Matt Yglesias asks, reasonably, why poisoned arrows (or bullets) are inherently worse than non-poisoned ones. (He’s replying to my post of yesterday commenting on Brad DeLong’s assessment of the contemporary relevance of some of the Homeric heroes.)
I think there’s actually a technical answer to Matt’s question. An arrow wound, poisoned or not, would usually have put the victim hors de combat; the additional risk of death or permanent disability created by the poison served no real military purpose other than terrifying the enemy. Thus the total severity-weighted casualty count would have been lower in a no-poison-arrows battle than in a battle where both sides used poison.
If that’s right, “nobody using poison” might have been Pareto-superior to “everybody using poison,” but there was a Prisoner’s Dilemma insofar as either side could get an advantage by using poison when its enemies didn’t. That’s the sort of situation where a strong convention might arise and be (imperfectly) observed and enforced.
But put that analysis to one side; assume that “not using poison” was a purely arbitrary convention with no substantive value. Even then, if “no poison” was the convention, then using poison was tantamount to what we now call “war crime.”
There is, it seems to me, an undeniable value in having war regarded as an activity defined in part by its limits; until that idea is in place, the task of defining the right limits is pointless. Homer’s Odysseus was doing what in him lay to make war even nastier than it had to be. So while I feel, as Brad DeLong does, that Odysseus is a much more comprehensible and sympathetic character than Achilles, Diomedes, Great Ajax, the rest of the homicidal maniacs who besieged Troy, and that Odysseus’s wiliness is more like your thinking and mine than, say, Hector’s somewhat ponderous nobility, I can’t take him, even provisionally, as a hero.