I think that in cases of horrific crimes, the death penalty is acceptable, or even required by retributive justice. However, I think this only applies to cases where there is absolutely no room for doubt. I also think that there really are such cases where there is 100% certainty e.g. the perpetrator was seen by many witnesses and confesses, plus as much additional evidence as you need. Unfortunately, if we only make convictions where we have the luxury of this certainty, we set the bar too high, and many guilty people escape conviction. Inevitably, under any reasonable judicial system there will be people charged for crimes they didn’t commit. But when you are charged with a crime, you are thereby unequivocally guilty, and there’s no way of charging someone with being guilty with the qualification, “he might not have done it” and another “he’s guilty of the crime and there’s no doubt”. In the eyes of the law, a guilty verdict is definitive; you did it, end of story. Is there a problem with this either/or approach nature of being charged? Shouldn’t the law recognise that someone is viewed as guilty on the basis of a balance of probability and not, or not always, because the truth has been uncovered as to guilt and innocence? A related problem is that maybe the situations in which there’s 100% certainty are those in which “everyone knows who did it” but there’s no code we can lay down to specify when those situations will occur, you just know it when you see it, kind of thing. But it seems bizarre that we can all clearly recognise cases in which the guilt of a person is not in question, but not be able to use capital punishment just in those cases without risking innocent people being treated in this way.

You make two good points. On the first, re death penalty, I would agree that there are cases where it's crystal clear that the accused is guilty. But this is really beside the point. The question is whether we can design a mechanism that correctly identifies these cases. In the absence of such a mechanism, we must be especially reluctant to use the death penalty.

Your other point is that perhaps we should incorporate into criminal verdicts an assessment (by the judge or jury) of the degree of certainty. But again, there is the question how accurate this assessment would be. And there is the further point that it would look quite bad to impose a severe punishment on someone with the comment that we've just barely reached the minimally required level of certainty. Better then perhaps to handle this issue unofficially: just as jurors sometimes acquit someone who very clearly did what the law proscribes when they feel the person did nothing wrong (example: the killing of a suffering and terminally ill spouse), a judge may sometimes impose a lesser penalty because she is less than fully convinced that the person really committed the crime.

And a possibly interesting footnote on this second point. In the case of acquittals, as well, one could incorporate an assessment (by the judge or the jury) of the degree of certainty. And this is actually done, I believe, in Scottish jurisprudence, which allows -- aside from "guilty" and "innocent" -- a third verdict of "not proven." The accused goes free, but has not been fully cleared of the charges brought against her or him.

Read another response by Thomas Pogge
Read another response about Law, Punishment