Obsolete fortifications

Why did walls that had protected cities for centuries stop working within a single lifetime?

A curtain wall that had turned away every attacker for three hundred years does not decay in fifty. The stone is the same stone, the masons no worse. Yet across the late fifteenth century towns that had been safe became indefensible, and everyone knew it. If the wall did not get worse, what exactly changed — and why was the answer that eventually worked a matter of shape rather than of stronger material?

Start by asking what a medieval wall was actually for. The threats were men on ladders, sappers at the base, and a ram at the gate. Against all three the decisive quantity is height: a wall you cannot get over, with defenders looking down on anyone who tries. Height is expensive in stone, so height is bought by keeping the wall thin. That is not an oversight — it is the correct trade for the attack of the day.

Now change the attack. Gunpowder artillery does not go over the wall. It applies a horizontal blow, repeatedly, at one spot near the base. Ask what the wall's two defining features do under that load. Thinness means the shot is not absorbed but transmitted through, and brittle mortared stone shatters rather than yielding. Height means the mass standing above the damaged course is leverage working against the wall, so once a bite is taken low down, the rest comes down of its own weight. Both of the wall's virtues have become liabilities, without a single property of the wall having changed. This is the whole point: obsolescence was not decay, it was the same object meeting a different question.

The change was not instant, and the emblem people reach for — Constantinople in 1453 — is less telling than the French siege train that went into Italy in 1494 and reduced fortress after fortress in days rather than seasons. What made that possible was less the guns' raw power than mobility and reliable enough carriages to bring them and aim them flat.

So what would you build instead? The tempting answer is more of the same: thicker, taller. Try it and see the problem. Taller only adds leverage and gives the gunner a bigger target standing clear against the sky. Thicker in stone is ruinously expensive and still brittle. So the useful moves are two, and neither is really about strength. First, get low and get soft: a rampart of packed earth, faced in masonry, that swallows a shot instead of transmitting it, and sits low enough that a besieger cannot see much of it to hit.

Second — and this is the real innovation — fix the geometry. A round tower's own curve creates a slice of ground at the foot of the wall that nobody on the wall can shoot into. Attackers had always known that: reach the base and you are safe to work. The answer developed in Italy through the early sixteenth century, the trace italienne, replaces round towers with angular bastions projecting like arrowheads, arranged so that the flank of each bastion looks straight down the face of its neighbour. Every stretch of wall is now covered by fire from somewhere else along the wall. The dead ground is not reduced — it is designed out.

Notice what kind of solution that is. Nothing about it is a better material. It is an arrangement of angles that converts the defenders' existing weapons into complete coverage: the fortification stopped losing not by resisting harder but by being reshaped so the attacker's new method had nowhere to work from.

Then follow the cost. Such works are enormous, they must enclose the whole circuit to be worth anything, and taking one requires an army large enough to surround it and stay for months. Charles Tilly's argument — that war made states as states made war — rests substantially on this: sieges of this kind demand sustained revenue, which demands a tax apparatus and the bureaucracy to run it, which is much of what a modern state is. It is worth saying plainly that the strong version of the claim, Geoffrey Parker's "military revolution", is contested; other historians read the changes as slower and more regionally uneven, and the arrow from fortification to state size is argued rather than settled.

The word "obsolete" hides the mechanism. A defence is never obsolete in itself, only against a specified attack, and it can become so overnight because the change happens on the attacker's side of the ledger. Call it a threshold rather than a decline: below a certain artillery capability the tall thin wall is optimal, and above it the same wall is worse than useless because it is expensive and it falls.

Two refinements are worth keeping. The bastion did not make cities safe again in the old sense — it made them slow. A well-designed fortress could be taken by anyone willing to pay for a long siege, and that shift from decision to attrition is what pushed the costs into the fiscal system. And the sequence repeated: rifled artillery made masonry forts obsolete in the nineteenth century, and the answer was again dispersal and geometry rather than a thicker wall.

• Why nineteenth-century masonry forts fell to rifled artillery, and what replaced them.
• How the cost of siege warfare shaped tax collection and standing armies in early modern Europe.

Every term the collection defines is gathered in the glossary.