Posts

Comments

Comment by nick8 on Against Devil's Advocacy · 2008-06-09T21:14:55.000Z · LW · GW

Knowledge is generally quantum in nature: there are many possible futures. Thus, Alice might believe Schrodinger's cat is alive, and Bob might believe it is dead. Such genuine disagreement can occur even in a purely deterministic Newtonian world where people have distinct bits of imperfect knowledge -- people can genuinely believe false things and have no way of knowing that they are false. Indeed, in many ways this is the normal state of knowledge, since by the pigeonhole principle no individual can host more than a tiny fraction of the world's data in his brain. Another reason for genuine disagreement is that people believe things that are true given the way they use words, but an opponent believes it is false given the way the opponent uses words. This kind of disagreement is one over description rather than over external reality.

Whether the disagreement is over semantics or external reality, both sides can have arguments in their favor, and arguments against, and it is often highly non-obvious how to reconcile the contraditions. People arrive closer to the truth by discussions in which it is understood that either side may be wrong.

Legal disputes follow the same "quantum" logic. We don't want to have cops go around shooting people just because they strongly believe they are guilty. Rather we go through a process that assumes that the party might be innocent or guilty -- much like a quantum state. Evidence is gathered and each side is allowed to put forward arguments in its favor, hopefully, the evidence causes ignorance to collapse, and a verdict can be more confidently reached.

In this world of uncertainty your opponent's argument can be as important as your own. An effective seeker after truth, as well as the effective advocate, understands counter-arguments, in order to discover the holes either in that argument or in one's own.

Michael Rose is thus quite right about this. If Richard Dawkins wants to debate creation or religion, he should be happy to exercise with "devil's advocate" arguments of his opponents. If he doesn't want to do that, he won't be an effective advocate and he's wasting his time. A dogmatic profession that "God does not exist" is almost competely uninformative -- we already know that millions of people are atheists.

(In fact, one of the beauties of Blind Watchmaker is that Dawkins understood and responded to creationist arguments, e.g. the watchmaker anaology, far better than other advocates of evolution, who usually neglected to explain the highly improbable design-like products of evolution. In the process he highlighted a crucial aspect of evolution, adaptation, that many others writing about evolution misleadingly downplayed, allowing creationists arguments based on accurate observations of the design-like nature of organisms to go unrebutted. Dawkins had to play plenty of devil's advocacy at least in his own mind to achieve this understanding of creationist arguments, and in the process we also learned more about evolution. It would be a shame if he's lost this skill and taken to just dogmatically asserting his beliefs).

Strong feelings and personal confidence often have little correlation to actual truth. Even if you believe strongly that you are right, or perhaps especially if you have strong beliefs, if you are going to engage in debate you should understand the "devil's advocate" arguments of your opponents. If you think it's a waste of time to understand arguments in favor of religion, asteroidal chocolate, or anything else that you find credible or incredible, you should not be surprised that (a) you'll be very bad at convincing people who don't share your beliefs to share them, and (b) if you do turn out to be wrong, you won't understand why you are wrong. It's fair to say that for the chocolate cake, since nobody believes it, there is nobody that needs convincing, and it's also fair to not be concerned about the risk that one's opinion about it is wrong. It's also reasonable to ignore religion and creationism for the second reason. But if you want to convince a religious person to be an atheist, or a creationist to be an evolutionist, you'll be far better off understanding their arguments first, and you might well arrive at more accurate forms of atheism and evolution in the process.