physically drag a country's leader out of their locked bunker, right? By the time anyone's knocking on that door, usually the war is lost and the country has surrendered.

And yet, if you're designing for defense, why NOT take such a cheap and easy countermeasure as putting a lock on the door or choosing the more defensible way to spiral your staircase? You might want to buy a few more minutes to negotiate in a desperate situation; you might want at least the option of taking that futile last stand; you might be facing not an invading army but a single lunatic with a sword who snuck past the outer guards.

[whiw]

I imagine that it would be more difficult to gain entry to an upper floor (at the top of a narrow staircase so single-file attackers) and a sturdy door with a couple of guards outside, than it would be to gain entry to rooms on the same level. Perhaps the women were tucked away on the upper floors, in relative safety.

[jvanderbot]

This comment is a perfect example of GGP's problem with historical analysis.

As is the nearly-exactly comment in TFA:

> Frankly, if you find yourself in this position the castle is probably already lost.

It's conjecture all the way down.

[qwytw]

> It's conjecture

Just like the entire staircase argument?

Except that we actually know that:

> Frankly, if you find yourself in this position the castle is probably already lost.

Was true in almost all cases.