Wednesday, July 23, 2008

A Nonconstructive argument about the election

I have heard several times in the media the following nonconstructive proof that Obama will win the election. They, of course, do not call it a nonconstructive proof. Since politics is far less predictable and rigorous than math I do not really buy the argument, but its of some interest to me that there is a nonconstructive argument in politics. Here is how we might phrase it. There are two cases.
  1. The Iraq war goes well. Then the Iraq war is off of the front pages. In this case, McCain's advantage, that he is seen (rightly or wrongly) as being better to have as prez when we are at war, is nullified. Hence Obama, which is seen (righly or wrongly) as being better on domestic issues will win.
  2. The Iraq war goes badly. Then Obama can say (or he might not even need to say so explicitly) that he was right about the war being a mistake in the first place.
What is wrong with this argument?
  1. The election may hinge on so many other things: a scandal, a mistatement, obvious things I am not mentioning, nonobvious things that have not come to light yet.
  2. The Iraq war might go (or be portrayed as going) some intermediary thing which is neither well or badly. In fact, the very terms well and badl are not well defined.
  3. More generally, its very hard to apply simple logic to an election. Or complex logic.


  1. In US presidential elections, the more charismatic candidate wins.

  2. Setting aside the validity or lack thereof of the argument, why do you characterize it as nonconstructive? In a nonconstructive proof, the existence of some object is demonstrated without showing how to actually produce such an object. I don't see how the given argument fits into that mold.

  3. Kurt: not only existence is interpreted differently in constructive logic, but also disjunction. The argument is based on a case distinction where you can't decide which of the alternatives holds, hence it's non-constructive.

  4. So any proof where you partition into cases and consider the cases separately is considered nonconstructive? That seems to cover an awful lot of ground.

  5. The beauty of living in a world where intrade exists is that whenever someone makes brash arguments for why some event is certain to happen, you can ask them to literally put their money where their mouth is, or shut up. Obama is at 65% right now.

  6. The American presidential election (like most elections) is in the long run, less about *who* wins than about what *ideas* win. The question of who is right about Iraq is therefore less important than whose ideas are right.

    From this point of view, the MNF-Iraq Commandant's Guidance is a very important (three-page) document.

    The 23 principles that it embodies are remarkably liberal *and* remarkably pragmatic. So much so, that *either* Obama or McCain might well choose to continue the war-fighting / nation-building / peace-making effort in Iraq along these lines.

    The point being that regardless of who wins the election, these ideas *already* have won.

    There is a present quite a substantial "nation-building" initiative among scientists ... to succeed it very likely will have to embrace many of the principles set forth in the Guidance ... for the same pragmatic reasons.

  7. Kurt: No, case distinction *is* constructive if the condition the cases are distinguished by is decidable.