In his online column at the New York Times, Stanley Fish distinguishes between what's right and what's true. Procedures yield the former by bracketing the latter. This often leaves us unsatisfied. Sometimes it also leads to absurdity:

[L]ast year I found myself talking to an insurance company about a claim I was making. The person on the other end asked me a series of questions (like when did the damage occur) and I answered them truthfully. She took pity on me and told me that those were not the answers I should be giving. I escaped from the conversation and called a private adjuster, someone who represents claimants and knows exactly what the right answers are. He told me that I should never speak to my insurance company, but instead refer everything to him. He was afraid I might say something true.The right answer is the answer a system invested in its own machinery will recognize no matter what the true facts may be.[...] This is almost always the case in the law, especially in a legal system like ours that privileges procedure over substance. Lawyers know that what they have to do is find the legal rubric that will allow them to frame an issue in such a way that when the systems questions are posed, the right answer, not the true answer, will be generated. Courts sometimes explicitly announce that the procedurally correct answer is preferable to the true answer, which is, legally, of no interest at all.

And sometimes the consequences are much worse than absurd. Fish cites Herrera v. Collins, a Supreme Court case in which a man found guilty of murder argued that his conviction ought to be overturned because of new evidence that proved his innocence. Writing for the majority, Chief Justice William Rehnquist claimed that it was the Court's job to decide not whether Herrera was guilty, but only whether his original trial had been a fair one—that is, procedurally correct. The Supreme Court ruled that it had been, and Herrera was executed. (Just like Fish, by the way, to put this last fact between parentheses. He is a prose proceduralist.) This is the kind of absurdity most people would call an outrage, but it isn't as easy to avoid as most people might suppose. Fish writes:

Is this bad? Should we go off the right standard and return to the true standard? A nice idea, but one that imagines a world where the judgments reached by systems are tested against a truth that is independent of any system. Where would that truth come from, how would it be identified and how could the endless disputes about what it is be resolved? (The laws project is to hold such disputes at bay.) It is because there are no answers to these questions that we will have to settle for the truths that systems create, deliver and validate in a sequence that may be reassuring but is finally without a foundation.

By "without a foundation," Fish means: without a foundation that can be demonstrated to everyone's satisfaction. Is this the only kind worthy of the name?