Saturday, August 25, 2007

Am I too rational for my own good?

In the second episode of Richard Dawkins' new show, The Enemies of Reason, (which you can watch below) Dr. Dawkins takes on "alternative medicine", and pays special attention to homeopathy. In case anyone reading (if anyone is reading) is not familiar with this particular flavor of bullshit, homeopathy is a form of alternative medicine in which substances -- and it truly makes no difference what they are -- are diluted to the point that "in order to get one molecule of the active substance, you'd have to imbibe all of the atoms in the solar system."

Dawkins correctly notes that the only possible use for homeopathic "medicine" -- which, amounts to so much water, exactly as pure as that used for dilution -- is as an expensive placebo. The placebo effect is well-documented and supported by scientific evidence. The psychological effect of believing that a treatment will work causes the body's self-healing mechanisms to function more effectively. In short it only works if you believe it will work. I don't believe that ignorance is bliss, but it's undeniable that under certain circumstances, it can be healthy.

I also considered this some time ago, a few years before I realized that the god that I had heard about all my life could not possibly exist. Before this realization came the realization that prayer could not possibly be efficacious. An all-knowing and unchanging god with an all-encompassing plan would necessarily have decided whether or not to do something long before it is requested, before the birth of the requester, and indeed, before the dawn of time itself. It could not be possible to sway the will of such a being, therefore prayer is useless. My pastor did not know how to respond to this conclusion, except to subtly scold me for thinking that way. Pastors are not, after all, paid to think. My mother, the theologian, told me that the purpose of prayer is not to change God, but to change yourself, which I immediately recognized as the placebo effect.

Of course, knowing the nature of a placebo negates its effect, which at first made me reluctant to share this knowledge. It was a moral dilemma: on one hand, telling someone that prayer is a placebo would make it as useless to them as it was to me, but on the other hand, keeping someone uninformed is always a disservice. This, to me, is perhaps the most important moral question, and it took several years for me to come to the conclusion that sharing the truth with someone is always the moral thing to do, no matter how unwelcome it may be. Each person has a right to their own beliefs, but that includes the right to base those beliefs on the best information available. By giving a person incomplete or false information, you deny them the right to believe what is true, and I can thing of no worse affront. This is why it is considered unethical in scientific medicine to prescribe placebos except in cases such as clinical trials where patients may be in the test group or the control, and thus know that they may receive placebos, and agree to be temporarily uninformed so that the placebo effect can be tested.

For doctors, it would be unethical to do any more than allow a patient to believe that a placebo will be efficacious, whereas religionists and practitioners of "alternative" medicine are free free to employ lies, dogma, and obfuscation to sell their snake oil. Deceiving anyone, yourself included, is harmful, and individuals only have the right to choose that type of harm for themselves. I prefer to rely on my body's own defenses whenever it is reasonable to do so, but I also avoid these types of self-deception that could enhance those defenses, so it is likely that I would be at least marginally more healthy if I were less rational. Perhaps I am too rational for my own good.

Saturday, August 11, 2007

Pascal's Wager

Pascal's Wager states that it is logical to believe in God based on a cost/benefit analysis. According to the wager, belief costs nothing, but the potential reward is infinite, as is the potential penalty for disbelief. If there is no god, nothing is gained and nothing is lost whether you believe or disbelieve. This, of course, neglects the value of a belief being true, presumably because it assumes that the chances of being correct in either situation is 50-50 and that no further study can significantly alter these odds.

There are many counter-arguments to Pascal's Wager, and I won't address nearly all of them here, but perhaps the most obvious is that it is a false dichotomy. There are many mutually-exclusive beliefs about the existence of a god or gods, and even if one chose to be a theist, one would still presumably need to choose the correct religion in order to gain the reward. Most people would then start by placing each religion on equal footing with one another and with atheism in terms of probability, but this is already mistaken because there is a true dichotomy between theism and atheism (when defined simply as a lack of theism). Religions posit a god, then make further assertions about that god, so the 50% probability that there is a god must be divided among them. At this point, it becomes clear that, in the absence of evidence either way, the atheist is more likely to be right than a member of any given religion, but this counter-argument misses the point a bit, because in this case it could still be desirable to forego all likelihood of being correct (a finite benefit) for the chance at attaining an infinite reward. This is a much truer wager, and resembles a lottery with a small buy-in for a small chance at big rewards, but in this case there is not only a huge pot to be won, but an equally huge penalty for not winning. You have to play to win, but you don't have to play to lose. Under these circumstances, it still makes sense to pick a religion and believe.

The real problem is that people assume that even if there's a 50% chance of a god, that god must resemble the god of some religion. It's tempting to divide the god side according to the world's religions, or even the individual (and unique!) concepts of god that each believer holds, but without consistent revelation, these gods are no more likely than any other conceivable gods. Herein lies, I believe, the real achilles heel of Pascal's Wager.

Let's assume that there is a god that is aware of you, gives a damn about what you believe, and will punish you for believing the wrong things with regard to its own existence. The probability of this is already far below 50%, but for the sake of argument, we'll assume that these are necessary properties of a god, and ignore deist gods who don't care or those who value other properties such as your taste in music, your favorite color, whether you donate $5 to this website, whether you have a cute mole on your ass, etc. etc. ad infinitum. This god is at least as likely to value skepticism, evidence-based belief, and critical thinking as it is to value blind faith. Add the fact that if there is a god, it goes to great lengths to hide its presence (as evidenced by the fact that we have to ask the question), and the benefit of pleasing the skepticism god is better than that of the faith god, as the latter would obviously be a sadistic psychopath who wants to punish more than to reward. Who would want to spend eternity with a jerk like that? In fact, in the watered-down theologies in which the punishment is only annihilation or separation from the god, I'd say that the punishment is better than the reward.

When you look at it this way, the greatest benefit possible comes from pleasing a god that values skepticism. Skepticism also offers the greatest chance of a benefit because you would have a 25% chance that you will be wrong but will gain the favor of a god that values skepticism on top of the %50 chance of being correct but receiving no further benefit, therefore skepticism is the only logical choice. Of course, adding gods that don't care about belief back into the equation significantly diminishes that 25%, but it diminishes the chances of having the correct theistic belief equally.

Sure, some would argue that a god that values skepticism is far less likely than one that values faith, but what evidence do we have that this is the case? Only the word of theologians who have no more evidence than you or I, can't agree on anything else, and would have to go get a real job if people didn't believe. That, Mr. Adams, does not "pass the sniff test".

Of course, it would be stupid to disbelieve in a god that probably doesn't exist simply for the purpose of currying favor with it, but I think I've demonstrated that it doesn't make any more sense to believe for the same reason.