What is the essence of proceduralism? I suggest: the criteria by which we judge that such-and-such is the case are constitutive of what it is for such-and-such to be the case. Or perhaps: the norms governing the validity of the 'output' of a procedure are identical to the procedural rules whereby fair are distinguished from unfair procedures. But I had better give an example, and pronto.
1. Lotteries. About (fair) lotteries we ought to be (pure) proceduralists. That is, we ought to hold that the procedure by which it is determined that a certain number is the winning number constitutes what it is to be the winning number. Of course I am assuming that the lottery procedure is fair: the selection process is random, and so on. So assume we have a fair lottery and that Wealthy Willy wins a pile of loot. Surely the following questions are senseless: Did Willy really win? Or was he merely judged to have won by the lottery officials? Isn't it possible that the lottery procedure, though fair and unbiased, selected the wrong winner? Isn't it possible that Impecunious Ike, or some other guy, is the real winner?
The point could be put as follows. With respect to lotteries, it is broadly-logically impossible that a person instantiate the property being judged to be a winner of lottery L and not instantiate the property being a winner of lottery L.
When it comes to lotteries, proceduralism is the only game in town.
2. Elections. The same, I think, goes for elections. If Snerdly gets more votes than the other candidates, he becomes president of the chess club. Surely it would be senseless to question whether Snerdly has really been elected, as if he could have an 'elective status' independently of the (fair) procedure by means of which he either wins the office or not.
In both the lottery and election cases one could of course maintain that there is something unfair, though not procedurally unfair, about the result. If Wealthy Willy, who needs no money, wins a million then that will be perceived as an unfair outcomes by the other players, all of them poor schmucks. But this is irrelevant to the point. It is also irrelevant to Snerdly's 'legitimacy' that he is not competent to be president if you assume that he was fairly elected.
3. Trials. Suppose a man is found guilty of a crime in a court of law in which all the proper procedures have been followed. Assume that the prosecution, the defense, the judge, and the jury are all highly competent and morally above reproach. The man has been found guilty of committing some crime; can one nevertheless reasonably ask whether he in fact committed the crime? In the case of Wealthy Willy, one cannot reasonably ask whether he in fact won: his having won just is his being-judged-to-have-won by the lottery officials.
If a sharp thinker insists that even in the Willy case there is a distinction between the two properties, I will cheerfully concede the point and congratulate the thinker on his intellectual acuity; but I will go on to insist that the distinction obtains only on the 'intensional plane' and not on the 'extensional plane.' For surely it is impossible that Willy or anyone instantiate the first property without instantiating the second.
So, extensionally, there is no distinction in the case of Willy between the property of having won and the property of having-been-judged-to-have-won. My question is whether, extensionally, there is a distinction between being guilty and being found guilty in a properly conducted court of law.
My answer is in the affirmative. This implies that there is at least a serious question as to whether proceduralism holds with respect to legal trials. A rational person must be a proceduralist when it comes to lotteries, but he needn't be when it comes to the law.
But I am prepared to make a stronger claim. Not only is there a serious question as to whether proceduralism holds with respect to legal trials, proceduralism does not and cannot hold with respect to them. For it is not perfectly obvious that people have been tried, convicted, and executed for crimes they did not commit? They were found guilty but they were not guilty in fact. All the procedures were properly followed and they were found guilty beyond a reasonable doubt. But they were not guilty in fact.
4. One might object to the foregoing as follows:
Look, we are agents, not transcendental spectators, and we have to come to intersubjectively binding and enforceable decisions in a timely fashion about matters of moment in conditions in which certain knowledge is rarely available. Sure, Jones is conceivably innocent despite having been found guilty in the eyes of the law. His guilt or innocence is not constituted by his being found guilty or innocent. But that is a merely theoretical consideration. For all practical purposes to be found guilty is to be guilty. After all, there is no other way we have of reliably determining guilt and innocence apart from the court sytem. We cannot call God on the phone. He is incommunicado, the hot line to the divine having been down for some millenia now. Given that there is no other equally efficacious procedure available, for practical purposes the 'ouput' of the legal procedure is identical to the objectively correct 'output.'
This objection concedes the main point I wanted to make, namely, that one cannot be a (pure) proceduralist when it comes to legal trials. This objection is the best I can do by way of a charitable interpretation of certain recent animadversions of commenter Philoponus in the comment thread to this post. But perhaps I haven't understood him.