Sir

The scientific community should think twice before it turns its back on non-lethal weapons, as Malcolm Dando suggests in his Opinion piece 'Biologists napping while work militarized' (Nature 460, 950–951; 2009).

It is true that fentanyl killed scores of civilians when it was used to end the Moscow theatre seige in 2002. But that was partly down to Russia's desire to hide its use of calmative agents, which meant that no life-saving antidote was to hand: the Russians had not disclosed the agent's identity even to their medical personnel.

Dando rightly points out that the US Army misused BZ and other chemical agents during the Vietnam War. But the US Army's use of BZ no more serves as a model for fighting with non-lethal weapons than Vietnam now serves as a model for conducting asymmetric wars — conflicts in which the two sides have very different strategies or levels of military power.

Modern asymmetric war poses a stiff challenge to military organizations searching for the means to defeat guerrilla and insurgency forces while sparing the surrounding civilian population. Many innocent civilians are still being killed as armies turn to drones, real-time surveillance and precision-guided missiles. Could chemical agents, electromagnetic devices and other non-lethal weapons reduce the death toll of modern armed conflicts? We don't yet know the answer.

The use of non-lethal weapons need not cause excessive casualties. Any weapon system carries the spectre of abuse. The purpose of humanitarian law in war is to prevent the abuses that Dando describes and to restrain the carnage of modern war. Sometimes it is successful and sometimes not. But it would be strikingly ironic if disagreement in the scientific community left military organizations free to pursue their penchant for high-explosive weaponry without considering any of the available non-lethal alternatives.