It rests on a prediction that cannot be tested, and so can justify anything

The argument

Opponents of preventive war argue that its justification has a fatal structure: it rests on a claim about what another state would have done, which is precisely the thing the war makes unknowable. If the attack never comes, the preventive action is declared a success; if intelligence about the threat proves wrong, the war has already happened. No outcome can disconfirm the original claim, which means the standard is not a standard at all but a form of words available to any government that wishes to attack. Opponents argue that this is not a theoretical concern. Assessments of another state's intentions are produced by agencies under political direction, in secrecy, and are routinely wrong even in good faith; threats are overestimated, ambiguous evidence is read in the direction leadership prefers, and the pressure runs one way, since no official is punished for having been too cautious about an enemy. A doctrine that licenses war on the strength of such assessments hands the decision to the least reliable and least reviewable part of government. The consequence, opponents argue, is that permitting preventive war in principle permits aggression in practice. Every state that has attacked another has described the attack as forestalling a danger, and a rule that accepts this justification cannot distinguish the sincere case from the pretext — which means in practice it validates both. Because preventive war depends on an untestable prediction about an attack that has not occurred, and such predictions are both unreliable and impossible to review, this argument concludes, preventive war cannot be justified.

Premises

Counter-arguments

Proponents reply that this is an argument about evidence and safeguards rather than about the principle, and that it would equally forbid all policing and intelligence work, which also act on predictions. They argue that a rule permitting nothing before an attack does not prevent bad-faith wars, which proceed on pretexts regardless, but does disarm good-faith action in the cases where prevention would genuinely have worked and lives would have been saved.

Rejecting the premises

[Rejecting P1] Unfalsifiability after the event is not the same as groundlessness before it. A great deal of defensible action rests on judgements no outcome will confirm, including evacuation, quarantine and deterrence, and the premise gives no reason to treat prevention differently from these. [Rejecting P2] That assessments of intention are often wrong is an argument for a higher evidential threshold, not for an absolute bar. The premise moves from unreliability to impossibility without supplying the step that would connect them. [Rejecting C] The conclusion rules out every case, including one in which capability, stated intent and mobilisation are open and independently verifiable. An argument from the difficulty of justification establishes that justification is demanding, not that it can never be met.