I think, Derek, that we need to assume certain moral obligations if we are going to accept morality at all, similar to the understanding that if we live under a particular law, we have a certain obligation to live accordingly.

My point is that if we are in a position to "break the moral tie-score", that there are certain rules we can follow that give us something to work with. One of those rules is one of general utility. While not my first choice for an ethical system, it at least provides us with tools to use in morally ambiguous situations, especially when discussed academically.

So, in the case of utility, there is a sort of "balance sheet" that we can work with. It is one of many systems, but I think one that works well for this particular question. One that would be much more difficult under a duty system or virtue system.

With that premise as part of my assumption (the premise that utility is the best system for answering this question and that the question is that we could possibly reduce the amount of suffering in the world through some personal sacrifice), then by offering ourselves up as the subject of torture, we take on a burden by choice that someone else would have to suffer presumably unwillingly. Since the alternative is to die (an entry of "0" on our balance sheet, assuming it was a deserved death or at least unavoidable), then we have a sum total positive on the moral balance sheet and the "greater good" consequence is achieved in some small way.

Without a rule of some sort, then all our morally ambiguous decisions are equal, and the choice becomes arbitrary. While this is perfectly possible, it does seem to stall the conversation unless offered as the ultimate conclusion (that all moral decisions are ambiguous and therefore in error)