Redux

I’ve rerecorded much of the vocals for Terminal, and remixed it a bit, too. I think it sounds a lot better now. (Or, at least, less out of tune :). I’ve also remixed How I Stay Safe, but it still suffers from those cheap mics, and I don’t think much will change that save a redoing of the whole song. If you actually downloaded these songs before, please download them again! I can’t wait to record the next song I have lined up (inspired by some California geography!), but it might be a whole month before I have the time to record it.

http://www.cs.berkeley.edu/~aj/music/Terminal.mp3
http://www.cs.berkeley.edu/~aj/music/How I Stay Safe.mp3

Everyone seems to be writing about David Hume these days. Hume really liked talking about miracles, but let’s generalize a bit.

Let’s say there’s some evidence that event X occurred. What does it take for you to believe that X actually happened? Hume says you should “proportion your believe to the evidence”, as follows: 1. Given what we know about natural law, how likely is it that this event occurred? 2. How reliable are the witnesses (the people or instruments that provide the evidence)?

I totally agree with this. What I wonder about is whether your threshold for belief of X should also vary based on how costly it is to believe that X occurred. My intuition is that it shouldn’t: your disposition to believe should rely only on the facts, not the circumstances of your believing. (Otherwise, you could say you’d be intentionally deluding yourself.)

But it gets confusing. Given event X, you can either say a) I believe it happened b) I believe it did not happen c) I don’t believe one way or another. Now let’s consider two events:
A) Ooga is the progeny of God and died for our sins. If you believe this, and you’re right, you go to heaven. Otherwise, you go to hell.
B) Blooga is the progeny of God and died for our sins. If you believe this, and you’re right, and you act according to the tenets of Bloogaism, you go to heaven. Otherwise, you go to hell.

Notice that both of these events essentially eliminate the “I don’t believe one way or another” option, since feeling that way is effectively identical to not believing.

The question is: do you need more evidence to believe Blooga? It’s totally cool to believe Ooga, since that doesn’t really cost you anything, and hey, you might get lucky and go to heaven. So your threshold of evidence for Ooga does not need to be high. But Blooga, on the other hand, well, now you have to start going to Blooga church, and read the Blooga scriptures, and stay away from those sinful Oogaers. The cost of believing is greater, even though the payoff is the same; do you need a correspondingly greater amount of evidence to believe Blooga? You could imagine a scenario in which the cost rises to infinity, in which case you better be damn sure that the event happened, or else you’re really screwing yourself over. So the amount of evidence you require to make up your mind might in some way be proportional to the effect that believing will have on you.

The analogy here, of course, is to Pascal’s Wager. I think Pascal errantly ignored the cost of believing.

This entry was posted in Uncategorized. Bookmark the permalink.

2 Responses to Redux