I tend to oppose the death penalty for pragmatic more than principled reasons. (That is, if someone's done something sufficiently wicked that we see no problem locking him in a concrete cube with gangs of thugs for 50 years, it's not obvious why a death sentence is inherently beyond the pale. But as currently applied, there are clearly too many errors and inequities for it to be necessary when the person’s no longer a threat.)Note Julian’s distinction between principled and pragmatic reasons. I share Julian’s position on this issue, I’ve often stated it in the same way, and I’ve made a similar distinction between principle and practice on other issues. But I always wince internally when I do so, because the wording implies that feasibility, logistics, and other practical matters are somehow not the stuff of principle. Now, if your ethical principles derive entirely from a priori, deontological considerations, that makes sense. But for people who admit consequences matter at all in ethical decision-making – and you don’t have to be a utilitarian to think that, though it helps – pragmatic considerations will inform your selection of principles. You can’t arrive at principles, at least not terribly useful ones, without practical knowledge about how the world works, how perfectly or imperfectly rules will be enforced, whether the institutions you favor will in fact select the right rules, how prevalent are bias and abuse, etc.
Principles exist in layers, from the broadest and most abstract (e.g., “Maximize human happiness”) to the narrowest and most concrete (“Don’t let cousin Lou have more than two drinks”). In between, there are principles with an intermediate degree of abstraction (e.g., “Let people make decisions that are primarily self-regarding”). It is the intermediate principles that provide the most guidance, since the most abstract principles are fairly useless without more information (“what makes people happy?”), while the most concrete principles provide little guidance outside their narrow purviews (“is it okay for cousin Carol to have another drink?”). All but the most abstract principles implicitly incorporate information and factual judgments, usually in the form of generalizations about the world.
When principles are highly specific, relying on a great deal of context and relevant information, people are likely to regard them not as principles, but as pragmatic judgments. This is especially true when we wish to draw attention to the factually-contingent nature of a conclusion. Hence statements like Julian’s about the death penalty: by calling his judgment pragmatic, he draws attention to his willingness to revise his opinion in the face of a different set of facts. Still, there is a clearly a matter of principle at work here, something like, “Don’t vest untrustworthy agents with the power of life and death over others.” Other opponents of the death penalty rely on a similar argument, namely, that agents of the state cannot be trusted to administer the death penalty without bias – but then they conclude with a categorical condemnation of the death penalty under any circumstances. Apparently, they must consider their factual judgment about the trustworthiness of state agents to be more universal in character, perhaps attributable to immutable human nature rather than historical or institutional features of the current system. (Or they might be deploying the trustworthiness argument for persuasion’s sake, while basing their own categorical opposition on more general principles – but I’ve met at least some libertarians who oppose the death penalty categorically based on their general distrust of the state.)