Wednesday, March 18, 2009

Bayes' Theorem and the Existence of God

This is an Scientific American article on the usefulness of Bayes' theorem in assessing such things as theism.

Well, well, of course, of course. Bayes' theorem doesn't give you objective antecedent probabilities. So we have to go with the ones we've got, and conditionalize our beliefs based on the evidence we received? Why go with the ones we've got? Can you suggest any others, without introducing an incoherence in our beliefs?

That's why I'm an annoying Bayesian subjectivist when it comes to prior probabilities. Let people have the antecedent probabilities that they have, and let us see what evidence moves the scale up or down.

This raises an issue, I realize, for advocates of the "outsider test." One one level, we can't get outside of our belief system and shouldn't try. Otherwise, we end up in a general skepticism, which is where Descartes should have landed if he had been consistent. But if the outsider test argument works, isn't it an argument attempting to show that Christian theism is less probable than you originally thought it was?


Jimmy said...

Did you just say something about John Loftus' Outsider Test? How dare you! Do you know how emails he get a week praising him and his arguments for showing some poor theist the rational light? Do you know that by not accepting John's arguments you are ignorant, irrational, dumb, and blind? I mean William Lane Craig is running scared from Loftus. What makes you, a mere philosopher, think you have what it takes to criticize the eternal Loftus Outsider Test?

Eric said...

"But if the outsider test argument works, isn't it an argument attempting to show that Christian theism is less probable than you originally thought it was?"

Interestingly, it not only tries to show this, but also presupposes it. John calls for the presumption of strong skepticism *as one undertakes* the outsider test, i.e. the presumption that one's beliefs are *probably false*. He justifies this by appealing to the fact that what one believes is in most cases a function of where one is raised (the 'religious dependency thesis').

Victor Reppert said...

Can you generate a real argument out of this, with, like, you know, premises and a conclusion?

Eric said...

"Can you generate a real argument out of this, with, like, you know, premises and a conclusion?"

Victor, I think it depends on what you mean by 'this'! As I've tried to point out to John in the past, he has to do more work to distinguish:

(1) Arguments justifying the outsider test (i.e. the presumption of strong skepticism)

(2) The outsider test itself

(3) Arguments from the outsider test

I find John's argument for (1) unpersuasive and possibly fallacious, since it rests on what he calls the 'religious dependency thesis,' i.e. the fact that most people's religious beliefs are determined by where they're raised. Obviously, there's a case to be made that he's committing the genetic fallacy here. He claims he isn't, since he's not arguing that religious belief is therefore false; I think he is, because the dependency thesis concerns the origin of a belief, and the outsider test is necessarily about justification because it just is a form of skepticism (i.e. I take the genetic fallacy to be a confusion of a belief's origins with issues about justification; John has a narrower understanding of it, according to which it requires reaching a conclusion of truth or falsity from the origin of a belief).

(2) isn't an argument, but a procedure for testing one's beliefs.

Any arguments characterized by (3) are always going to be, it seems to me, problematic. If a Christian undertakes the outsider test and finds, in the end, that his faith is still intact, what then? Will the atheist question his commitment to the test? Does the atheist understand the test in such a way that it's only 'properly' undertaken if it results in a loss of faith?

In short, the answer to your question is yes with respect to (1) and (3), but the resulting arguments are not, it seems to me, very strong at all.

John W. Loftus said...

Here's the fullest statement of the OTF one can find online.

The comments can be read as instructive as well.

Mark Frank said...


I think you are letting Bayes distract you from the more fundamental issue. Does it make sense to talk of the probability of theism?

If you view a probability as just a numerical way of expressing your strength of belief then it makes a sort of sense. But then - so what? No one denies the strength of your belief and it has marginal importance for the strength of anyone else's belief.

But all other definitions of probability fall apart for an outcome such as "the existence of a deity" or "the value of the fundamental constants of the universe". For example, imagine someone asked what is the probability of the earth being 93 million miles from the Sun (within suitable margins). It makes no sense unless you specify some context (explicit or implied). Do you mean, given the mass of the Earth and the Sun and Newton's laws? In that case it is close to 1. Or do you mean, given Newton's laws and the mass of the Sun, when the solar system was formed what were the chances that the third planet would be 93 million miles away? You can easily think of infinitely many more such contexts. Given the context you might come up with a classical, subjective or frequentist probability. Without a context - well it is a meaningless question.

What is the equivalent context for probability that a God (or the God of Christianity) exists?

oscarspaz said...

After reading Loftus's OTF blog, it seems that there is a disconnection between this:

This whole inside/outside perspective is quite a dilemma and prompts me to propose and argue on behalf of the OTF, the result of which makes the presumption of skepticism the preferred stance when approaching any religious faith, especially one’s own. The outsider test is simply a challenge to test one’s own religious faith with the presumption of skepticism, as an outsider. It calls upon believers to "Test or examine your religious beliefs as if you were outsiders with the same presumption of skepticism you use to test or examine other religious beliefs." Its presumption is that when examining any set of religious beliefs skepticism is warranted,

and this (second half of the sentence):

since the odds are good that the particular set of religious beliefs you have adopted is wrong. "

The best conclusion one can draw from OTF is one should reserve room for one's faith can be wrong - so be a Mulism Skeptic, Christian Skeptic or Budda Skeptic or Hindu Skeptic, etc.

How can Dr. Lofus jump from "be a skeptic" to the conclusion that one should "abandon the ship" boggles my mind. What kind of "odds" is he referring to? Whether you belief in this way or claims not to belief in anything, there is no certainty that you are absolutely right.

In the "six objection" section, Dr. Lotfus deals with the issues of people who escaped outside of "their culturally adopted faith" by expressing his "opinion" that these cases are "exceptions" or these people had not think through the issues. It is kind of like these points do not fit my curve so please ignore them or let me explain to you why they are not important. This is "un-scientific" and arrogant to ignore any DATA.

Also, how dare Dr. Loftus say that "They have no initial way of truly investigating the proffered faith." These people know the cost that came with the change of faith. They put their lives on the line by being ostracized or be killed. They did not have Dr. Lutfus luxury to sit down and have a debate over it. It is a matter of life and death, yet they chose to die.

In my opinion, Dr. Lotfus's idea is dangerous to all people not only to religious people. It promotes anarchy under the disguise of "intellectual argument" hiding his opinion in arguments and chose to argue away data that does not fit his conversion experience.

Dr. Lotfus's OTF is a product of our culture today.