Tuesday, May 12, 2009

How to Argue: Pascal's Wager

Hello again. Today is the first in a series of posts that I am choosing to call the How to Argue entries. These are intended as a guide to debating people of faith (an all too common situation for those of us without it) in a manner that won't create enemies, and that will hopefully help you make your point clearly.

More importantly, I will be using these discussions of common arguments to illustrate the principles of critical reasoning that will help you in all areas of your rational life.

Before we begin, note one important fact. You will almost never win a rational argument with a person of faith. Faith is by definition irrational. Expecting a person who whole-heartedly believes something without proof to change their mind is ridiculous. Instead, try to settle for pointing out that the "rebuttals" often offered to your heresy are intellectually and philosophically bankrupt. If you are calm and kind, you may even have some impact.

Today we will be discussing a classic argument for faith first posed by the mathematician and physical scientist Blaise Pascal. I've chosen this particular argument for three reasons. First, you will hear it constantly. Pascal's Wager, as it's known, is a classic last-ditch argument for faith in the face of reason. Second, the faithful tend to like it for these debates because it sounds scientific and reasonable. Lastly, it is a particularly easy argument to rebut, and so makes for a good example of today's topic. So let's start there:

An argument is only as sound as the assumptions upon which it is based.

Repeat that to yourself a few times, and keep it firmly in mind as we discuss today's argument. At first blush you might be tempted to ask, "Assumptions? I thought we were talking about reason and empiricism and rational skepticism and.... shouldn't we not have any assumptions at all?" Disabuse yourself of this idea. All philosophy is based on assumption. It's not like maths, where axiomatic truth (and therefore rigid proof) is possible. In order for reason to exist at all, a few basic assumptions have to be made.

Just consider: you must assume yourself to exist. This seems obvious, but it can not be proved, Voltaire be damned. There is no way to assert that you're not the "dream of a butterfly who thought himself a man," to borrow a classic. You can not say with certainty that you are not part of a larger consciousness, a computer simulation, or a mass hallucination. I won't belabor this point, and you can read up on David Hume's notion of self for further reading.

Likewise, you must assume a number of other things for reason to be worthwhile. You must assume that the universe has substance outside of your own perception. You must assume that your senses, while imperfect, are not farcical. You must assume that inference itself is possible.

This is a much, much larger topic than I care to devote this post to. The idea I'm trying to get across here is that all philosophical arguments are based on assumptions at some level. This does not mean, however, that all assumptions are equally valid (when you think about it, "valid" in this context actually means "useful"). The idea "Objects continue to exist even when I can not see them." is useful because it leads to further useful ideas. Without this assumption I would have a hard time finding my keys. "There are no purple apples." is a much less useful assumption. For one, there is no need to make it, as there either are or aren't purple apples (within the framework of assumptions like "the physical world exists"). More importantly, it precludes a conclusion that could be made from simpler, more direct propositions. Confused yet? Let's get to the meat of this post.

Pascal's Wager states that it must be more reasonable to believe in god than not to, as the reward if right is infinitely good, while the outcome of all other scenarios ranges from negligible to infinitely bad.

You might think that sounds a bit like a Decision Theory proposition, and you'd be right. Pascal himself actually expressed it as an outcome matrix:



God ExistsGod does not Exist
FaithEternal Salvation (infinite reward)Nothing
DoubtEternal Damnation (infinite loss)Nothing


You're more likely to hear it from a person of faith in a form like this:

You can't prove that god doesn't exist, so doesn't it make sense for you to believe in him "just in case"?
The short answer is no. The long answer is, well, long.

Pascal's Wager makes at least six unjustified assumptions. I'm going to cover four in some depth; see if you can spot the others. It shouldn't be difficult. You could drive a truck through the holes in this argument.

Take a look at the table of outcomes above. One thing should be readily apparent. Pascal's Wager assumes that god rewards belief. What reason is there to believe this to be true? Most faiths have belief as a central tenet, but almost none have it as the sole requirement for salvation. Catholics have the requirements of confession and penance, Hindus the notion of Satyaghara, Buddhists the search for enlightenment, Muslims the communal aspects of "right living", and on and on. So what reason is there to believe that simply believing a god exists is enough to earn his favor?

I would go even farther, though. Most religions recognize "faithful" behavior done only for self-interest as hypocrisy. This is exactly what Pascal proposed. The whole crux of his wager is to adopt a belief solely in the hope of gaining personally from it, an idea that most faithful people, regardless of religion, would find sinful. So do we have reason to believe that god rewards belief, and that he rewards belief based only in self-interest?

Pascal's next failure is that his idea presents a False Dichotomy. In this argument, there are only two possible outcomes: god exists or does not. But what's the assumption behind that? The assumption that Pascal made invisible to his argument was that there was only one possible god, the Christian one (this is born out in his other writings). What if we account for other possibilities, though? How do the permutations change if suddenly there could be multiple warring gods? What if there were a non-specific, non-sentient creative force? What if we only added rows for the gods that others already believe in? Mustn't we then have outcomes mapped for Yahweh and Krishna and Aphrodite and Osiris and a multitude of others? Most of these faiths are exclusionary, meaning that faith in any other is a sin or heresy. What follows from this is that the number of possible outcomes is huge, and in most of them the best you could hope for is for nothing to happen. Most of the time (in this probability matrix) the outcome is eternal damnation or some sort of other punishment.

The third faulty assumption I'll touch on is similar to Pascal's assumption that only one god was possible. Pascal assumed that god's existence necessitated an afterlife, and likely a split one (heaven / hell) at that. This again came from him only considering a single faith as the possibility. In fact, he had already made up his mind, and was only justifying his conclusion.

I will be repeating this over and over as I keep posting, but starting with a conclusion is the most dangerous thing you can do, intellectually.
The idea that there must be an afterlife is not a reasonable one. For one, we have no evidence for it, but that's not the point I'm trying to make. Assuming that there are a heaven and a hell is exactly as reasonable as believing, like the Jews do, that there is only a heaven. It is the same as the Hindu notion that life after death continues on the mortal plane. Even the Catholics have a third option in purgatory. When you start adding these options to the matrix, it becomes apparent that the number of possible outcomes is far too large for any approach to make sense without some additional knowledge. i.e. "Just in case" no longer makes sense, as you are overwhelmingly more likely to make the wrong choice.

Lastly: Pascal's greatest failure. He makes the assumption that living with faith, even if it is misplaced, does not do any harm. You need only read Wes's last post ("Cognitive Dissonance") for an excellent example of why this is not true. Faith means abandoning reason, it means changing your identity, and it means creating an area of yourself that is immune to criticism and debate. The first two are not always bad*, but the third is. We can only learn when our ideas are challenged. Accepting that any part of our conception of the world is inviolate is the height of arrogance, and a sure-fire way to miss a lot of really important things.

When you have knocked the legs out from Pascal's Wager, it becomes clear what it really is. It is not a logical proposition, it's a cheap dodge to make an irrational belief appear reasonable.

This is what you should try to do with any argument that you are presented with, regardless of whether you agree with it or not (I would even say particularly when you agree with it). Look for the assumptions. What isn't being said in the argument itself? What's being taken for granted, that if untrue, would completely reframe the argument? Is this castle just built on a swamp?

Everyone said I was daft to build a castle on a swamp, but I built it all the same, just to show them. It sank into the swamp. - Monty Python and the Holy Grail


Be well, and come back soon for more really, really long posts.

-Chris



* I don't hold that everything can be dealt with rationally, because I've been in a relationship. Nor is pure reason an intelligent pursuit. Love and laughter matter, even if they aren't always reasonable. It's just a good idea to recognize those parts of yourself that aren't rational for what they are, rather than believing them to be a valid way to find new knowledge or explain your world. More on this later, I promise.











No comments:

Post a Comment