I joined in the discussion, which was a mix of the good, the bad, and the ugly (mostly good though).
Eventually, I posted the following:
IF WE GRANT THAT
G: There are occasions in which opting for the use of torture (on, say, a captured terrorist) will increase the chances of acquiring information – specifically in cases in which such information possibly would prevent moral disaster (e.g., the detonation of a nuclear bomb in a city)—then (and I take this to be Shri’s core insight) it is morally odd (and by no means incontrovertably wise) to adopt an absolute prohibition/condemnation against the use of torture, for such prohibition would seem to allow moral disasters—situations even more regrettable (from a moral perspective) than the instance of torture.
Philosophers have long known that, in contemplating extreme circumstances—as actually arise in, say, the setting of national leadership—paradoxes (I use the term somewhat loosely) emerge. Thus, for instance, Gregory Kavka* once argued that, to bring about the morally best outcome, it may be necessary for some individuals within a society to become morally corrupt—in order to act under special circumstances in a manner in which no decent moral being would act (namely, by retaliating against a nation’s nuclear onslaught with a reciprocating [and pointless] nuclear onslaught, for the sake of effective deterrence).
Thus, paradoxically, a utilitarian would not seek that everyone be a utilitarian.
Many years ago, some philosophers began considering morality and moral principles from two perspectives: from the individual’s perspective and the perspective of a moral being who has the opportunity to decide on the rules and practices that would be adopted by everyone in society. Arguably, one would be in a much better position to maximize happiness (or minimize violence, pain, etc.—or, indeed, to achieve any overarching goal) if one had the latter perspective and could somehow enforce it or cause individuals to act accordingly. (Example of such theorizing: John Rawls’ mid-50s essay, “Two Concepts of Rules.”**)
It seems clear that no society can flourish in which free and informed members are permitted acts of violence (including torture) as means to their goals, however noble. On the other hand, arguably, a community can together reflect on that general perspective mentioned above and in this way see the wisdom or desirability of permitting torture for cases such as G above.
In a sense, we already approach some matters in this fashion—or at least some of us individuals do. I would argue that no large society (perhaps there should be no large societies!) can survive (and thus flourish) without a military and a substantial army. Further, no large army could function if it allowed its soldiers to exercise moral autonomy. And in fact, actual armies (certainly ours) operate in a manner that discourages autonomy among individual soldiers. (Unsurprisingly, we are not very honest with ourselves about this.)
I have generally refrained from applying my usual demand (of persons) that they exercise moral autonomy in the case of soldiers (and other classifications) exactly because of this recognition that, as a matter of practical fact, no military can function if it encourages precisely the sort of character that, normally, we hope to instill in our children.
So my perspective (here) parallels Kavka’s.
But I don’t see how a large army could function (especially during time of war) were that autonomy to be permitted or encouraged for soldiers qua soldiers. And so one confronts a kind of dilemma.
And, for me, the dilemma is mitigated or alleviated by that higher perspective that sees the necessity of soldier non-autonomy relative to our society’s gravest desiderata. That is, that perspective is in some sense compelling, from a moral perspective. It is as compelling, perhaps, as the need for the condemnation of torture.
I offer the above with no slight tentativeness.
*Some Paradoxes of Deterrence (Gregory Kavka, 1978)
**Two Concepts of Rules (John Rawls, 1955)