Deterrence is a parent of paradox. Conflict theorists, notably Thomas Schelling, have pointed out several paradoxes of deterrence: that it may be to the advantage of someone who is trying to deter another to be irrational, to have fewer available options, or to lack relevant information.
Consider a typical situation involving deterrence. A potential wrongdoer is about to commit an offense that would unjustly harm someone. A defender intends, and threatens, to retaliate should the wrongdoer commit the offense. Carrying out retaliation, if the offense is committed, could well be morally wrong. (The wrongdoer could be insane, or the retaliation could be out of proportion with the offense, or could seriously harm others besides the wrongdoer.)
The moral paradoxes of deterrence arise out of the attempt to determine the moral status of the defender's intention to retaliate in such cases. If the defender knows retaliation to be wrong, it would appear that this intention is evil. Yet such "evil" intentions may pave the road to heaven, by preventing serious offenses and by doing so without actually harming anyone.
Scrutiny of such morally ambiguous retaliation intentions reveals paradoxes that call into question certain significant and widely accepted moral doctrines. These principles are what I call bridge principles. They attempt to link together the moral evaluation of actions and the moral evaluation of agents (and their states) in certain simple and apparently natural ways. The general acceptance, and intuitive appeal, of such principles, lends credibility to the project of constructing a consistent moral system that accurately reflects our firmest moral beliefs about both agents and actions.
For a system of morality to reflect our firmest and deepest convictions adequately, it must represent a middle ground between the extremes.
~ Gregory Kavka