Combatant identification

Why do armies dress their soldiers to be seen when staying hidden would save lives?

The question contains a hidden assumption worth dislodging first. Modern uniforms are camouflaged; no army dresses its soldiers to be conspicuous at eight hundred metres. So what is being bought is not visibility to the enemy's eye. It is something narrower: that when a soldier is seen, he is identifiable as a member of an armed force rather than as one of the people among whom he fights. Why would a body of law build itself around that?

Begin not with the soldier but with the person about to shoot him. If force may lawfully be directed at combatants and not at civilians, someone must be able to tell which is which, in poor light, at speed, from a distance. Nothing about a human body announces its legal category, so the category has to be carried by something the person does or wears.

That is exactly how the law constructs it. The old Hague Regulations, carried into the Third Geneva Convention's definition of who qualifies as a prisoner of war, set four conditions for militias and volunteer corps: command by a person responsible for subordinates, a fixed distinctive sign recognisable at a distance, arms carried openly, and operations conducted according to the laws and customs of war. Regular armed forces are simply presumed to satisfy them. Three of the four are about conduct and organisation; only one is about appearance, and even that asks for a sign, not for conspicuousness.

Now ask what the soldier gets for accepting it. Two things, and they are large. He acquires the combatant's privilege: he may kill enemy soldiers and not be tried for murder afterwards. And if captured he is a prisoner of war, held until the end of hostilities rather than prosecuted. A fighter who meets none of the conditions is not automatically a criminal for that alone, but he keeps none of this either — he can be tried under ordinary law for the very acts a uniformed soldier commits with impunity.

So it is a bargain, and a bargain has a defector's option. What does a defender gain by abandoning the sign? Surprise, survivability, and the enemy's paralysis. The cost lands on the civilians he is now indistinguishable from, and on the attacker who must decide without the information the rule was supposed to supply.

Here is the part that makes the whole structure work, and it is the uncomfortable part. The law does not respond to that defection by relaxing the attacker's duty. Quite the opposite: in case of doubt whether a person is a civilian, that person is to be treated as a civilian, and civilians lose their protection only for such time as they take a direct part in hostilities. The cost of ambiguity stays with the attacker's judgement, no matter who created the ambiguity.

Do you see why that forces a second rule? If distinction can be defeated by one side while the burden remains on the other, then non-compliance is not merely unrewarded — it is profitable. So the specific exploitation has to be prohibited in its own right. That prohibition is perfidy: killing, injuring or capturing an adversary by inviting his confidence that you are entitled to protection, and then betraying it — feigning civilian status, surrender, incapacitation by wounds, or the protection of the red cross. Note what is not prohibited: ruses of war, from camouflage to dummy positions to misleading radio traffic, are lawful, because they deceive the enemy's eye rather than his trust in a legal protection.

Three refinements hold this together.

First, the rule is about identifiability, not exposure. "Recognisable at a distance" means recognisable when observed, which is why camouflage is entirely lawful and why the answer to the opening question is that armies do not, in fact, dress to be seen.

Second, the categories track conduct and status rather than appearance alone. That is why the hardest modern cases are about behaviour: a civilian who takes a direct part in hostilities may be attacked while doing so, and regains protection when it ends. Where exactly that boundary sits is genuinely contested — the ICRC's interpretive guidance on direct participation is disputed by states and scholars, and I am simplifying a live argument.

Third, this is not a claim that irregular fighters are villains and uniformed ones virtuous. The First Additional Protocol of 1977 loosened the requirement where the nature of hostilities prevents a fighter from distinguishing himself, asking at minimum that he carry arms openly during each engagement and while visible in the deployment preceding an attack. That concession is one of the most argued-over provisions in the whole body of law, and part of why several major states have not ratified the Protocol.

• Where the line between a lawful ruse and perfidy falls in practice, and why wearing the enemy's uniform sits so awkwardly on it.
• What "for such time as" means for someone who fights intermittently.

Every term the collection defines is gathered in the glossary.