Siege camp attrition

Why did a besieging army so often sicken and starve before the city it had surrounded?

The besieger holds every advantage. He is in the open air, free to move, with a whole countryside behind him and the choice of when to attack. The defenders are shut in, crowded, unable to harvest, cut off from everything.

So why is it so common to find the besieging army breaking up first, forced to withdraw from a city still standing? Something in the obvious reading is wrong, and the useful question is which of those advantages is not one.

Start with the one thing that is true of a besieger and of no other kind of army: he stays put. Everything else follows from that.

Take supply first. A moving army passes through fresh country every day and can eat some of it. A besieging army eats the country around it once, and then has to reach further out for the next load — so the effective haul distance grows with every week it sits there, in a widening ring of exhausted ground. Notice what that means: duration is silently converted into distance. This is the same limit its neighbour Military logistics describes as a supply radius: transport consumes what it carries, so past some distance a convoy arrives empty. The besieger meets it without going anywhere — time does the walking for him.

Now the sanitary side, which is usually the larger killer. Concentrate a great many people and animals in one place for months. They produce waste at a fixed daily rate and it accumulates, because nothing carries it away. The camp was sited for military reasons, not for drainage. So the water being drunk is very often water that the camp's own waste has reached, and the diseases that follow are exactly the ones that account describes: the dysenteries and enteric fevers that travel by contaminated water, and typhus, which travels by lice and so is a disease of crowding and of clothing that is never washed.

Now compare the city, and the asymmetry inverts the intuition we started with. It has sanitary infrastructure — wells, cisterns, drains — because people built it to live in permanently. It has food deliberately laid in, it knows how much, and it can ration against it. It can also cut the number of mouths by expelling those who cannot fight, a brutal move but an arithmetically decisive one. The besieger has none of that: no infrastructure, a supply that arrives or does not, and no idea how long he must last.

So the contest is not "who is more comfortable" but "which side's stock runs out first", and one side has planned its consumption while the other is improvising it.

Then the piece that makes it self-worsening rather than merely bad. Sickness reduces the number of effective men. Fewer effective men means the lines cannot be pressed, assaults cannot be mounted, and the siege takes longer. A longer siege means more time in the same fouled camp, drawing on a hinterland that is emptier still. That is a genuine reinforcing loop — the illness lengthens the siege and the length feeds the illness — and it is why these armies tend to decay slowly and then collapse rather than declining steadily.

Why is the loop not simply broken by leaving? Because the decision is not really the commander's. An abandoned siege is a visible, expensive defeat, the investment already made is enormous, and the political cost of walking away frequently exceeds the military cost of staying. The loop persists for institutional reasons, not because nobody could see it.

Here is the test. If the mechanism is duration multiplied by density in a fixed place, then losses should scale with time-in-place and with how tightly the camp is packed, not with the general hardship of the campaign. A force of the same size, in the same country and season, that kept moving should sicken markedly less. And armies that shifted their camps periodically, sited latrines below the water they drew, or loosened the investment into a distant blockade should show the effect weakening. The refuting observation is straightforward: besiegers falling sick at the same rate in their first fortnight as in their fourth month, or moving armies of equal size losing just as many.

Three refinements keep this honest.

First, "starve" is usually the wrong verb, and it matters for the mechanism. Outright famine in a besieging army is less common than a slide into a state where men are underfed, filthy, cold and therefore far more likely to fall ill, and where the sick recover badly. Hunger and disease are not two causes running in parallel; the first is largely a multiplier on the second.

Second, this was understood, and the countermeasures follow the mechanism precisely, which is itself evidence for it. Roman practice of building a properly laid-out fortified camp with sited latrines, the later habit of moving camps and rotating troops out of the lines, and the option of a distant blockade rather than a close investment are all attacks on the same two variables: density and duration. A distant blockade trades a slower result for a camp that is not sitting in its own accumulation.

Third, the numbers. I have deliberately given none. Returns from besieging armies are patchy, causes of death were recorded by people without a germ theory, and the surviving figures are frequently rhetorical rather than administrative — which is why the argument here rests on the structure and not on any body count. The relative contribution of contaminated water, of louse-borne fever and of undernourishment varies between sieges and cannot be settled from the sources, and retrospective diagnoses from period descriptions are contested among historians of medicine.

• Why the distant blockade eventually displaced the close investment, and what it gave up to do so.

Every term the collection defines is gathered in the glossary.