Sunday, October 28, 2012

But how shall we follow probabilities?

Loftus: We should think exclusively in terms of the probabilities.


VR: How in blazes do you calculate probabilities? Probability theory tells you how you get from a prior probability to a posterior probability. What it does not tell you is what prior probabilities are correct. Hence I can begin with a probability of 1 for the Resurrection and end up with a probability of 1 for the resurrection. Ditto for a probability of zero. So telling me to think exclusively in terms of probabilities tells me squat. Probability theory does tell you how, given enough evidence and a small enough split between probabilities, we can come to an agreement about whether something is true or not. But if there is a large split between antecedent probabilities, we can easily have rational people taking opposite beliefs to their graves.

I happen to think that there are no right or wrong antecedent probabilities. We start with the probabilities we have and go from there. My view is that a Bayesian-rational person can conclude that Jesus rose from the dead.

24 comments:

Crude said...

Victor,

How in blazes do you calculate probabilities?

Could you clarify what you mean here? I know you've discussed calculating bayesian probabilities before - so I take it what you mean here is that 'initial' probabilities are not calculated?

B. Prokop said...

That's why they're "initial".

Crude said...

I figured, but I wanted to clarify so there's no one derping about with "he doesn't even know how to calculate probabilities!" So, there we go, nice and covered.

John W. Loftus said...

How do we calculate the probabilities? In science it's easy. In life it can be hard. But we SHOULD do it if we want to achieve success in knowledge and in life.

I really don't think believers can comprehend what I'm saying. They are not even trying. They can't. It won't be until they reject faith that it'll be possible. But if and when that happens it will be crystal clear.

You see Vic is not addressing the issue I raised. I said we should think in exclusively in terms of probabilities, and his response is to question how we can do this, not whether we should try.

When it comes to gods and goddesses, virgins births, miraculous cures, Demi-gods or incarnate gods who walk the earth, claims of resurrection, and the like, my argument is that we should treat all of these claims with the same standard, as non-believers. Not to do so falls prey to what Stephen Law argues, not to do so is to apply a double standard, a hypocritical standard. Not to do so allows one's cognitive biases to run a muck.

Stephen Law argues that "Anything based on faith, no matter how ludicrous, can be made to be consistent with the available evidence, given a little patience and ingenuity." (Believing Bullshit, p. 75). Because of this it is essential that we think exclusively in terms of probabilities, the probabilities of a non-believer in all extraordinary claims, that is, the concrete examples I have given.

Crude said...

From the other thread:

It won't be until they reject faith that it'll be possible.

Pretty much everyone in this faith rejects faith as you've so far defined it - even Bob would have to now that you've defined 'pursuing a goal with extremely low odds' as not necessarily being an act of faith.

And yet here we are, remaining theists and Christians. So where did you go wrong?

Take once again Vic's question. You see he's not addressing the issue I raise. I said we should think in exclusively in terms of probabilities, and his response is to question how we can do this, not whether we should try.

Er, he's addressing the issue you raise. You said 'we should think exclusively in terms of probabilities'. He asked how this is possible, and how you set your initial probabilities. I asked you as well - how do you 'think in terms of probabilities' with regards to your axioms or foundational beliefs? I left an obvious out for you - we can't 'think exclusively in terms of probabilities' when it comes to choosing our axioms or starting beliefs. Or are you going to somehow make the move where axioms should only be chosen based on probabilities you estimate?

Then GK Chesterton's book based on it should be trashed too.

Chesterton didn't make the mistakes you did with regards to the Outsider test as you propose it. As has been said before, what's worthwhile of it isn't original, what's original with it isn't worthwhile.

Meanwhile, you've failed to defend your definition of faith on its own terms, contradicted yourself about the effects of faith, refuse to clarify regarding how you ascertain initial beliefs, and have accidentally lended support to Pascal's Wager.

cl, the many gods objection to Pascal's Wager defending his Catholic faith at the time destroys the force of his Wager. Which god, which religion should we wager for? It only works if there is one true faith and non-belief. So, which faith is the true one?

If the force of the wager is merely reduced to 'There are multiple gods, and I can't tell if any of them is more likely to exist than the rest', then you've still destroyed your argument: the result becomes 'worshiping any god is preferable to atheism'.

If, however, the stipulation becomes 'some claims of gods or religion are more credible than others', then you've already put Christianity back in the running pending an evaluation of the evidence.

So, John, which religion will you be subscribing to? Given that your gambling analogy supports Pascal's wager, while at the same time makes it so accepting Pascal's Wager is absolutely not an act of faith as you define it? And I say this as someone who is pretty damn leery of Pascal's Wager. ;)

With some added comments:

Not to do so allows one's cognitive biases to run a muck.

"Run amok".

Because of this it is essential that we think exclusively in terms of probabilities, the probabilities of a non-believer in all extraordinary claims, that is, the concrete examples I have given.

You've been asked to give concrete examples of how to determine your initial probabilities, and more. You've failed. Just as you failed to justify your definition of faith, according to your own stated standards.

You're all over the map, John. This isn't doing wonders for the probability that you're worth taking seriously on this matter.

B. Prokop said...

"I really don't think believers can comprehend what I'm saying ... But if and when [they reject faith] it will be crystal clear."

John, can you not see that what you've just written can be re-expressed as "As soon as people agree with me, then they will agree with me!" ???

This sounds fun - let me try it! "I really don't think that Loftus understands Christianity. But as soon as he acquires Faith, then all will be crystal clear!"

B. Prokop said...

Crude,

I think in John's case, "run a muck" might actually be appropriate!!!

Victor Reppert said...

JWL: When it comes to gods and goddesses, virgins births, miraculous cures, Demi-gods or incarnate gods who walk the earth, claims of resurrection, and the like, my argument is that we should treat all of these claims with the same standard, as non-believers. Not to do so falls prey to what Stephen Law argues, not to do so is to apply a double standard, a hypocritical standard. Not to do so allows one's cognitive biases to run a muck.

But, just for starters, nonbelievers disagree amongst themselves about how likely these events are. Consider C. S. Lewis just before his conversion to Christianity. He went from being in a state where he didn't believe in the Christian miracles, but thought they had a chance of being true, to actually believing that they really took place, based on evidence. How did this happen? Other skeptics are looking for extraordinary evidence, and then you have someone like P. Z. Myers, who would remain a skeptic even if Jesus were to come back. Or so he says. There's a whole continuum of people who don't accept these miracles with respect to their antecedent probability. If I have to dump my priors, whose priors should I adopt. Which outsider should I be?

Not this guy, surely.

http://old.richarddawkins.net/discussions/642394-there-can-be-no-evidence-for-god-revisited

Crude said...

Personally I think RD Miksa makes a powerful case for his view that, if anything like an 'outsider test' is to be taken, the ideal position is one of a deist.

SteveK said...

For those that experienced God firsthand, the probability of God=1. The math is pretty easy.

BeingItself said...

"For those that experienced God firsthand, the probability of God=1."

LOL. Wrong.

John W. Loftus said...

Rados Miksa criticizes the default position of OTF. It should not be agnosticism, or a reasonable skepticism, as I have argued, but rather deism. However, Miksa simply misunderstands deism, saying it represents the particular deist conclusion that there is a creator God. That’s not the case at all. Deism began with Herbert Cherbury in England during the seventeenth century as a natural religion where religious knowledge is acquired solely by the use of reason, as opposed to the Bible or the church, or faith. Deists affirmed that reason must support any theological belief. If it can’t, then that religious belief is to be jettisoned. That’s the heart of deism, not any particular conclusion arrived at by deists because of their method. Deism went through four different stages and traveled from England to America and France. The final stage is largely of French origin, where God is seen merely as the creator of the universe. God created it and set it in motion, but does nothing to intervene in its affairs. It soon evolved into atheism after Darwin published his book, Origin of Species, in 1859, because after all, it was based on reason not faith. If deism is properly understood then it is welcomed. Contemporary deists who still hold to the conclusion there is a creator God are simply not being consistent. After all, as I argue in chapter ten there is no such thing a reasonable faith, or for that matter, a reasonable religion.

See J. O’Higgins, who distinguishes between four types of Deism in “Hume and the Desists: A Contrast in Religious Approaches,” Journal of Theological Studies 23, no. 2 (October 1971): 479, 480, which is summarized in Norman L. Geisler and William D. Watkins, World’s Apart: A Handbook on Worldviews (Grand Rapids, MI: Baker Book House, 1989), pp. 148–49.

Crude said...

However, Miksa simply misunderstands deism, saying it represents the particular deist conclusion that there is a creator God. That’s not the case at all.

No, what Miksa does is describe what he means by deism, and goes on to defend why he believes the position of the deist, rather than the atheist or naturalist, is the ideal outsider position for evaluating faiths.

"Deism" did not "evolve into atheism" at any point. Now, maybe some deists became atheists - always possible. But atheism was around well in advance of deism, and deism continues to exist to this day as a live option - hence the various contemporary deists.

It's certainly an outsider position with regards to religion, and Miksa argues persuasively that it makes a superior position at that.

John W. Loftus said...

Naw, you only wish he made a good case. Deism is as much built on faith as any religion that is not based on reason.

Ah, but why do I bother?

John W. Loftus said...

That is, his kind of deism.

steve said...

John W. Loftus said...

"I really don't think believers can comprehend what I'm saying. They are not even trying. They can't. It won't be until they reject faith that it'll be possible."

What's the probability of a royal flush? Well, by one reckoning, it's

649,739 : 1

But that assumes the deck was randomly shuffled. What's the probability of the deck was stacked?

Loftus would have to know ahead of time whether or not the deck was stacked before he could crunch the numbers.

He confronts the same issue with miracles. What background information is he including or excluding?

steve said...

John W. Loftus said...

"Because of this it is essential that we think exclusively in terms of probabilities, the probabilities of a non-believer in all extraordinary claims, that is, the concrete examples I have given."

Is rolling sixes 10 times in a row ordinary or extraordinary? Depends on whether or not the dice are loaded. If so, then it's not extraordinary.

Loftus needs to know in advance what variables are feeding into the prior probabilities. He can't take a shortcut and just declare something to be "extraordinary."

"Because of this it is essential that we think exclusively in terms of probabilities, the probabilities of a non-believer in all extraordinary claims..."

In which case, Loftus has his thumb on the scales. To make the unbeliever the standard of reference begs the question. That's prejudicial.

Crude said...

Naw, you only wish he made a good case. Deism is as much built on faith as any religion that is not based on reason.

That is, his kind of deism.

No, Miksa's 'deism' is demonstrable not built on faith - it's simply the position of the outsider he uses for the OTF, and it happens to be superior to the naturalist outsider for the reasons he states.

If you're saying that you can't have an outsider position for the OTF which you think is irrational or incorrect, you're skunking the OTF before it even gets off the ground.

Ah, but why do I bother?

You don't really bother. You don't put in effort, or if you do, you fumble with it. Learn what people are saying to you, John. Understand them. You're never going to get very far if you just keep on misconstruing your critics and having ungrounded faith that all of them are deluded.

Crude said...

I really don't think believers can comprehend what I'm saying. They are not even trying. They can't. It won't be until they reject faith that it'll be possible. But if and when that happens it will be crystal clear.

This is worth pondering.

According to John, you have to reject your faith in order to agree with him. But if you reject your faith first and THEN agree with him - what exactly has John accomplished?

So, John just said that he's pretty useless in the attempt to get believers to deconvert. He functions instead as a kind of after-the-fact supporter/cheerleader of atheists.

steve said...

John W. Loftus said...

"How do we calculate the probabilities? In science it's easy. In life it can be hard. But we SHOULD do it if we want to achieve success in knowledge and in life."

Okay, John. Why don't you crunch the numbers.

You think the Resurrection is highly unlikely. You'd have to prove one of two things:

i) Calculate the improbability that God exists;

ii) Even assuming God exists, calculate the improbability that God intended to raise Jesus from the dead.

The mere infrequency of God raising the dead doesn't, of itself, tell us anything about God's intentions regarding the resurrection of Jesus.

Anonymous said...

Y'all give John too much slack.

Hold him to his comments on probability. I mean, REALLY hold him to them. Remind him that, on randomness, the probability of us even being here is near-negligible. Then when he retreats to "multiverse" speculation, remind him of his own claim that we must have POSITIVE EVIDENCE for what we believe (cf. Loftus, "Listing of Cognitive Biases"). Then, if he insists that the speculations of theoretical physicists count as "positive evidence," remind him of Vesa Palonen's paper, Bayesian Considerations on the Multiverse, which uses Bayes theorem to conclude, "because multiverse hypotheses do not predict fine-tuning for this particular universe any better than a single universe hypothesis, multiverse hypotheses are not adequate explanations for fine-tuning."

I literally guarantee that he will not be able to adequately respond. He will itch, squirm, dodge, weave, insult and reply with non-sequiturs, but he will NOT successfully rebut the reasoning.

Then you can stop wasting precious time on somebody who is neither committed to reason nor in earnest pursuit of truth.

John W. Loftus said...

Steve, you are obtuse if you think that by the word "God" you intend to mean the Christian one.

So let's look at this, quoting from you:

You think the Resurrection is highly unlikely. You'd have to prove one of two things:

i) Calculate the improbability that Allah exists;

ii) Even assuming Allah exists, calculate the improbability that Allah intended to raise Jesus from the dead.

The mere infrequency of Allah raising the dead doesn't, of itself, tell us anything about Allah's intentions regarding the resurrection of Jesus.

None of you get it. Which god you refer to has everything to do with which god prevails in your culture.

John W. Loftus said...

cl said: "Hold him to his comments on probability. I mean, REALLY hold him to them."

Fine. You seem to think the only alternatives are your own particular faith and the randomness answer.

That is emphatically NOT the case!

Every believer would use the same argument and then conclude afterward for their own natural theology based on it.

You just don't get it. I don't expect you too. Given the numerous different and contradictory religious faiths out there each owing to their own culture, it's better to think along the lines of science. Faith has no answer to the these kind of problems. Given that faith has no answer let's see what science tell us. That's my position. No religious answer based on faith, which is the warp and woof of theology) I'm willing to await the results of science.

Eric said...

cl: "Hold him to his comments on probability. I mean, REALLY hold him to them."

John: "Fine. You seem to think the only alternatives are your own particular faith and the randomness answer."

Recently on DC, John posted about some odd comments that William Lane Craig has made about John's move to atheism.

First, let me agree with John: If Craig has no grounds for his pornography remarks, he should apologize *publicly* to John, and he should make it clear that he was either careless or mistaken (I seriously doubt that he's deliberately lying).

Second, however, I want to hold John to his probability claims, as CL suggested. In his post on Craig;s remarks, John certainly seems to think that it's true that publicly saying false things about another person is wrong. Here's my question for John: How can you make sense of the notion that you reached this conclusion by assessing the probabilities? Is it merely more probable than not that Craig is in the wrong if he's falsely accusing John of a pornography addiction? That's ludicrous, right? But if you didn't reach this conclusion by assessing the probabilities, then you're irrational to hold it (by your own reasoning).

I think that this example clearly illustrates the inadequacy of the "we must think exclusively in terms of the probabilities" slogan.