Saturday, July 21, 2012

Probabilities, miracles, and design

I do believe that a case can be made for the claim that the Christian story about what happened in the life of Jesus makes more sense of the evidence than any possible naturalistic story, so long as you are willing to allow for a God who might do such a thing. 

But here Hume keeps coming back. If you base probabilities on what is most frequently found in nature, and you don't introduce the possibility of a non-human designer, then it looks as if the frequency of dead people who stay dead defeats anything but "extraordinary" (read virtually impossible) evidence for, say, a resurrection. 

But, the believer responds, causing the miracles in the life of Jesus seems like something a God might do. It makes sense from a theistic perspective, as opposed to, say, claiming that God caused a bunch of people to hallucinate, or caused a bunch of people to propagate a hoax that would ultimately result in them ending up on the kind of cross that Jesus was crucified on. 

But, the reply goes, likelihoods about what a divine agent might or might not do can't be brought in. They aren't based on experience, the way, say, the frequency of dead people who stay dead does. If you bring God in, you play a wild card. Anything goes. 

But, the theist replies, we can draw inferences about possible divine designers from analogy to human designers. 

That's why I think Lydia McGrew's paper on design and probabilities is relevant to this whole debate, which I linked to a few days ago.

26 comments:

BeingItself said...

"so long as you are willing to allow for a God who might do such a thing"

Is this a joke?

Bradley C. said...

SO in regards to the first paragraph, it seems to be missing an assumption. Even on the assumption God exists, it is still incredibly unlikely that he would facilitate Jesus' miracles and resurrection unless Jesus was actually God's son.

Because of this, it seems question begging to use things such as the resurrection to argue for the truth of Christianity, because the only worldview under which that event isn't prohibitively unlikely is a Christian one.

B. Prokop said...

"It seems question begging to use things such as the resurrection to argue for the truth of Christianity, because the only worldview under which that event isn't prohibitively unlikely is a Christian one."

It seems question begging to use such things as the 2012 Olympics to argue for the existence of sports, because the only worldview under which that event isn't prohibitively unlikely is one in which sports exist.

rank sophist said...

SO in regards to the first paragraph, it seems to be missing an assumption. Even on the assumption God exists, it is still incredibly unlikely that he would facilitate Jesus' miracles and resurrection unless Jesus was actually God's son.

Lewis's trilemma deals with this contingency. If Jesus's claims about being God's son were faithfully represented in the New Testament (the main point on which people attack the trilemma), then Jesus was either lying, insane or telling the truth. The first two options seem out of the question.

Further, a real resurrection seems like the best historical explanation for countless later events. Even if we determined that Jesus's claims were being misrepresented, the resurrection seems somewhat more likely than the other options--particularly if we believe that our world can be made to operate in such a strange way. As a result, God's existence still massively increases the probability of this explanation compared to the other, semi-conspiratorial ones.

BeingItself said...

"then Jesus was either lying, insane or telling the truth"

Or he was just mistaken.

And why are lying and insane out of the question?

Bradley C. said...

To B. Prokop: I don't understand the analogy. We pretty much know the probability of the existence of the 2012 olympics as well as the probability we live in a world that supports the existence of sports.

An analogy (NOT a comparison): I saw a video of a guy who claimed to get all his nourishment from the sun. He said he hadn't eaten in 20 years, he just charged himself up. A doctor was there verifying his claims. For this video to be true, we have to live in a world where a) it is possible to get nourishment from the sun and b) it is somewhat probable this man would know how to do it. On the other hand, the evidence that he does in fact get nourishment this way could be used to support a and/or b. It doesn't go both ways though. You can't assume a and b are possible to lift up the prior probability of the claims, then turn around and say that his being able to eat sunlight proves a and b are correct.

We know that plants use photosynthesis to convert the sun's light into useable energy, and solar cells do the same for electricity. that means that in theory, we could say "a" is correct. B, however, seems incredibly unlikely, given our current knowledge of our bodies, the sun, and energy. We also have overwhelming evidence that people die when they don't eat. This all lowers the probability that the video I saw was telling the truth.

If we assume that A and B are at least somewhat probable, we would need a lot less evidence to accept the claim, but we have no reason whatsoever to believe that is the case.

The situations don't map one to one, but my point is this. If you argue from the resurrection to the truth of Christianity, you can't presuppose the christian framework to artificially inflate the prior probability.

Papalinton said...

"Lewis's trilemma deals with this contingency. If Jesus's claims about being God's son were faithfully represented in the New Testament (the main point on which people attack the trilemma), then Jesus was either lying, insane or telling the truth. The first two options seem out of the question."

Lunatic? Liar? Lord?

No. That trilemma has been long superseded, thanks to biblical scholarship undertaken without the baggage of Apologetics.

Lunatic? Liar? Lord? or Legend?
The fourth point is powerfully shaping up in confirmatory research.

Crude said...

SO in regards to the first paragraph, it seems to be missing an assumption. Even on the assumption God exists, it is still incredibly unlikely that he would facilitate Jesus' miracles and resurrection unless Jesus was actually God's son.

This seems incorrect, or at least is aimed poorly.

If all we can be certain of is something like 'God exists' (let's define God as an omnipotent, omniscient being), then something like a resurrection is at the very least possible. But likely or unlikely? That's a statement about God's nature beyond mere power and existence.

So no, it's not "incredibly unlikely that He would facilitate Jesus' miracles and resurrection unless Jesus was actually God's son". All God's existence gets you to - and I believe, all Victor suggests that God's existence gets you to - is the possibility. Which in turn gives you a reason to investigate, and possibly believe, a resurrection claim.

There's no "presupposing the Christian framework" in this case. The closest Victor gets to determining what God would or would not do is his suggestion that analogies to human designers may or may not be appropriate. But even there, that's not presupposition of Christianity - that's just analogical reasoning based on evidence.

Crude said...

B, however, seems incredibly unlikely, given our current knowledge of our bodies, the sun, and energy. We also have overwhelming evidence that people die when they don't eat. This all lowers the probability that the video I saw was telling the truth.

Also, this seems really badly stated. A was already shown to be problematic by you, rightly, on relatively tame grounds ('photosynthesis'), and B is way too open-ended, since technological explanations are in principle possible.

To use a comparison: if tomorrow a headline reads "Scientists devise way for man to get necessary nutrients from the sun", can you say "Well, we've studied humans for years, and those who do not eat, get no nutrients" in response and write it off?

Mr Veale said...

Vic

I have found that this paper by Hambourger complements Lydia's beautifully.

http://philosophyfaculty.ucsd.edu/faculty/rarneson/Courses/hambourgerphil1reading.pdf

Rather than using frequency data, it uses "conceptual" considerations to predict/detect God's actions. Well worth the read.

Graham

Bradley C. said...

To Rank Sophist: I agree. If the trilemma holds it provides some other evidence for Jesus' claims apart from the resurrection (Though he could have been mistaken). Personally, I agree with the critiques that argue that Jesus' claims weren't entirely faithfully represented in the gospels... I don't think God's existence massively increases the probability of the resurrection however, only makes it possible.

Bradley C. said...

Crude: I apologize, I did rush that off a little bit yesterday, so I will try to clarify. Our positions might not be so far apart though.

I agree that the existence of God simply makes resurrection possible. I would hope you would also agree that even under this assumption, resurrections don't seem to happen very often. I am not making a frequency argument here, just an observation. Resurrections aren't the typical thing that happens when somebody dies. Furthermore, most resurrection claims come with the claim of divine revelation.

Jesus, according to the gospels, made strong claims about God and His nature. Those claims were much more likely to be believed if he could be shown to be of miraculous character. The resurrection, therefore, does make an assumption about the character and motives of God. Namely, that he endorses Jesus' teachings. If He didn't, it would still prohibitively unlikely that the resurrection occurred, and we are back where we started.

I admit Jesus didn't HAVE to be God's son, I was wrong there. If Jesus never claimed to be Gods son, or meant it in some metaphorical fashion, God still may want to endorse his teachings.

to be continued (to clarify the analogy)...

B. Prokop said...

Bradley,

I find the language you use to be rather bizarre. Even if you're not a Christian, it simply makes no sense to say something like "God ... endorses Jesus' teachings". According to Christianity, Jesus is God. so what you're really saying is "God endorses Himself".
So when discussing Christianity, one ought to use language that at least has some relevance to the subject.

And as to using the Resurrection to argue for the truth of Christianity, how on Earth can that possibly be "question begging"? That is basically the entire argument! It's how Paul argued in Athens (and elsewhere). Heck, it's how Peter argued on Pentecost, a mere 50 days after the event. In fact, the Resurrection was what convinced the Apostles that they hadn't been following a fake for the past three years. 2000 years later, it remains the one, critical issue. Do you believe that Jesus rose from the dead, or not? Rather than "question begging", it is The Question.

Bradley C. said...

Crude:
"All God's existence gets you to - and I believe, all Victor suggests that God's existence gets you to - is the possibility. Which in turn gives you a reason to investigate, and possibly believe, a resurrection claim." I agree with this, though I really don't think there are too many people that wouldn't agree (I could be wrong here, close-mindedness is pretty wide-spread).

On your second comment, let me try to restate a bit. in the analogy, granting proposition "a" is the equivalent of granting the possibility of God raising Jesus from the dead. Even if this possibility of gaining sustenance from the sun is granted, it is still incredibly unlikely for this man to actually be able to do it, unless we grant the probability that he has some knowledge that nobody else in the world has. Why would we want to allow for that without good reason?

The evidence for his feat would have to be VERY strong (not EXTRAORDINARY, but more than for a less fantastic claim) in order for me to believe that he is actually doing what he says he is. I absolutely wouldn't write off a scientific discovery that says we can get our nutrients from the sun, though I would be skeptical until I was at least able to read the paper on it, and I would really want to see repeatability of the results. (See the recent kerfuffle with neutrinos going faster than light speed)

The McGrews (I think it was them) have written an article evaluating the resurrection from a bayesian perspective. I think this is the right way to go about it (though I disagree with some of their methodology). By looking solely at the evidence we have, they try to demonstrate how low your prior probability would have to be to reject the resurrection hypothesis. I think this works better than manipulating your priors in such a way as to lower the amount of skepticism warranted.

Bradley C. said...

B. Prokop:

What is wrong with saying God endorses His own teachings?

All I mean is that raising Jesus from the dead is a way to prove that what he says is true. If God didn't want people to believe what Jesus said (whether or not Jesus is God) he would not have raised him. Why is that controversial?

B. Prokop said...

I didn't imply (I hope) that it was controversial - just that it's a very odd way of phrasing what you're trying to say.

Crude said...

Bradley,

I would hope you would also agree that even under this assumption, resurrections don't seem to happen very often.

They're certainly closer to moon landings than, say, landing on Florida's shore, sure.

Jesus, according to the gospels, made strong claims about God and His nature. Those claims were much more likely to be believed if he could be shown to be of miraculous character. The resurrection, therefore, does make an assumption about the character and motives of God.

The specific, Biblical claim of resurrection, absolutely. But the possibility of resurrection, full stop? It doesn't.

Say you're a bare theist. You have belief in God, but that's it - no doctrine, maybe a quasi-deism. If you hear a report of a resurrection, you wouldn't necessarily rule it out right away, even with your bare theism. You'd have reason to investigate this claim, or at least not write it off entirely from the outset.

Now, in the course of the investigation you may start inferring something about God's character, etc. But you don't need any commitments re: that character to take the prospect seriously at the start. That's all.

Even if this possibility of gaining sustenance from the sun is granted, it is still incredibly unlikely for this man to actually be able to do it, unless we grant the probability that he has some knowledge that nobody else in the world has. Why would we want to allow for that without good reason?

This just doesn't seem right. Was quantum physics "incredibly unlikely" circa 1900? The ideas certainly ran right up against the prevailing (fundamental, even) scientific knowledge at that point.

Let's run with Plantman for a bit. Let's say people observe him for months on end. Sunlight and water is all he's taking in. He's healthy. He's alive. Are the odds that he's really living off the sun now different? Or were the previous stated odds just plain wrong?

Now, I can appreciate skepticism of the claim. But I think there's a lot more going on with that skepticism than some appeal to the odds. I think it's a lot messier than that. You bring up the neutrino example, but there's one important thing to remember: lots of people were skeptical, but most of the people who were skeptical had very little knowledge of the scientific theories and the reasons why a FTL neutrino was just incredible. They were just running with the crowd. (I'm talking layman here.) Should they have been agnostic on the issue?

Also, I don't think it's right to talk about 'manipulating your priors', as if this is some intentional thing. One's priors are what they are, and that seems to be all there is to it for the most part. I'm a lot closer to Victor's... I think he called it subjective bayesian view, where you acknowledge your priors from the start, as well as the priors of others.

Bradley C. said...

B Prokop: I am sorry, I didn't mean to misinterpret you.

Crude: I think I agree with you for the most part. I would never suggest dismissing the claim out of hand without investigation, merely that a high amount of skepticism is warranted, even on basic theism. The investigation could change one's mind, but I often hear arguments go along these lines:

Skeptic: There is simply not enough evidence to justify such an extraordinary claim.

Believer: Well if you allow that God might exist, then you have to allow for miracles such as this one. You are just dismissing it based on your naturalistic presupposition.

This suggests that Skeptic has unreasonable priors, which prompted my comment about manipulating them. I meant it in a purely mathematical sense, nothing bad.

Are you suggesting that I think we should throw out anything that is highly unlikely? I don't mean to imply that at all. Quantum physics is a great example of something that seems unlikely but ended up being true. Relativity is another. I would even throw in feats like Philippe Petit's tightrope walk across the twin towers in that category. I think we have every right to doubt such things, until something comes along to change our minds on them (video & eyewitness in one case, repeatable experiments in the other two).

In the case of plantman (who is a real person by the way, I think you can find the video on youtube) I don't believe we need to be agnostic because we don't know that it can't happen. Significant observation could certainly change our opinion, but for now I am content believing he is a liar or a lunatic ;-). The stated odds aren't plain wrong, just subject to revision with more information.

Doug Benscoter said...

Probabilities are less relevant when it comes to decisions made by personal agents. It's unlikely that if you put one green ball in a lottery with a trillion red balls that the green ball will be chosen by chance. However, if someone chooses to rig the lottery, then the probability greatly increases. This is why the existence of a cosmic designer makes the resurrection of Jesus more plausible than it would be otherwise. Jesus would be unlikely to rise from the dead by solely natural means. But, if the designer of the laws of nature chooses to "fidget" with things, then there's nothing to prevent the resurrection from occurring. Whether Jesus did, in fact, rise from the dead is a discussion that concerns additional background information, and so forth.

B. Prokop said...

Doug,

What a great post! You've put your finger squarely on something that has always bugged me about all this discussion of probability. Of course the calculations are entirely different when we are speaking of free agents, rather than about purely physical events. The following two questions may linguistically be quite similar, but in their essence are nothing alike at all:

1. What is the probability of that volcano erupting tomorrow?

2. What is the probability of Joe Smith eating an apple tomorrow?

Doug Benscoter said...

Thanks, Bob! You might even modify (2) with an intentional equivocation: What is the probability of Joe Smith erupting tomorrow? :)

Papalinton said...

"It's unlikely that if you put one green ball in a lottery with a trillion red balls that the green ball will be chosen by chance. However, if someone chooses to rig the lottery, then the probability greatly increases. This is why the existence of a cosmic designer makes the resurrection of Jesus more plausible than it would be otherwise. Jesus would be unlikely to rise from the dead by solely natural means. But, if the designer of the laws of nature chooses to "fidget" with things, ...."

A-wishin' and a-hopin', and a-dreamin' ...... if, if, if. It is called teleological wish-listing, a sub-routine of the agency-detection function of the brain's 'theory of mind' process.

Bradley C. said...

"Probabilities are less relevant when it comes to decisions made by personal agents." This seems wrong. It certainly makes the probability harder to calculate, but not less relevant.

You can think of probability as how certain we are of any variety of outcomes to an event. Only one outcome obtains or obtained, so the probability doesn't reside in the event itself, only in our cognition of it.

We can use whatever information we have to make a prediction on how likely something is to occur / have occurred. The more info we have the better our prediction will be.

The fact that a personal agent could be involved could increase our uncertainty in all of the outcomes possible because we don't have access to the agent's mind. We can still calculate the probability though, the uncertainty just changes some of the numbers.

Doug Benscoter said...

Bradley, I think I'm with you on your first two paragraphs. With respect to the last one, though, that's where we might have a divergence of opinion. It's highly unlikely that the money from a safe would disappear all on its own. However, suppose that Jones's fingerprints are found on the safe. In this instance, the plausibility that Jones stole the money is higher than it would be in the absence of his fingerprints. We don't need to know what his motivations were in order to infer that he likely stole the money.

Bradley C. said...

There are two situations which I think are getting confused here. One is where there is an event that might have an agent or non-agent cause. The other is where there is an agent that can choose to actualize two or more outcomes to an event. Each case has different considerations.

Doug, your last comment seems to be referencing the first case, whereas my last paragraph is in reference to the second case (note the use of could).

I will try to lay out the probabilistic considerations in the case of Jesus's resurrection if I have some time later, because BOTH of these come into effect there.

Bradley C. said...

When looking at the resurrection of Jesus, what we have is a body of evidence: The NT documents, the rise of Christianity, etc. There are a number of competing hypotheses to explain the evidence we have. For simplicity I will separate it into 4 main hypotheses, but one of those (H3) is very expandable.

H1: God raised Jesus from the dead
H2: Jesus was raised from the dead by chance
H3: Some known naturalistic explanation (hallucination, legendary embellishment, swoon theory, mythicism… I just include them under one heading for convenience)
H4: Some yet unknown naturalistic explanation

All 4 of these hypotheses are consistent with basic Theism (T), but H1 requires it. I can see no reason that T would favor H1 over the other hypotheses up front, it simply doesn’t give us any information regarding the matter.

The McGrew paper gives a very good reason to favor H1 over H2. It seems quite unreasonable to assume H2 is probable, even if H3 and H4 are somehow ruled out. What it doesn’t show, however, is why H1 should be favored over H3 or H4, even given the assumption that T is correct. It seems the H1 would require some sort of assumption on the motives of the intelligent agent in order to garner support up front.

It is very much the case that the probability for any permutation of H3 is relatively low, and I suppose the evidence could be strong enough to effectively rule out H3 and H4, though I don’t think it is. If we have some background evidence that naturalism is likely to be true, this does increase the likelihood that H3 or H4 is true over H1.