# Christianity is True

Last time I wrote a long discussion of what I think is the best evidence for and against Christianity.

When it comes to the historical testimony to the Resurrection, I analyzed the data in terms of a set of minimal facts which I think even a "liberal" biblical critic ought to accept.   By the "liberal" view I mean someone who thinks that the Gospels, Acts, half the letters of Paul, and the general epistles were pretty much all written in the closing decades of the 1st century, by people other than their traditional authors—but that nevertheless the texts are based on identifiable earlier sources which go back to some real, and much earlier, written or oral traditions about Jesus.  (Because some parts of the Gospels are obviously rooted in actual historical events.)

This was a concession, not my actual beliefs: in reality I think this "liberal" view, despite its acceptance by the majority of secular biblical critics, is supported mostly by circular reasoning, and is more than a little reminiscent of the claim that Bacon wrote Shakespeare.  Nevertheless, if a good case can be made for Christianity even under such hostile circumstances, then obviously things would work even better on a more "conservative" view.  For a Bayesian approach based on a somewhat more conservative view of biblical criticism, see this paper by Timothy and Lydia McGrew (they argue for the conservative biblical view, but don't incorporate it into their probability analysis).

I also made no reference to the "inspiration" of the Scriptures by the Holy Spirit; rather I am treating them as ordinary historical documents containing what is claimed to be testimony.  Just to be clear, I do believe that the entire Bible is inspired by the Holy Spirit, and that therefore all of it is an important communication from God to us.  I disagree with fundamentalists in that I don't think the point of the Bible is to communicate scientific facts, or even matters of historical trivia.  For this reason it doesn't bother me that the Gospel accounts contain apparent contradictions: although most of these contradictions would probably be resolvable if we knew the full circumstances, it's not important to me whether if the documents contain mild inaccuracies regarding trivial matters, as all genuine eyewitness testimony does.

So now I have to analyze how I think the probability analysis actually goes.  To summarize the previous post, the main factors weighing against Christianity are:

(A1) prior probability due to specificity of Christian doctrine,
(A2) prior probability due to the weirdness of Christian doctrine,
(A3) prior probability due to Jesus being only one of billions of people.
(B) the Argument from Evil

while some of the factors weighing in favor are:

(C) the Fine-Tuning version of the Argument for Design.
(D) circumstantial facts of Jesus' life prior to his death, making him more likely to be the Messiah
(E) multiple testimonies to the Resurrection,
(F) modern-day miracles.

This is by no means a complete list, but I think it contains most of the highlights.  I've excluded some other arguments for Theism, and various Christian experiences, as less likely to be persuasive to someone coming from a skeptical point of view, even though I rate some of these things as significant.

Each of these factors leads to a shifting of the relative odds of Christianity : Naturalism.  (I'm assuming here that Naturalism is the main rival hypothesis to Christianity for skeptics of a "scientific" sort.)  Since these factors are approximately independent, each of them multiplies the relative odds by some number.  Since we aren't going to agree on precise ratios, I'm going to just estimate how persuasive these are to me, in terms of orders of magnitude (i.e. powers of 10).

I think that (B) and (C) are both good for about 2-4 orders of magnitude and therefore they approximately cancel each other out.  Each of them, despite being grounded in certain observable facts about the universe, is a controversial philosophical argument about metaphysics, and I think it's quite rare for any such argument to establish a conclusion with an error rate less than $10^{-3}$ or $10^{-4}$.  You could say that I'm skeptical about Philosophy, but that's not really fair since shifting the odds by a factor of 1000 should count as a "success" even if this can be easily overwhelmed by more definitive evidence.  For similar reasons I don't think (A2) should be much less than $10^{-4}$.

(A3) and (D) cancel each other out nearly exactly, because if Jesus is in a very small group of people more likely to be the Messiah, then it doesn't matter how many other people there are in the control group.   See the discussion of the prosecutor's fallacy.

(E) contains a factor which cancels out (A3) exactly.  That's because the Christian testimony is to Christianity specifically, and not to any other religion.  Whereas a made up religion is going to be selected from some large set of possible religions.  See the discussion of the Farmer Jones case in the comments of this thread.

The remaining part of (E) has to do with how plausible the testimony is, leaving aside its specificness.  Now the testimony of a single individual is has to be worth at least 2 orders of magnitude, because most of the time people don't lie about things.

Somewhat counterintuitively, the value of testimony towards implausible claims like miracle reports is actually somewhat greater than this, because people are less likely to make up claims that are implausible.  There are many people who claim to have crossed the street, but only a few people claim to be abducted by UFO's.  The number of people who claim to have seen people physically resurrected is quite low; certainly much less than $10^{-2}$ of the population.  To the extent that this effect exists, (E) also partially cancels (A2).  But this cancellation only works up to a certain point, since nothing says that the rate of false claims has to mirror the true a priori probabilities.  For the small fraction of the population who is willing to make absurd claims, there is no limit to how unlikely their assertions may be.  So some residue of (A2) may remain.

What remains to be considered in (E) is the gross improbability of seeing testimony as strong as this about anything false.  The multiplicity of eyewitnesses, the priority of the women, the apparent unanimity in the pre-existing group of Twelve (minus Judas), and the fact that they were willing to die for their testimony all weigh in its favor.  All of these facts are strongly supported by the historical data.  Their initial skepticism and doubt, mentioned explicitly in almost all of the Gospel accounts, weighs heavily in its favor, so the skeptic is probably better off trying to argue (somewhat less implausibly) that this was made up later to prove a point.  The lack of any reasonable ulterior motive also makes the testimony less likely given Naturalism.

St. Paul presents a special case: since we know he was previously hostile, his testimony is more or less independent of the others.  That in conjunction with the unusual circumstances surrounding his conversion is a factor of $10^{6}$ all by itself; the group testimony of the remaining apostles is at least $10^{8}$, and that's rather generous, since on that hypothesis we'd expect to find about 100 similarly strong examples of false dramatic claims by large groups, in the most recent generation of $10^10$ people alone.

Taking everything together I think that the implausibility of a set of testimony looking like (E) given Naturalism is at least $10^{-14}$, and much lower if one takes a more conservative approach towards biblical criticism.  If this is right, then (E) pretty much overwhelms every other factor in play, and (F) is just icing on the cake.  Therefore, with high probability, the Resurrection seems to have really happened.  In other words, (some unspecified version of) Christianity is far more probable than Naturalism is.

I am a postdoctoral researcher studying quantum gravity and black hole thermodynamics at UC Santa Barbara. Before that, I studied the Great Books program at St. John's college Santa Fe, and got my Ph.D. in physics from U Maryland.
This entry was posted in Theological Method. Bookmark the permalink.

### 103 Responses to Christianity is True

1. lavalamp says:

I'm still thinking about our other thread (haven't had large blocks of free time), but I don't agree with this analysis very much.

I think you're missing any term regarding the rarity of the claimed events.

More importantly, you glossed over the thing that I think loses the most probability, which is the reliability of the documents. Even under your "liberal" scenario (which I generally accept to be the case), the probability that the text is wrong is highest at the most unlikely claims. Even though I agree that a few of the texts date from the first century (Mark and/or Q, about half of Paul's letters), we don't actually have first century copies, and even the earliest dates for Mark (70 ad) are almost 40 years after the events they concern. 40 years is at least one and maybe two generations (average Jewish male's lifespan in the first century was possibly as short as 22 years according to one source I saw); it's quite possible that there weren't many people around remembering the events to keep the accounts accurate.

I do think it's likely that Paul believed in a resurrection, but it's not at all clear to me that he believed this for epistemically justifiable reasons. Paul was not an eyewitness and doesn't claim to be one; he claims to have seen a vision. I wouldn't believe (his interpretation of) a story like his if it was told to me in person, I'm certainly not going to believe it when it's gone through 50-150 years of manual copying by people motivated to make it say a particular thing. (Do you find the Mormon story of Joseph Smith's visions remotely persuasive?)

Ancient historians report fantastical events with some frequency, and from what I can tell, modern scholars disbelieve those bits with impunity. I feel like treating biblical accounts the same way.

Anyway, given the nature of these omissions, I strongly suspect that if you'd been raised in an Islamic environment and did this same sort of analysis, you'd find a strong chance of Islam being true.

2. g says:

You claim 140dB of evidence from (E). Words cannot express how wrong I think this is, but let me have a go.

The evidence for the resurrection -- the evidence available to us, that is, not some hypothetical evidence that might have been available to people 1900 years ago -- consists, basically, of some documents from the New Testament and basically nothing else. (Note: all the evidence you cite falls into this category.)

So your claim takes the following form. "There are these documents, derived from originals about 2000 years old. The existence of these documents constitutes 10^14:1 evidence in favour of the proposition that someone rose from the dead."

I don't quite want to claim that there is no possible way that any document in the real world could give 10^14:1 evidence for anything, because it's easy enough to construct counterexamples. But I do think it's about as obvious as anything can be that the New Testament documents do not provide 10^14:1 evidence for the resurrection. Actually, no, one thing is more obvious than this, namely that you haven't shown that the New Testament documents provide 10^14:1 evidence for the resurrection :-).

(There's a lot of overlap between what I'm saying here and what lavalamp said. Too bad.)

Let me sketch what, from the perspective of anyone not already committed to Christianity, we have here. A modest number of documents (let's say five: three Synoptics including Luke-Acts as one, John, and the few relevant fragments in Paul's letters), written decades after the events they purport to describe, plainly not independent (because the Synoptics are clearly working from one or more common sources or copying each other) but also far from consistent (there is scarcely anything in the resurrection stories, beyond the barest outline, on which they quite agree); written by partisans, in at least one case with the explicit goal of making converts (John says so in so many words). We have not the originals but copies of copies of copies of copies; for the most part we can make very good guesses at what the originals said, but there is pretty good evidence that the transmission process wasn't terribly high-fidelity (because there are so many minor disagreements, and some cases where texts have clearly been changed for theological reasons). As well as disagreeing with one another, these documents contain plenty of things that there's other reason to think probably wrong (if there'd been the sort of large-scale zombie sightings described in Matthew 27, we'd expect other sources to mention it; demon possession appears not to be a real thing, which makes it odd for Jesus to have cured so many cases; the dates of Herod and Quirinius don't match up right for the birth narratives; etc.) which suggests that their authors were, for whatever reason, not so careful as to make their historical claims very trustworthy.

This is not the sort of documentary evidence that can give you a factor of 10^14 for a single fairly simple proposition.

(You can get that sort of evidence for stupid contrived compound propositions like these: "At least 80% of the following propositions are true: There was a person alive in 20AD called Yeshua or something of the sort. He had a close acquaintance called Thaddaeus. He had a close acquaintance called Nathaniel. He went to Capernaum several times. He told a story along such-and-such lines: [...]. He told a story along such-and-such lines: [...]. Etc." What's difficulty is getting, so to speak, more than a certain "density" of likelihood ratio.)

For a more extended, more referenced, but also more oblique, version of this argument, see here.

3. Mom the linguist says:

Surely a moment's thought will tell you that even if the average lifespan is only 22 years, that hardly means that there are no 40 year-olds alive. There could even be a lot of them, depending on how much variation from the average there is. If practically everyone died at 22, then there would be nobody left to raise the children at all, which is unfortunately starting to be nearly true in some parts of Africa right now. Though even in those places, the children left behind are being raised largely by grandparents, who exist, though in much smaller numbers than the children they are raising. Average life expectancy is very much lowered by high child mortality. If more than half the people die before the age of five, you could still get a lot of people that are 50, 60, even 70 and 80 and still have an average of 22. Anyone who has studied classical history ore literature finds that there are plenty of old people around.

4. lavalamp says:

Don't forget that infants can't witness things. Actual witnesses couldn't have been too much younger than 60 at the time Mark/Q was written. I'm not saying there weren't any, but they were rare, and for us to have heard dissenting voices, they'd have to have been literate (extremely rare) and able to get their documents into the right hands for them to be passed on to us.

5. Mom the linguist says:

I'm not arguing with the statement that 40 years is a generation removed. It's when the eyewitnesses start dying that people figure they better start writing things down. But to claim 40 years is 2 generations removed from eyewitnesses based on "average life expectancy" is just bad math. It's like the argument that the apostle John couldn't possibly have lived to be a really old man like tradition says he did because "people didn't live that long back then", followed by a statement of average life expectancy, which proves nothing.. Eyewitnesses don't have to be literate either, just have access to someone who is. I don't think literacy was "extremely rare" in that sense. People who needed things written down (lots of people including the lower middle class) had ways of making that happen. Whatever you think of the early Christians, they were certainly good at getting their documents "into the right hands", as you say, mostly by distributing them very widely. It's amazing how many many copies we have of the NT books as compared to anything else of that age. There are thousands of NT manuscripts less than 100 years removed from their time of writing, as opposed to hundreds of the next best preserved thing (Homer). Most manuscripts of this time period exist in 10 copies or less.

6. lavalamp says:

Let me clarify what I'm claiming, since it's not very similar to what you're responding to. :)

I'm saying that the relative sparsity of actual eyewitnesses reduces the chance that errors in the written accounts (both accidental and deliberate) would have been detected and corrected. I'm also claiming it reduces the chance of *contrarian* accounts having been created and preserved (so the fact that we don't see any isn't necessarily much evidence that there weren't any).

7. g says:

It doesn't appear to be true that "there are thousands of NT manuscripts less than 100 years removed from their time of writing".

For instance, according to this page (written by a Christian and arguing that there are more early manuscripts than is commonly believed), "there are at least ten and as many as thirteen NT MSS that are possibly or definitely from the second century". These contain somewhat less than half of the NT. (About 43% of NT verses are found at least in part in these manuscripts, according to the same page.)

On the other hand, I'm not aware that there are any manuscripts of Homer dating to within 100 years of composition.

(I think there are thousands of NT manuscripts in Greek, and hundreds of Homeric manuscripts in Greek. Perhaps that's what MtL is thinking of?)

8. Mom the linguist says:

I admit I did a quick and sloppy check because I was running out the door. There are thousands of early manuscripts, the oldest of which are first century. The Homeric manuscripts are certainly farther removed because Homer is much much earlier than any of the other things being discussed. I believe that the oldest ones are at least 500 years later. Whether the New Testament documents are "accurate" in the sense of "corresponding to reality" is arguable, but almost everyone agrees now that they are accurate in the sense of "corresponding closely to what was originally written" with a few exceptions that are well known, like the end of Mark.

9. Aron Wall says:

Mom,
I believe you probably meant to say that there are thousands of (pre-printing press) manuscripts, some of which are less than a century old (of which the earliest uncontroversial example is a fragment of John from the early 2nd century.) But everything else you say is true, especially the general point that the statistics are much better for the NT than for other ancient documents.
[Note: this was written before I saw the comment just above.]

Why is that important? Because of this (g):

We have not the originals but copies of copies of copies of copies; for the most part we can make very good guesses at what the originals said, but there is pretty good evidence that the transmission process wasn't terribly high-fidelity (because there are so many minor disagreements, and some cases where texts have clearly been changed for theological reasons).

and this (lavalamp):

...we don't actually have first century copies...I'm certainly not going to believe it when it's gone through 50-150 years of manual copying by people motivated to make it say a particular thing.

Textual critics make their "very good guesses" (or should I say "scientific conclusions"?) by constructing an evolutionary tree and then figuring out where all the mutations happened. It's similar to what biologists do when they study the fossil record to work out the family relationships between different species of life. Because there are so many copies of the NT, this process is extremely reliable. So long as the mutations occur after the first branch point such that each branch has extant manuscripts, you can always tell that the mutation occured. (Telling which variant is the original and which the mutation is trickier: sometimes it's obvious, but other times it boils down to rules of thumb. This doesn't matter, because I don't plan to rely on any controversial assumptions about which variant is correct.) In the case of documents which were as widely copied as the NT, I would be surprised if the first branch point were more than a very small number of copies from the original. So the "fidelity" of the copying process is not really that relevant.

This is not the hill the skeptic wants to die on. You'll do much better to concede that we have in all essential details the original NT manuscripts, and instead stick to arguing that they aren't truthful as originally written. Here we come to textual criticism's sexier younger sister: "higher criticism", which seeks to analyze and date the sources that went into the original documents.

lavalamp:

More importantly, you glossed over the thing that I think loses the most probability, which is the reliability of the documents. Even under your "liberal" scenario (which I generally accept to be the case), the probability that the text is wrong is highest at the most unlikely claims. Even though I agree that a few of the texts date from the first century (Mark and/or Q, about half of Paul's letters), we don't actually have first century copies, and even the earliest dates for Mark (70 ad) are almost 40 years after the events they concern.

Actually, even "liberal" scholars usually date the other gospels (esp. Matthew and Luke) and several of the other books of the NT to the first century.

Normally the liberal scholars date Mark to almost exactly 70 AD. Do you know why? Because in it, Jesus predicts the destruction of the Temple at Jerusalem. It is assumed that predictive prophecy is impossible (and I suppose they also think that no one could possibly have seen it coming through regular foresight?), even though all the other evidence points to a somewhat earlier date. So they put it right at 70 as the best compromise.

Then they assume that Matthew and Luke, because they use Mark as a source, have to be 10 or 15 years later. (Although it hardly took that long to communicate across the Roman Empire.) That's how they date the synoptics.

Against this, we have at least three different reasons to date the Gospels much earlier. First, if the Pastorals were really written by Paul (and there must be some evidence for this since even some of the "liberal" critics have proposed that the supposed forger must have had access to some genuine letter of Paul), then since 1 Timothy quotes Luke, Luke and Mark must be prior to St. Paul's martyrdom in the 60's. Secondly, it's rather surprising that the book of Acts makes no mention of St. Peter or St. Paul's martyrdom if it was written after that, and of course Mark and Luke were written earlier than Acts. Finally, we have the writings of St. Papias (disciple of St. John), St. Iraneus, and the Muratorian Canon which state explicitly that the gospels were written by their traditional authors, which would of course require them to be within the lifespans of those individuals.

Anyway, in my post I assumed the liberal view for the sake of argument (although any implausibilities in this view further reduce the probability of Naturalism). That's why, instead of arguing directly from the Gospels and Acts, I assumed only a small set of "minimal facts" which are supported by multiple distinct lines of evidence in the New Testament. Several of the facts I mention in (E) have support in BOTH in Paul's undisputed letters AND either Acts or multiple Gospel "sources", which is a very cynical standard of evidence. Which if any of these "minimal facts" are you prepared to deny? If none, then the reliability of other aspects of the manuscripts is irrelevant.

g:

Let me sketch what, from the perspective of anyone not already committed to Christianity, we have here. A modest number of documents (let's say five: three Synoptics including Luke-Acts as one, John, and the few relevant fragments in Paul's letters)

When I open my Bible, I also see James, 1-2 Peter, 1-3 John, Jude, and Revelation. Admittedly the evidential support and relevance of these books are not all equal, but these books are also claimed to be of apostolic origin, and I won't allow them to be dismissed without argument.

lavalamp:

Paul was not an eyewitness and doesn't claim to be one; he claims to have seen a vision.

Huh? Where are you getting this from? The ordinary meaning of "Resurrection" in 1st century Judaism involves physical bodies, not "visions". In his 1 Cor 15 list of witnesses he makes no distinction between what happened to him and the other apostles, except as to the matter of timing. If you are using Acts as your source for this claim, you have to reckon with the fact that the other people there were affected as well. (St. N.T. Wright also has some things to say about "visions")

lavalamp:

(Do you find the Mormon story of Joseph Smith's visions remotely persuasive?)....
I strongly suspect that if you'd been raised in an Islamic environment and did this same sort of analysis, you'd find a strong chance of Islam being true.

In the case of Mormonism and Islam, the historical data is extremely well explained by the assumption of a single unscrupulous liar. Once you eliminate the testimony of the main prophet, there's little left to build any kind of case on. (Joseph Smith did try to assemble a team of witnesses for the golden plates, but within a few years the living witnesses pretty much all recanted, so that hardly counts for much.) Both Joseph Smith and Mohammed gained from their testimony, rather then being prepared to die for it. (Smith was shooting at his assailants when he was "martyred" in jail.) There are no records of formerly hostile witnesses becoming eyewitnesses, as there are in Christianity. The differences here are numerous and obvious.

g:

plainly not independent (because the Synoptics are clearly working from one or more common sources or copying each other) but also far from consistent (there is scarcely anything in the resurrection stories, beyond the barest outline, on which they quite agree)

So when the Gospels agree with each other, that's a reason to disbelieve them. And when they disagree, that's also a reason to disbelieve them. Interesting.

It has been noted that the Resurrection accounts in the Gospels have a greater degree of independence then the other parts do. That combined with the apparent contradictions makes it clear that they go back to different sources from each other. The fact that they do agree in broad outline then implies strong historical support for the "minimal facts" of that outline. As for the contradictions, aren't there always minor contradictions in eyewitness testimony? (Also, many of the supposed "contradictions" in lists circulated by atheists are quite silly, such as things mentioned by one author and omitted by another.)

g:

written by partisans, in at least one case with the explicit goal of making converts (John says so in so many words).

In other words, if the people who saw the evidence found it convincing, and expected that it might convince other people too, then their testimony must be discounted for that reason. Also interesting.

g:

if there'd been the sort of large-scale zombie sightings described in Matthew 27, we'd expect other sources to mention it

The text in Matthew is obscure, but I interpret it as most likely implying that a small number of Christians known to St. Matthew reported temporary visits by certain resurrected individuals other than Jesus. I'm not sure why you expected this to be reported by others as well? The whole Jesus thing overshadowed it, I guess. Anyway, this is a fairly weak argument from silence. By the way, there seems to be some extrabiblical support for the
crucifixion darkness and earthquakes.

demon possession appears not to be a real thing, which makes it odd for Jesus to have cured so many cases

You're showing some provincialism here. There are many parts of the world where possession by evil spirits is fairly common. Of course, the Christian and the Naturalist may disagree as to whether these evil spirits really exist, or whether these people are just some other sort of crazy. But leaving aside the explanation, it's simply a fact that some people really are sadly afflicted in this way, and that exorcisms sometimes cure them. There's a couple of case studies in psychologist St. Peck's People of the Lie.

Of course, it's quite possible that the limited knowledge of people in 1st century caused them to misdiagnose cases of e.g. epilepsy or schizophrenia as demon possession (the demon-possessed boy in Mark 9:14-29 seems to me to really be an epileptic; interestingly Jesus says it is a different "kind" than usual).

g:

the dates of Herod and Quirinius don't match up right for the birth narratives; etc.) which suggests that their authors were, for whatever reason, not so careful as to make their historical claims very trustworthy.

You're really picking at straws here. Suppose that the histories are accurate. The probability that, after 2000 years, there would be some apparent discrepencies is nearly 100%. For example, Luke 3:2 states that there were two High Priests at the time that John the Baptist started his ministry. This is an obvious contradiction with Jewish ritual, indicating that St. Luke has no idea what he's talking about. Or at least, that's what people would say, if we didn't know the following additional fact: that the Romans had deposed Annas in favor of his son-in-law Caiaphas, but that the Jews still refered to Annas as High Priest out of respect for his former position.

The article I linked to by the McGrews mentions several notable cases where things that scholars originally took to be inaccuracies in Acts and the Gospels, proving that they were written later, were confirmed by later archaeological discoveries. It's reasonable to believe that the other "inaccuracies" could mostly be resolved in the same way, if we knew everything. Note that I would say the same thing about historical sources outside the Bible, if they seemed reliable otherwise.

g:

The evidence for the resurrection -- the evidence available to us, that is, not some hypothetical evidence that might have been available to people 1900 years ago -- consists, basically, of some documents from the New Testament and basically nothing else. (Note: all the evidence you cite falls into this category.)

That's right. Well, there's also a passage in Josephus, but many scholars think it must be an interpolation by Christians, because of how strongly it seems to support Christianity. (This game is rigged; it's not possible to win it.)

g:

I don't quite want to claim that there is no possible way that any document in the real world could give 10^14:1 evidence for anything, because it's easy enough to construct counterexamples. But I do think it's about as obvious as anything can be that the New Testament documents do not provide 10^14:1 evidence for the resurrection.

I don't think it's at all unusual for historical documents to confirm facts with exponentially large probabilities, especially when there are multiple sources. Remember that the probability of being wrong falls off exponentially with the amount of evidence, so a factor of $10^{14}$ isn't necessarily that immodest. It only takes 46 "bits" of evidence doubling the odds ratio to get there. I think the probability that Julius Caesar wasn't really assassinated is probably lower than $10^{-30}$. If I'm wrong about this, it's because I'm wrong about the strength of historical testimony in general, not because of special pleading in this case.

In any case, what matters is the relative weight of my (primarily) historical evidence for (D & E), versus your (primarily) philosophical evidence against (A & B). What reason to do have to believe that Philosophy settles questions more definitively than History does?

lavalamp:

I think you're missing any term regarding the rarity of the claimed events.

What did you think (A3) was? By definition, if Christianity* is true, the Messiah event occurs just once. (A3) is a tiny number which keeps track of that rarity, equal to 1 over the total number of people who have ever lived. It ends up being cancelled by (D), some actual historical evidence for the specialness of Christ.

10. lavalamp says:

Wow, really? 10^-30 is an amazingly tiny number. I am not certain about anything at that level, with the possible exception of mathematical proofs with small numbers of steps. I am certainly not *that* sure that Julius Caesar was assassinated, and I think it's ridiculous to be that sure. (I'm about 55% sure he was, but that could go up as high as 95% if I had studied the relevant documents.) We accept physics discoveries at 5 sigma, which is something like 10^-6. If you think we're billions of times more certain about history than physics, I'm not sure what to say.

More later, maybe, but I think this is our major disconnect.

11. lavalamp says:

Hm. It's hard to keep track of what comes from where when you learned all this stuff as a kid. You're right, I accept Corinthians as more likely to be accurate than Luke/Acts. And you're right that Paul uses the same language about Jesus appearing to him and the rest of the crew. In English, anyway, that passage doesn't have to be read with a bodily resurrection in mind, though ("appeared" is ambiguous). Even if you can find a passage in one of Paul's (authentic) letters that unambiguously requires a physical, bodily appearance, I'm still going to have trouble accepting Paul as an eyewitness because he didn't see Jesus both before and after (how does he know what he saw was Jesus?). If you can somehow overcome that objection, I'm willing to give two orders of magnitude (20dB) from Paul's testimony.

I don't think it counts as extra that he died for what he believed, that's bayesian evidence that he believed what he believed strongly, not that what he believed was true. I also don't believe it's a one in a million event for people to drastically change their minds, which is good for you, because if it were, you'd have to supply evidence that it actually happened, as an assertion in an ancient text isn't going to be sufficient for me. (IOW, you're treating this as a +60dB event, when I would treat it as more like a -20dB event (+20dB each for the assertion in Paul's letters and Acts). As it is, though, I see it as something like +10dB for Paul sincerely believing, which is not the same as him being correct.)

I'm also willing to give you another 20dB from Mark. I keep talking about Mark because it's earliest and I don't believe the others are independent, and even though the resurrection story isn't in Mark, it came from *somewhere* and I'm feeling generous. ;)

So, that adds up to +40dB of evidence from the bible for the resurrection, assuming you can convince me that Paul should count as an eyewitness. That leaves it at something like -50dB. (0dB is of course 50%, so more is required to make me actually believe this.)

Oh, that reminds me, I don't think A3 adequately captured prior improbability of the event. I guess I agree it's necessary, you do need to locate your messiah among all the people that ever lived, and I agree that the bible is sufficient for that purpose. I'm talking more about believing that anything miraculous happened in the first place.

Sorry, this is kinda rambly.

12. lavalamp says:

One other thing, even though I should be sleeping. I'm not sure how you get 20-40dB of evidence from the fine-tuning argument. I don't think it's bayesian evidence either way, since P(we'll find ourselves in a universe we can live in) is near 1 by definition, no matter what. But if I hadn't thought of that, I'd expect you to need to do some calculation figuring the proportion of possible universes given our physics (10^500 unless string theory has changed since I last heard) which can evolve life. I'm not sure if we can figure that out without simulating some amount of them to see what happens?

If anything, the fact that we seem to have evolved in this universe is bayesian evidence that there's no god-- it's an essential feature of the world given naturalism and an optional feature given theism. It could even be strong evidence if you believe there's lots of possible ways god could have created us without evolution. The nature of abiogenesis could still conceivably be questioned, however, so I won't demand that you count this. :)

13. lavalamp says:

Ah, a bit more.

So the "fidelity" of the copying process is not really that relevant.

If you're claiming a -90dB event happened, and the copying process corrupts things at the ridiculously low rate of 1 out of a million things, a copy error is 1000x more believable than the -90dB thing. You can't just handwave low probability things when you're trying to establish an extremely low probability thing.

Several of the facts I mention in (E) have support in BOTH in Paul's undisputed letters AND either Acts or multiple Gospel "sources", which is a very cynical standard of evidence. Which if any of these "minimal facts" are you prepared to deny? If none, then the reliability of other aspects of the manuscripts is irrelevant.

Believing the things you listed under E is not a binary thing. I'll give all those things +20dB for each independent source, so I'll believe them if that's enough to overcome their prior improbability. (And, btw, I feel this 20dB per source thing I've been doing is extremely generous given the route the text has taken to get to us.)

I think there's a disconnect. You're using Bayes theorem to evaluate christianity given a set of facts. I'm using it to evaluate your facts. If your "facts" aren't facts, there's nothing to explain.

14. g says:

On textual criticism: I have no particular wish to die on any hill, and I see no way to get from "this is far from the weakest point in Christians' text-based arguments" (with which I certainly agree) to "I should concede that the NT texts are accurately enough preserved for everything Christians want to do with them" (with which I don't necessarily agree). I will agree that probably for a sizable majority of locations in the text we know with (say) 90% probability or better what originally went there. But we know that, even early in the textual history of these documents, there were accidental errors and deliberate changes in some copyings, and I don't see that we are in a position to know with very high confidence that all the particular points on which a Christian might want to rely when arguing for the Resurrection have been correctly transmitted.

(In order to tell whether any such issue actually arises, we'd need to see an actual argument for how the NT documents as we have them provide the alleged 140dB of evidence that the Resurrection happened.)

An argument that Luke must be early because the Pastorals are genuinely by Paul and they quote Luke seems to me pretty weak, especially if the only evidence you're going to offer for an early date for the Pastorals is that some liberal scholars say their author must have had access to a genuine letter by Paul.

It looks as if when I said "three Synoptics including Luke-Acts, John, and Paul" you thought I was attempting to list all the books of the NT. I have no idea why, but plainly I was unclear and I'm sorry about that. I was listing the documents that could be claimed to provide evidence for the Resurrection (beyond the observation that the Christians believed in it, which of course some of those other books corroborate -- though the first one you listed, the letter of James, doesn't even do that). I was attempting to "dismiss" them only in the following sense: I don't think they say anything relevant to the question of whether the Resurrection happened.

(Aron's comment on this was addressed to lavalamp and not to me, but I'll respond anyway.) The only descriptions we have of Paul's alleged encounter with Jesus are in Acts. They do not, even slightly, look like descriptions of meeting a physically resurrected, embodied being. We're told that there was a bright light from heaven, that Paul was struck blind, and that he heard a voice. We're told that his companions heard the voice but saw no one. (And later we're told that Paul said that his companions heard nothing but saw the light. Obviously we're not dealing here with a very reliable source.) No one (including Paul's companions, who you will recall were not struck blind) is said to have seen a person. I'm not sure that "vision" is the right word for what is said to have happened to Paul, but it certainly doesn't sound at all like the same sort of thing that happens at the end of the gospels and the start of Acts.

I did not say "when the stories agree, that's a reason to disbelieve; when they disagree, that's also a reason to disbelieve". I am not a complete idiot, and you will arrive at more accurate interpretations of my meaning if you make use of this fact in cases of doubt. (The same heuristic would have led you not to assume that I didn't know that James etc. are in the NT, too.) When you are presented with a number of pieces of testimony, they are more valuable if they are of independent origin (100 identical newspapers are little more evidence than one; 10 different newspapers all citing the same Reuters report, likewise), they are more likely to be right if they agree on at least the essentials of the story and preferably somewhat more, and their details are more likely to be right if they agree in detail too. But some kinds of agreement -- e.g., outright textual copying -- are strong evidence of non-independence, so no such principle as "more agreement is always good" is correct. In the particular case of the NT documents allegedly providing evidence for the resurrection, we have rather few of them and there's good evidence that they are not all independent (e.g., near-identical wording in many places; I forget whether any of these are in the resurrection sections, but the point is that the Synoptics are known to have common sources); and they disagree a lot on points of detail (including some that might reasonably be thought important, such as who it was that first went to the tomb and found it empty; all the gospels agree that it was Mary Magdalene plus zero or more other women, but no two agree on which other women if any).

The point about contradictions between the accounts is not -- at least, not when I make it -- to say "see, they don't agree about everything, therefore they're all just made up". It's to say "see, they disagree about just about everything but the bare outline, which means that we can't trust any specific detail that they describe because they're known not to be accurate at that level of detail, which I predict is going to make a lot of trouble for any argument that tries to get 140dB of evidence out of what they say". Because I've seen several such arguments before, and they all involve things like "we know that X didn't happen, because Matthew says there was a Roman guard by the tomb", and if we don't actually know that then the argument is greatly weakened. (Note: that was a randomly chosen example and I am not in fact asserting that Jesus's body was stolen.)

I remark in passing that it's foolish to caricature someone else's argument as "when the stories agree, that's a reason to disbelieve; when they disagree, that's also a reason to disbelieve" and then immediately afterwards, in the same paragraph go on to make an argument of your own that could be caricatured with at least equal fairness in the exact same way, with only the obvious change of sign.

The point about the documents being written by partisans is this: all the textual evidence we have for the Resurrection comes from writers who plainly were fully convinced of the Resurrection before they started writing; it looks to me, though of course it's impossible to know at this historical distance, as if they were fully convinced of it before they started looking in earnest at the evidence they recorded. Works of advocacy for religions (and indeed for other kinda-religion-like things) almost-universally present whatever evidence they have in the most favourable light. They very commonly don't present whatever evidence there might be on the other side, even if their authors know of it. They sometimes just lie outright. That puts a limit on how much we can rely on the evidence they present.

Now, to be sure, that means that if (e.g.) Jesus really was raised from the dead, and if the roughly-contemporary evidence for this was incredibly strong, so that anyone who looked into it properly would have become a Christian and started writing works of advocacy rather than impartial history, then it's not possible for textual evidence now to be as convincing as you might like. (And the same goes for any other proposition that would tend to make advocates of its believers.) Perhaps that's unfair, but the unfairness isn't created by skeptics who point it out; it's just a genuine weakness of historical textual evidence of this kind.

I am not claiming that "philosophy settles questions more definitively than history does"; that would be far too broad and sweeping a claim. History tells us nothing about whether divine-command theories of ethics are any good. Philosophy tells us nothing about whether the Romans invaded Britain in 53AD. And of course in the present instance some of the questions are mixtures of history (broadly conceived) and philosophy. For instance, an Argument From Evil is a philosophical argument, but it draws much of whatever strength it has from a bunch of empirical facts about suffering and death and error and misery and so forth. And also from other facts (which aren't exactly either historical or philosophical) about the attributes the god under discussion is supposed to have.

I agree that historical documents can sometimes provide quite a lot of evidence (though 300dB for Julius Caesar's assassination seems outrageously optimistic to me). However, it seems to me that there are several "skeptical" hypotheses about what happened around the time of the crucifixion that aren't disfavoured by anything like 140dB relative to any "real-resurrection" hypotheses by the existence of the texts we have. I think that if we're to take the claim of 140dB of evidence seriously, we need to see an actual argument for that claim.

15. g says:

Oops, I just noticed a few things Aron wrote that were directed to me and to which I failed to respond. In order:

Matthew's zombies: if many bodies came out from their tombs near Jerusalem and went into the city to say hello, wouldn't you think that would be the sort of thing others in Jerusalem might take notice of? If only because of the large-scale grave-robbery there would seem to have been? I agree that this is far from conclusive evidence that it didn't really happen (though, for me, it doesn't seem that much positive evidence is needed to conclude that a zombie uprising didn't happen). It's just one small example among many of how the gospels are full of the kind of stories that, generally, we find best explained by saying that someone made something up, or imagined something, or was deceived.

No, I don't think my comments on demon possession show provincialism. I am well aware that there are places where demon possession is claimed to happen. I think the people who make that claim are wrong. (I'm sure you don't think that every instance of the schema "Most people in population X believe Y; Z says they are wrong" is rightly described as provincialism; in particular, I'm sure there are many such instances that put you in the place of Z.) Again, you are free to disagree, and to think that the best explanation of claimed demon possession is actual demon possession; but I suggest that from the point of view of someone who is trying to decide between Christianity and naturalism (and it is presumably to such a person that arguments like yours are addressed) the fact that a document attributes things that -- as you say -- look a lot like epilepsy, schizophrenia, etc., to demon possession, is evidence against the reliability of that document (and of any character recorded therein who endorses that attribution). And yes, Jesus does say there's something a bit different about the case of "demon possession" that you say is probably epilepsy. What he says is different is that that sort of demon "cannot be driven out by anything but prayer", which doesn't seem to me like it describes epilepsy any better than it does other problems of the kind that the gospels attribute to demons, nor like it describes epilepsy better than it would actual demon possession.

In case it isn't clear, the most important thing about these cases of "demons" in the NT isn't that the NT authors misdiagnose them. It's that they say that Jesus diagnosed them as demons, and typically treated them by, e.g., giving commands to the demons.

It's hardly "grasping at straws" to point out a respect in which the NT documents appear to be historically inaccurate. Do we actually, as you say we do, know (or even have strong evidence) that it was common to refer to Annas as if he were still high priest? So far as I can tell -- but on the basis of very limited investigation, and I'm very willing to be proven wrong -- there's no actual evidence for this other than the fact that that's a way to save Luke 3:2 from being wrong. (I don't think it's at all implausible, and of course I wasn't the one who introduced that particular passage into the discussion. I just want to get a better idea of what standard of evidence you require before saying that we "know" something like this.)

The McGraws' article gives some examples of things found in the NT for which there was no historical evidence, which some critics said were probably false, and for which evidence later turned up. OK, fine. What does that have to do with things in the NT against which there is historical evidence? The two situations seem quite different to me. Perhaps there are examples of the latter in the McGraws' article and I missed them; I looked through it only quite cursorily. (The portion I found that treats of such matters begins "And the conjectures of the form critics ..." and goes on for a few paragraphs. Is there more?)

It is a gross misrepresentation of the arguments against the authenticity of the "Testimonium Flavianum" to say "many scholars think it must be an interpolation by Christians, because of how strongly it seems to support Christianity"; whether they're right or not, their justification is not nearly so simplistic or question-begging. (For instance, they argue: 1. that it not merely supports but assumes something like Christianity even though Josephus is known not to have been a Christian; 2. that some of its usage of words is very unlike that of Josephus elsewhere; 3. that lots of early Christian apologists (before Eusebius, in whose writings the earliest text of this passage is found) make reference to Josephus but none of them cites what he says about Jesus.)

16. Mom the linguist says:

Anyone who thinks that there is only a 55% probability of Julius Caesar really having been assassinated given the historical evidence obviously has set their priors for historical documents really low. It seems like arguing about specific historical documents ought to be put aside for now and the question of accuracy of history in general needs to be addressed instead.

17. lavalamp says:

55% is an estimate of my own uncertainty. If I examined the evidence, it could go up or down, probably up if what you guys are saying is correct. I haven't updated this to reflect the evidence from your opinions because, well, it seems like Aron at least is wildly over-confident in the one collection of historical documents I *do* know a little bit about... If Aron's assessment of the matter is correct, then 95% would be my actual estimate.

18. Aron Wall says:

lavalamp,

A lot of our disagreement seems to be about the weight which should be assigned to historical documents in general. Bickering about the details of the NT won't help us very much until we get that straightened out. As you say:

Wow, really? 10^-30 is an amazingly tiny number. I am not certain about anything at that level, with the possible exception of mathematical proofs with small numbers of steps. I am certainly not *that* sure that Julius Caesar was assassinated, and I think it's ridiculous to be that sure. (I'm about 55% sure he was, but that could go up as high as 95% if I had studied the relevant documents.) We accept physics discoveries at 5 sigma, which is something like 10^-6. If you think we're billions of times more certain about history than physics, I'm not sure what to say.

$10^{-6}$ is regarded as a minimum for accepting a result in phyiscs (in certain fields). That doesn't mean that this is a typical level of confirmation in Science. Rather, scientific results are frequently even better than this! For example, if the 5 sigma result is replicated independently a hundred times in laboratories across the country, then (assuming that the systematic errors are independent in each experiment) the result is confirmed up to an error of $10^{-60}$.

If we turn to biology, and you asked me what are the odds that the theory of evolution is wrong (I mean "just plain wrong", not an approximation with limited validity, or something normally true with some exceptions somewhere), then I think the odds of us being wrong have to be less than $10^{-1000}$. (Of course, at a certain point skeptical hypotheses like we're really in a simulation, or people have been lying to us all our lives, may become more likely ways out, but I'd like to ignore that sort of thing here.)

So you see, your "billions of times more certain" comment is misleading. Getting to $10^{-14}$ only takes 2.3 times as much evidence as getting to $10^{-6}$ does. That's why there must be many historical propositions which are certain up to an error of $10^{-30}$. Suppose you find a single manuscript which states that a person named King Alfred existed in some time period, and suppose you conclude on the basis your overall sense of the reliability of that manuscript that there's a 99% chance that he really did. Well, if there's any person for whom there's 15 times as much (independent) evidence as King Alfred, then you should be $10^{-30}$ sure that he existed. And if you have a person like King Henry VIII or Abraham Lincoln who leaves records in a lot of different places, then the odds should skyrocket up to gazillions.

Now, if you want to consider really weird hypotheses like Abraham Lincoln actually had a twin brother who assassinated him and took his place, then of course the records aren't really independent because they were all tricked for the same reason. But that's so crazy it should itself be assigned a miniscule probability, since it still stacks many different improbabilities together.

And that's why your assignment of 95% probability to Caeser's assassination is still laughably low. That's equivalent to saying that there's a 1 in 20 chance of some really weird conspiracy theory being true, involving a giant coverup of an event having multiple witnesses and recorded in multiple contemporary sources. Do you think that 1 out of every 20 times a public figure is reported to be killed in the presence of numerous witnesses, there's actually been a successful conspiracy to conceal the cause or fact of death? Would you really be no more surprised to discover this, than to roll a "20" on a d20? I don't think so.

19. Aron Wall says:

I also don't believe it's a one in a million event for people to drastically change their minds

You're downplaying some rather important features of Paul's experience here. Would you also say that my hypothetical of

Hitler converting to Judaism after being struck by lightening, and then later being accepted into the Jewish community and becoming a highly respected rabbi.

is no more implausible than any other instance of someone "dramatically changing their mind"?

I admit that if we go by the book of Acts, there's some significant differences between what happened to Paul and what happened to the other disciples. It's true that there's nothing in the text which explicitly says that Paul saw the form of Jesus, although 9:7 seems to imply it by contrast with his companions. It says explicitly that he saw a bright light, fell down, that Paul had a conversation with the voice, and was blinded for 3 days. It also says that the people around him were affected by the "vision" (9:7 vs. 22:9 is only a contradiction if you assume the point is to distinguish between the senses of sight vs. sound, rather than that they experienced something indistinct but didn't have the same experience that Paul did.)

Even if you don't accept Acts and just go by Paul's letters, you won't get a lot of these details. However, it seems clear that something stronger happened than just Paul drastically changing his mind about something. Something specific happened to him, which (1) he interpreted as seeing the resurrected Jesus, and (2) caused him to stop murdering Christians and try to join their community instead. Call it a "vision" if you like, but it's not at all parallel to what would happen if I persuaded you, or you persuaded me.

Assuming Naturalism, how often should such "visions" be expected to happen? You think it's generous to give Paul's conversion a factor of 100. Well let's see. If such events happened 1 out of 100 times, that would mean that there should be about 70 million people in the world right now who have also totally re-oriented their beliefs as a result of a dramatic vision in which a person seems to speak to them and tell them things. Do you think that 1 out of every 100 people you meet is such a person? If not, you have to rank the probability lower than that. $10^{-6}$ is equivalent to there being about 7,000 people in the world today who report experiences similarly dramatic to Paul's. Do you think that it happens more frequently than that?

Yes, the fact that Paul and other apostles were martyred for their testimony only helps show that they were confident and sincere, not that they were correct. The degree to which this is relevant depends on which alternative naturalistic scenario you try to construct. If you want to argue that he was a liar, it's highly relevant. If you want to argue that he was crazy, it's not as relevant. You haven't yet committed to a particular alternative explanation, you see, so I have to cover all the bases.

It may help if you explain in more detail what you think is the most likely naturalistic alternative scenario. For each "minimal fact" I mentioned, you'd have to decide whether the fact is true (in which case you have to explain why it is true) or false (in which case you have to give an alternative explanation for the evidence of that fact).

20. Mitchell Porter says:

Now add up the number of people who have been tortured to death during human history, and calculate the probability that God is good.

21. lavalamp says:

You think it's generous to give Paul's conversion a factor of 100.

Er, I think experiences like Paul's are something on the order of 1 in 1000, interpreting "vision" rather broadly. However, I accept the acts story and Paul's own epistles as at least somewhat separate, and give them +20dB each, so I do think it's likely that Paul experienced some event that caused him to change his mind. You haven't convinced me that he should count as an eyewitness, though, and I don't see a reason why I need to accept his interpretation of his life-changing vision as the actual truth behind it. People are mistaken about the things they experience with some frequency.

If you really think that such a drastic change is a 10^-6 event, then I will put a non-negligable amount of probability mass on the idea that Paul's pre-conversion persecution of christians has been exaggerated.

It may help if you explain in more detail what you think is the most likely naturalistic alternative scenario.

I don't feel the need to explain away -70dB events. (That number being the prior probability of the resurrection, + 20dB for the synoptics.)

Moving on,

For example, if the 5 sigma result is replicated independently a hundred times in laboratories across the country, then (assuming that the systematic errors are independent in each experiment) the result is confirmed up to an error of 10^−60.

When you come up with numbers like that, obviously if it's wrong it's because there's some unknown systematic error.

Don't forget that I'm throwing out these certainty numbers for *me* individually. I make mistakes understanding people at *far* greater rates than 1/10^-60. At numbers like that, lots of unlikely things start looking quite probable. Like world-wide conspiracies of physicists.

And that's why your assignment of 95% probability to Caeser's assassination is still laughably low. That's equivalent to saying that there's a 1 in 20 chance of some really weird conspiracy theory being true, involving a giant coverup of an event having multiple witnesses and recorded in multiple contemporary sources.

The further back in the past some PR cover-up story was invented, the less able we are today to tell the difference. Caesar's death is nowhere near as well documented as stuff like Lincoln or Kennedy. About those, I do give the various conspiracy theories less than 1% chance. But not less than 10^-6, I think.

Let me put it this way: I think you are wildly underestimating your own uncertainty, and the scope of possible errors in the information you've encountered. Like a true bayesian, all I can do to prove this is to offer to bet. It's not much use betting on history, as new information is rare. But if you can come up with some number of physical laws or constants that you're sure about at 10^-60, I will happily bet that they're wrong at only 10^6:1 odds (give or take, depending on what it is).

Would you really be no more surprised to discover this, than to roll a "20" on a d20? I don't think so.

Actually, if you can list 20 events that you're equally sure about, I'll give 50:50 odds that one of them is wrong or at least debated by historians, and we'll know this in the next 50 years.

If we turn to biology, and you asked me what are the odds that the theory of evolution is wrong (I mean "just plain wrong", not an approximation with limited validity, or something normally true with some exceptions somewhere), then I think the odds of us being wrong have to be less than 10^−1000.

10^-1000? What!? Words cannot express how overconfident I think that is (literally; I tried and deleted it because I couldn't think of anything large enough). I think young-earth creationism *alone* is vastly more probable than 1/10^-1000 (god/satan planted the evidence), not to mention all the other various weird alternative explanations for the evidence we observe. I'd say something more like 10^-6, and this figure is probably constrained not by the evidence, but by my own personal ability to examine the evidence.

(Of course, at a certain point skeptical hypotheses like we're really in a simulation, or people have been lying to us all our lives, may become more likely ways out, but I'd like to ignore that sort of thing here.)

That's sort of my whole entire point, you absolutely *cannot* ignore those "weird" theories when you're dealing with such enormously tiny/large probabilities.

In studies, people who are 99% sure about things are sometimes actually wrong 40% of the time. "Once subjects had been thoroughly warned about the bias, they still showed a high degree of overconfidence." What evidence do you have that you're so much better calibrated than the average human that you can throw around such huge certainties? Bayes' theorem is not enough, you also need to be certain that you did the math right and that you did the right math and considered every possible source of error...

22. Aron Wall says:

lavalamp:
If you go back and re-read my sentence about evolution, you'll see that I inserted an important caveat:

(Of course, at a certain point skeptical hypotheses like we're really in a simulation, or people have been lying to us all our lives, may become more likely ways out, but I'd like to ignore that sort of thing here.)

I'm sorry if this was confusing, but I actually meant this parenthetical as an anticipatory concession that the point you make here is right:

10^-1000? What!? Words cannot express how overconfident I think that is (literally; I tried and deleted it because I couldn't think of anything large enough). I think young-earth creationism *alone* is vastly more probable than 1/10^-1000 (god/satan planted the evidence)

When I said I was ignoring "that sort of thing", I meant to ignore things which totally invalidate our knowledge, like people lying to us our whole lives (God or Satan planting the evidence is a hypothesis of this kind, I think). Of course, you're quite right that these skeptical hypotheses may be more plausible than $10^{-1000}$, but I meant this figure to be something more like "probability that the data would appear by chance, without a skeptical hypothesis of this sort" (or for that matter, us just being really stupid when we do the Bayesian analysis).

If I said something like "the probability that a (heads/tails) coin will land heads every time when flipped 1000 times is $2^{-1000}$," you would know what I mean by this, even though in reality some other hypotheses (like a circus coin-flipper who can always get it to land the same way by some trick) is vastly more probable. I meant something like, the probability of the evidence appearing this way by chance is this low. So I'm not really disagreeing with you here. My point was just that sceientific theories are supported by many, many times more evidence than 5 sigma (conceding that, due to skeptical hypotheses, this evidence doesn't really stack completely independently).

I hope that with this explanation for my comments, your probability that I'm incredibly stupid has gone down a bit :-).

23. Aron Wall says:

g writes:

Works of advocacy for religions (and indeed for other kinda-religion-like things) almost-universally present whatever evidence they have in the most favourable light. They very commonly don't present whatever evidence there might be on the other side, even if their authors know of it.

It's quite fascinating then, that several times the gospels do include evidence which, at first sight, appears to contradict Christian theology about Jesus. Two of the many examples are Jesus' cry of forsakenness at the cross, and his refusal to accept the title of "good teacher" from the rich young ruler in Mark 10:18. The presence of such examples makes it more likely that other features of the text are accurate.

Unfortunately, I don't have the time right now to respond to all your other points in as much detail as I'd like, but I think it's fair to say that you place the bar for accepting the New Testament accounts so high that it's unreasonable to expect, even if it really happened, that anybody of ancient literature could clear that hurdle (all ancient histories contain some apparent discrepencies with other things we know, they always have some motivation to say what they're saying, it can seldom be proven that the claimed set of witnesses to an event were independent, the textual criticism issues are far worse than for the NT, et cetera). You seem to admit this explicitly here:

Now, to be sure, that means that if (e.g.) Jesus really was raised from the dead, and if the roughly-contemporary evidence for this was incredibly strong, so that anyone who looked into it properly would have become a Christian and started writing works of advocacy rather than impartial history, then it's not possible for textual evidence now to be as convincing as you might like. (And the same goes for any other proposition that would tend to make advocates of its believers.) Perhaps that's unfair, but the unfairness isn't created by skeptics who point it out; it's just a genuine weakness of historical textual evidence of this kind.

If you aren't even willing to accept the sort of documents which you think would be produced if the event really happened, then there's really no point in continuing to discuss the details. Your a priori prejudices are too severe to overcome with historical data (at least from the ancient world; we haven't discussed modern miracles yet).

Let me observe here something about the rhetoric of this conversation. You and lavalamp are trying to make it seem like I'm being cocky and overconfident by assigning so many orders of magnitude to a set of ancient historical texts. I've been trying to argue that this isn't as unreasonable as it looks, because of the mathematical fact that the probability of being wrong scales exponentially with the amount of evidence available for that proposition.

But let's take a step back and look at this from another angle. Suppose we ask how much evidence is necessary to make Christianity $50%$ probable. I wouldn't need so many orders of magnitude in (E) to get there, unless you had already claimed the exact same number of magnitude evidence against Christianity coming from other sources (The Argument from Evil, plus whatever a priori factor you assign to Christianity being weird). I think that you are being unjustifiably cocky to assign these things so many orders of magnitude.

In other words, you're skeptical about History, but you're quite confident about what God would or would not do if he existed. While History can't prove anything with 100% certainty, I'm far more confident of my abilities to find out what really happened by studying ancient texts, then of my abilities (as a backseat driver) to tell how God ought to have created the universe.

24. Aron Wall says:

lavalamp writes:

Oh, that reminds me, I don't think A3 adequately captured prior improbability of the event. I guess I agree it's necessary, you do need to locate your messiah among all the people that ever lived, and I agree that the bible is sufficient for that purpose. I'm talking more about believing that anything miraculous happened in the first place.

If it's because miracles are weird, that goes in (A2). Seriously, I've given you a very detailed model (with three different a priori factors) for analyzing the probabilities, and I also explained why I think (D) and (E) include factors which cancel out (A1) and (A3), based on our discussion of the Farmer Jones example. You seem not to accept some of these cancellations, but you haven't really explained why not, except in vague terms.

I don't feel the need to explain away -70dB events. (That number being the prior probability of the resurrection, + 20dB for the synoptics.)

Huh? What do you mean, you don't feel the need to explain away "-70 dB events"? Surely you accept that there has to be some naturalistic explanation for the records, if Christianity isn't true? My submission is that if you take the time to imagine what it would take to produce these records naturalisitcally, you'll see that a grossly improbable assumption comes in somewhere. The question is whether this is sufficient to overcome your sizable a priori prejudice which (if I understand what you're saying correctly) is $10^{-9}$. (Not sure where you got this number, but if it's related to the rarity of the claimed event happening relative to the total number of people in the world, I've already explained several times why I think that factor is cancelled out by (D).)

But even if I grant you a prior probability of Christianity at $10^{-9}$, it seems that you're willing to give me 2 additional orders of magnitude for the synoptics and maybe 3 for Paul's vision, getting us down to $10^{-4}$ or so. Great! That means I'm about half-way there to convincing you that Chrisitianity is more probable than not. If it turns out that the evidence for some set of modern-day miracles is equally good or better to the NT evidence, then it seems that the remaining gap here might be closable. Agreed?

25. Aron Wall says:

By the way, if it wasn't clear from my previous comments, I'm trying to wrap this particular discussion up soon, since I need to spend at least some of my free time thinking about other things. (Regrettably, I find myself obsessing about how to make the best possible replies, when I should be thinking about other things.) I'm grateful you came over here to express your opinions, and hopefully we'll be able to make further progress on specific issues in future posts.

I'm not sure how you get 20-40dB of evidence from the fine-tuning argument. I don't think it's bayesian evidence either way, since P(we'll find ourselves in a universe we can live in) is near 1 by definition, no matter what. But if I hadn't thought of that, I'd expect you to need to do some calculation figuring the proportion of possible universes given our physics (10^500 unless string theory has changed since I last heard) which can evolve life. I'm not sure if we can figure that out without simulating some amount of them to see what happens?

If anything, the fact that we seem to have evolved in this universe is bayesian evidence that there's no god-- it's an essential feature of the world given naturalism and an optional feature given theism. It could even be strong evidence if you believe there's lots of possible ways god could have created us without evolution.

With respect to the Fine-Tuning Argument, I think it's a bad idea to discuss this in depth at the bottom of a comment thread about something else. Let me just make some brief comments. With respect to the string theory multiverse: 1) string theory is not currently supported by any experimental data, 2) even if it were, this would not immediately imply that the $10^{500}$ possible universes of the string landscape actually exist somewhere, 3) how to reason about probabilities in the context of a (possible) multiverse is extremely controversial, and all approaches that I know of lead to horrible paradoxes. It's not obvious to me that the multiverse should count as an explanation for anything.

Of course the probability that you exist given that you make some observation is 1. But that doesn't mean that your existence doesn't call for any explanation, because it's still true objectively considered that you might not have existed at all. So if you have a prior over the space of possible laws of nature etc., you should still update that prior based on the information that you exist. If you grew up on a desert island without knowing who your parents were, that wouldn't make it reasonable to assume you don't have any parents, on the grounds that the probability of you existing is 1 anyway. If you face a firing squad and end up surviving, you still have to come up with a explanation (e.g. the execution was faked) even though if you had died you wouldn't be around to ask the question.

Yes, it's possible that God could have arranged things so that life could only evolve or exist through continual supernatural intervantion. But conditional on theism, I think the probability that he would choose to use natural means is $\sim .5$. Whereas, the probability that a single naturalistic universe (with randomly chosen constants) will produce life is exceedingly low. So the standard Fine-Tuning Argument is much better evidence than the reversed-polarity argument you mention.

26. g says:

Aron,

"Difficult" things in the NT like Jesus's cry of dereliction are interesting evidentially. They make it less likely that the NT authors both (1) were what's now considered doctrinally orthodox and (2) were making stuff up rather than reporting sincerely. But if they're accurate then of course they also make it less likely that (3) what's now considered doctrinally orthodox is correct. What they do, overall, to Pr(Christianity) probably depends on exactly how you define "Christianity".

I'm not sure what you mean by "clear that hurdle". I'm fairly sure I don't place the bar so high that there could never be good documentary evidence for anything 2000 years ago. (For instance, I think there's pretty good documentary evidence that there was, at least, a real person on whom the NT stories about Jesus are based; full-blown Jesus-mythicism seems improbable to me.) But I do place it high enough that it would indeed be very surprising if we had documentary evidence strong enough to justify (what Christianity requires) a very strong commitment to an otherwise very improbable proposition, such as that Jesus was killed and then miraculously raised from the dead.

You can call this "a priori prejudices" if you insist, but that seems to me unfair and uncharitable. It's not my fault how difficult it is to get really good evidence from historical documents. And of course I could be wrong about that; maybe, in this particular case, the evidence really is as strong as you say. I'm willing to be persuaded. It seems like a question that could be addressed by actually looking at that evidence, considering some rival hypotheses, etc. (Unless, of course, you find it more convenient to say that anyone who currently disagrees with you on this point has insurmountable a priori prejudices and isn't worth arguing with, which would be disappointing.)

In other words, I am not convinced that "there's really no point in continuing to discuss the details".

I think we will have a more productive discussion if we concentrate on the actual issues rather than on whether anyone here is "unjustifiably cocky". If you think I've been trying to make you look bad rather than focusing on the issues, then (once again) I apologize for doing so if I have been (which I don't think I have, but who knows?) and for appearing so in any case. Anyway, of course I agree that the amount of evidence required to make Christianity 50% probable is less than the amount needed to make it (say) favoured 99:1 -- only about 7 bits less, of course -- but I'm not sure why that makes me any cockier than I would have been otherwise.

I wouldn't say that I'm globally skeptical about history and globally confident about my ability to tell what God would do if real. I'm skeptical about some things you might want to do with history, that's all.

Perhaps it would be useful for me to give a brief summary of where I think the various big odds ratios come from. Let's consider three candidate positions: "conservative" Christianity (roughly speaking, what an evangelical Protestant and a traditional RC could reasonably be expected to agree on; probably not a million miles from your position, Aron), "liberal" Christianity (say, that of David Jenkins, former Anglican bishop of Durham), and scientific naturalism (roughly my position). I'll give triples (C,L,N) of log-odds-ratios, expressed in dB; obviously (C,L,N) means the same as (C+k,L+k,N+k) for any k, and it turns out I'm taking N=0 in each case. This is all necessarily rough and simplistic, not least because the various things here are far from being independent. I'll try to err on the side of generosity towards Christianity.

Mere specificity, together with the fact of specific testimony: I agree that these kinda-sorta-roughly cancel, but plainly some complexity penalty remains (testimony or no testimony, traditionalist Roman Catholicism is less probable than Christianity, which is less probable than theism, which is less probable than non-naturalism, simply because of specificity differences). Let's say something like (-20,-15,0). L does better than C because it's vaguer.

Non-naturalism in the face of the very impressive extent to which the world appears to behave exactly as predicted by a bunch of precise scientific laws: (-30,-10,0). L does better than C because it has much less "intervention" and can more credibly argue that God generally lets the universe run according to its own rules.

Weird doctrines (incarnation, atonement, etc.): maybe (-20,-10,0) after accounting for mere non-naturalism. Depends a bit on exactly what we put in this bucket.

Evil, divine silence, etc. (note that my preferred versions of these arguments differ somewhat from the "if God exists then there would definitely be no suffering in the world, ever" strawmen sometimes addressed by theists): somewhere around (-40,-40,0); L is maybe better than C in being able to claim more consistently that God has a general non-intervention policy, but maybe worse in that it's more committed to the idea that God is nice.

General "quality" of the Bible (fulfilment rate of prophecies, quality of poetry, insight into human nature, wisdom of laws, etc.): (-20,-5,0). L does better because it's less committed to a strong view of inspiration.

Biblical evidence for the resurrection: (+10,+5,0). I think I'm being very generous here, mostly because I can't be bothered trying to deal with figures less than 5dB and I think C should do strictly better than L here because on L there maybe isn't anything much to explain, so to speak.

Assorted arguments for Christianity (fine tuning, apparent design of living things, alleged excellence of Jesus's character and teaching, modern miracles, etc., etc., etc.): after a bit of estimating and tallying, I reckon this is about (+5,+2,0). I don't think any individual argument makes more than about 5dB of difference. I've tried to be quite generous to Christianity in this reckoning.

Of course there are other factors, but those'll do for now. Adding them up (I hope correctly), I get (-115,-73,0). I repeat: just adding them up is of course terribly simplistic. On the whole, these things probably correlate positively in the obvious sort of way -- if I've missed some reasons why arguments from evil are no good, I'm more likely also to have missed some reasons why various Christian doctrines are more sensible than they look, etc. -- so purely handwavily the "right" figures for me might be more like (-70,-50,0) or something. Maybe as low as (-40,-40,0) if actually everything else gets completely "absorbed" in the biggest figure; maybe as high as (-100,-65,0) if there's more independence than one would think. L is substantially more probable than C, but still very improbable and also much less interesting :-).

You'll notice that there are no 140dB blockbusters there; there's no single point I'm anywhere near as (over?)confident on as you are of the documentary evidence for the Resurrection, and there are no obviously-unclosable gaps. But I'd need to see some high-quality evidence or arguments, not previously known to me, to close those gaps. You could call that "a priori prejudice" if you want, but I do hope you won't.

27. g says:

(I wrote all that before seeing Aron's remark that he was trying to wrap up this discussion. My apologies, Aron, if it leads you to spend more time doing things he doesn't actually consider a good use of that time...)

28. g says:

... doing things you don't actually consider ...

(The perils of rephrasing things.)

29. imho says:

Wow... I can barely wrap my head around a religious Physics PhD. I'm not trying to be rude, I really am having a hard time understanding how any objective scientist could reconcile Physical laws with magic. BTW, your analysis misses the only factor that matters... That magic is utterly incompatible with the laws of Physics.

Prior to his resurrection, Jesus walked on water. Please explain what laws of physics Jesus broke in order to perform this feat. Did he adjust the masses of all the particles in his body such that he became lighter than water. Or perhaps he forced a macroscopic violation of the 2nd law of Thermodynamics causing the water to momentarily freeze then unfreeze beneath his feet as he walked. Or maybe he temporarily suspended gravity in that particular location.... I can only imagine what laws would need to be broken to resurrect then float away into the sky.

So what precisely do you believe? That the true laws of physics contain special God terms, allowing him(her) to tune said laws at will? That these God terms are completely undetectable in every experiment ever, but the proof of said terms can be found in 2000 year old books?

It seems to me that a belief in miracles (magic) requires a disbelief in the Laws of Physics as presently understood. How do you reconcile these things?

30. g says:

imho,

It doesn't seem obviously absurd to say that (1) the universe normally operates according to fixed laws, but (2) occasionally God (who, being God, can do as he pleases) causes those laws to be broken a bit in order to bring about some special purpose. If it's occasional enough, then that breakage won't show up in laboratory experiments. (Nor if God goes out of his way to avoid having his tinkering show up in experiments. That would be a pretty horrible bit of special pleading, but it's one with some precedent in the Christian scriptures.)

I do think the fact that the universe seems to operate according to fixed laws *is* evidence against "magic" (of whatever sort, religious miracles included) -- but it doesn't seem to me to be the strongest evidence against, say, Christianity. (I don't think Aron's exactly missed it; it fits into the category he calls A2. But it's true that he hasn't said anything about it specifically.)

For the avoidance of doubt: I am not Aron; his views surely differ somewhat from mine. Like you, I'm an atheist.

31. lavalamp says:

My point was just that sceientific theories are supported by many, many times more evidence than 5 sigma (conceding that, due to skeptical hypotheses, this evidence doesn't really stack completely independently).

I hope that with this explanation for my comments, your probability that I'm incredibly stupid has gone down a bit :-).

OK, I think we're on the same page here, and I did note your paranthetical, I just think that you can't ignore all the weird theories and expect to get the right answer. And I don't think you're stupid, or I wouldn't be leaving comments. (Do you have any idea how hard it is to find someone who both knows what Bayes' theorem is and doesn't already agree with me??) :)

But even if I grant you a prior probability of Christianity at 10^−9, it seems that you're willing to give me 2 additional orders of magnitude for the synoptics and maybe 3 for Paul's vision, getting us down to 10^−4 or so. Great! That means I'm about half-way there to convincing you that Chrisitianity is more probable than not. If it turns out that the evidence for some set of modern-day miracles is equally good or better to the NT evidence, then it seems that the remaining gap here might be closable. Agreed?

5 orders of magnitude is actually a little more than I'd be comfortable granting from the historical evidence, I meant for it to add up to 4 orders of magnitude. Still, yes, if there weren't any evidence against Christianity you would indeed be something like halfway to making the resurrection seem 50% probable to me. (I picked 10^-9 as a prior because surely resurrections happen less frequently than that, so it seems like an upper bound to me, and upper and lower bounds seem like the best you can do if you're essentially doing Fermi estimates with Bayes...)

As for modern day miracles, I'm actually willing to grant you 10dB from the Fatima thingy; it's the only thing I've investigated that seems difficult to explain (although, now that I write this, perhaps I should apply that only to Catholicism). To accept more modern stuff, at a very minimum, I'm going to require video evidence that professional stage magicians can't explain. The fact that this stuff isn't on youtube with hundreds of millions of hits is evidence to me that it doesn't exist.

I actually like g's breakdown a fair amount. I can add a few.

Lack of souls. It's extremely difficult to form a coherent Christianity without souls and without turning god into a simulator (at which point, IMO, you've ceased talking about Christianity). I'm quite confident that there aren't souls, almost as confident as I am about evolution. So to me, this seems like -20dB or so against all Christianities.

Trinity. You could lump this in with weird doctrines, but I think asserting logical impossibilities deserves a mention of its own. However, Christians could misunderstand their own religion, so I'll count this as only -5dB or so.

Fine tuning; again, I still think it's +0dB. I don't know how to sum over all possible worlds, and neither do you, and without that, we don't know if theistic or naturalistic life bearing universes are more common. It's not clear that positing a god helps any. (What kind of universe does the god inhabit, and why is it so fine-tuned that it produces god(s)?) (Also, I take the position that any universe that can be simulated must exist, for the same reason that the 10 billionth digit of pi will be the same for all possible observers who calculate it. But even if you don't buy this, surely it's still valid to imagine these possible universes as some sort of guide to the probability of our "real" universe?)

Hm. I went back to read the thing I wrote (for my folks to explain my deconversion) to recall the rest of the arguments that convinced me. Then I thought, why not just post it? It's not explicitly written in the form of a Bayesian argument (I tried to make it accessible, I was aware of Bayes when I wrote this but not fluent-- footnote 130 is *painfully* wrong), but it still serves as a decent summary of my position. It's long. I think it's funny. If history is any guide, it's not an entirely safe read for a thinking Christian wishing to remain so.

https://dl.dropbox.com/u/412445/apologia.pdf

32. lavalamp says:

Perhaps as a politer version of imho's comment, I am curious. Do you see god as working within the laws of physics, as a creature of the universe? Or do you think he works from outside the laws of physics, and may change the state of the universe at his whim (as I thought while a Christian)? Is there a third option?

In other words, if I try to make up the most reasonable version of Christianity possible, I find myself needing to choose between space alien and simulator for god; both seem antithetical to Christianity, and I can't think of another option that doesn't fit in either category and doesn't also seem very improbable (e.g., the universe doesn't actually run on math).

I'll go ahead and say that lack of a good answer for this costs theisms at least 10dB in my mind. I think I derive a lot more evidence from philosophy-type arguments than you do, because I consider them separately.

33. Mom the linguist says:

First briefly on the matter of souls: If by soul you mean an immaterial something that lives in our bodies and can be detached from it and become or ghost, or go live in some immaterial heaven, Christianity does not require this belief, nor does the Bible teach it (though there a handful of passages, mostly in poetry that are kind of like this, and one difficult passage in I Samuel possibly featuring a ghost). Neither the Old or the New Testament uses the word usually translated "soul" in this way. Psyche in Greek, and whatever the Hebrew word translated by psyche in the Septuagint is (I'm not a Hebrew scholar, but have checked personally with one), is associated with breath, and refers to things that are immaterial in a way, like speech, thoughts, attitudes, personality, but that have roots in our physical selves. You might characterize psyche as an epiphenomenom of our physical selves without upsetting me in the slightest. Animals have psyche too, in the Bible. The "soul" idea to which you refer, which admittedly most Christians who haven't thought deeply about the matter probably hold, comes partly from the "ghost in the machine" hypothesis, which is very naively plausible, and has therefore been held and debated over centuries, and the particular folklore of early modern European cultures that still holds sway, based on a melange of deism, theosophism, and Victorian sentiment. Mind you, I'm kind of fond of Victorian sentiment, but that doesn't mean I believe in it. Even as a child, it was pretty clear to me that all those "Christian" fairy tales about child angels and so on were dreadfully inconsistent with what the Bible actually says if you sit down and read it. The Bible is all about physical resurrection, with a physical setting that includes plants and animals and rivers and stuff. Yes, Christianity requires life after death, but actually not immaterial souls. I think the psyche is part of our pattern, which includes our physical selves, and God presumably has to recreate this and sustain it in a somewhat different medium eventually, but you'll find some discussion of this in I Corinthians 15. I suspect that what was characterized as "God as simulator" isn't that different from what I and many other Christians actually believe--that we live in a medium created and sustained continually by God's love, attention and will. Philosophically it's hard to see what the difference between things "really" being there and being there to every possible sensory test would be if it's God that's sustaining it.

34. g says:

FWIW I agree that Christianity doesn't strongly require souls. (I was for many years a believer in Christianity and a disbeliever in souls. I know at least one person who is still both of those things.)

35. lavalamp says:

MtL, that's certainly not a response I expected! Clearly you have spent some time thinking about it.

It seems there's a lot of stuff you have to discount to hold this, though, no? Off the top of my head, in addition to the witch/Samuel passage, there's a reference to the souls of the righteous in Revelation, all the cases of demon possession, and the parable Jesus tells about the rich man speaking with Abraham in the underworld. I will understand if you don't wish to believe anything based on Revelation, but the latter two seem to be harder to discount.

If I accept your assertion that "soul" came mostly from the Greeks (and it wouldn't surprise me too much, as it seems like that's where hell mostly came from), then I find it very disturbing that the vast majority of Christians are wrong about their religion. I mean, you're the first Christian I've ever met that doesn't believe in souls. It just feels like, to me, that after very many rounds of, "well, actually, the popular Christian perspective on X is actually incorrect," what you're arguing for is no longer best described as "Christianity", at least not as popularly conceived.

(And actually, this sort of disagreement between Christians is another source of improbability, for me. Call it -10dB; it's not what I would intuitively expect a true religion to look like.)

36. lavalamp says:

Actually, 10dB is quite a lot of evidence, so call it -5dB. And, given g and MtL's comments, I'll reduce my improbability due to seeming to require souls to -5dB.

37. Mom the linguist says:

The "souls" in Revelation is from one of the poetic passages I was referring to (there's another in Job). Jesus's story about the rich man and Lazarus is a parable, and I'm not committed to the literal reading of parables. Most of Jesus's stories about the afterlife describe it as a big party (with food!) of some kind, not much like the immaterial heaven naively believed in. I admit, I hadn't actually considered demon possession as an argument in favor of souls. It's certainly an argument in favor of "spiritual beings", whatever those are. It doesn't necessarily follow that our "souls" are immaterial and separable from us, unless you are imagining demon possession as like one of those TV SF devices where you put the headsets on and switch minds. I always assumed that the poor person was still "there" inside, however you think of that.

As for the belief in immaterial souls and ghosts and so on, I should probably have characterized this position as "what most people believe, unless they have been taught differently". Most Christians, like most people, are not very well informed on many subjects. This is a philosophically interesting but somewhat minor point in Christian thinking (unlike, say the Resurrection). In my 40 odd years of Bible teaching I point this out when it comes up, but it's not one of the main things one is trying to teach. I could show you many books by Christian authors taking the position I hold, and many notes in many study Bibles to much this point. You wouldn't criticize scientific positions by popular misunderstandings either. Although I know many devout people who firmly believe in the immaterial soul, this is not a point of doctrine required by any statement of belief that I have come across.

38. imho says:

Hi g.

Breaking the laws of Physics "a bit", anywhere in the universe, DOES seem obviously absurd to me. THE fundamental pillar of all science is that the natural world can be explained by laws that hold for everyone everywhere. So as not to offend, I won't talk about God. Instead, lets imagine an invisible sky fairy that does magic. You can either believe that the laws of Physics hold for everyone everywhere, or you can believe that an invisible sky fairy secretly and un -observably performs magic, which necessarily exists outside of those laws... You can not believe both as they are mutually exclusive...

Ultimately, it comes down to belief. Religion and physics are incompatible to anyone with intelligence and intellectual honesty. As I sit here typing into the computer that science built, reading an article about the mars rover that science built, with the sun (that science explains) beating down on my brow, I choose to believe in the laws of Physics... Perhaps others choose to believe in magical invisible sky fairies that listen to their thoughts... to each his own I guess?

In my view, the real issue is the mental and emotional journey one must take in moving from fairy myths to belief in physical law. It is a long and hard journey filled with self doubt, fear, and uncertainty. I myself was a Christian raised in a very religious household. I was taught than anything or anyone contradicting Christian teachings was evil and trying to hurt me. My journey began in undergrad QM deriving the Hydrogen atom. Seeing Schrodinger's masterpiece; that simple operator equation. I remember watching the Frobenious power series effortless produce spherical harmonics. The Sturm-Louisville properties that ultimately lead to selection rules. Everything coming together to beautifully and perfectly explain absorption spectra... It was the most breathtaking and powerful thing I had ever seen. Testable and with unlimited prediction power. It was perfection. Once I saw Hybrid orbitals and Hartree-Fock, which basically caricature the rest of solid matter, I was completely converted. Many body field theory was just icing on the cake.

Here is an interesting probability calculation. What is more likely, Santa Clause or God. I know for sure there were presents under my tree and I did not buy them all. I know it's possible to pick locks, and I know Alcubierre technology could produce a FTL sleigh. Now that I'm older and wiser I understand the sheer absurdity of science vs religion debates. It's like discussing the merits of travelling by city-bus or flying-unicorn-dragon (although flying-unicorn-dragons are well within natural law).

The first step is the hardest, but it's well worth the journey.

39. g says:

imho,

First of all, I have the impression (perhaps wrongly) that you have a wrong impression about my own position. I'm an atheist; I'm a mathematician and know quite a bit of physics; you don't need to tell me how beautiful and impressive science is!

So, anyway, is it absurd to suppose that science basically works but that God makes exceptions occasionally? I don't see how it could be (beyond whatever absurdity the idea of God has on other grounds). I see the aesthetic appeal of saying that the world runs purely on natural laws, and (I repeat) that is in fact my opinion and I think it very well supported by the evidence; I think one can make a good argument that this is probable given the great success of science in explaining and predicting (not to mention the other evidence pointing in the same direction); -- but I don't see what justifies going beyond that and saying it's "obviously absurd" to suppose that there's a god who makes occasional exceptions to those natural laws, and I don't see what there is in your comments on this that goes beyond saying "see, it's obvious!".

God versus Santa Claus? I think God wins this one because the naturalistic explanation for the Santa phenomena is so good: all the specific bits of evidence for Santa get good solid naturalistic explanations once one grows up and one's parents stop lying. (To the small child who isn't aware of the tradition of parents lying to their children about this, I think Santa Claus wins and actually is rationally not all that implausible given what a typical child of that age knows.)

40. lavalamp says:

imho, imagine you have a computer with enough memory to simulate our universe. You press the pause button. You change some bits in memory. You press the go button.

I personally don't think this is plausible, nor is it anything like the god Christians are looking for. If it turns out that Aron holds this view I will make an argument against it. But is it really completely inconceivable?

MtL, elsewhere on this blog it is mentioned that concepts are defined by their centers, not their boundaries. I think of religions the same way. If you and g and a few other people know that souls are exceedingly unlikely, but people like you make up only 10% of christians... OK, sure, you're still Christians. And perhaps another (different?) 10% agree with you that hell is not a literal place of eternal torment (I'm guessing that you hold this). How small of a subset of Christians do you have to be before it's no longer fair to the other Christians to call the particular set of beliefs you hold "Christianity"?

I am also curious how one ought to determine which parts of the bible are poetic, and which parts are for real. I mean, it should be obvious, yet somehow Christians have pretty large disagreements about it. (And explanations such as "by genre" don't work for me, because different people implementing the same rules *still* don't agree about what's for real and what's not.)

41. Mom the linguist says:

I'm only assuming a very commonplace definition of poetic, such that any reasonable person would use. I also don't think the poetry is to be ignored, just that poetry conveys truth in different ways than other kinds of statements. If someone says, "I'm just dead today", you don't say to yourself, "Wow, they look alive to me, therefore this statement is meaningless and to be ignored." So for instance, the souls in Revelation crying out for justice might be like Abel's blood crying out from the ground--not even the most literalist Bible interpreter I know thinks there was a little voice coming up from a pool of blood. For another example, actually coming from the Bible, when the Psalmist says "All that is in me praises the Lord" do I really think that his spleen is sitting there singing? Of course not. Nevertheless, he is making a meaningful statement.

By the way, I do believe that there is eternal torment, but I would hesitate to call Hell a "literal" place. I bet the percent of Christians who think Hell is a "literal" place is much less than 90% if you mean what I think you mean by "literal"--a place in the same sense that Los Angeles is a place.

42. lavalamp says:

If the hero of your story tortures people for eternity... Well, I suppose that may not change the probability that it exists.

It certainly changes the probability that it's evil.

43. Mom the linguist says:

Okay, I think the subject of Hell is way off this thread, but I will just briefly say that most Christians don't believe in the little demons with pitchforks whose job is to torment people in Hell either. This is also a conception from popular culture, largely based on confusion between ideas about Hell and purgatory. What I believe, and this is a very mainstream position, is that given that we have free will we can choose to refuse God, and that eternal separation from God is itself torment. This does not require me to believe in a Medieval torture chamber. Any somewhat well-informed Christian knows that there has been debate for centuries within Christianity as to whether this torment has an exit door.

44. lavalamp says:

Yes, I'm well aware of the debate and the inadequacy of the popular (mis)conception. Eternal-torment-with-an-exit-door may or may not be acceptable, g argued against it in another thread. Eternal torment without an exit door is obviously evil, no matter the type or intensity of the torment. A debate should not have been necessary, and it's not clear that the correct side has won.

45. g says:

Just to make one of lavalamp's points more explicit: the problem many skeptics have with the idea of hell has nothing whatever to do with demons and pitchforks specifically, and everything to do with the idea of horrendous eternal torment. Anything that counts as belief in hell, I suggest, had better be at least about as bad as the everlasting mediaeval torture chamber; else why call it hell, and why would it be described in terms of inextinguishable fire?

MtL, is your opinion that those who end up in hell get a genuine choice, where they actually have the option of saying no to hell, or that being an ordinary human being with the usual set of failings while not being convinced by Christianity counts as choosing hell, or what? For my part, I would rather cease to exist than face an eternity of anything for which what Christians have said about hell is at all an adequate image; do you expect me to have that option?

(I don't think hell is any more off- topic here than any other specific issue relevant to the truth of Christianity, though of course perhaps Aaron was planning to dedicate a post to "troublesome Christian doctrines" or something.)

46. g says:

Oops -- Aron, not Aaron. Sorry about all the typos.

47. lavalamp says:

I guess if I could attempt a summary of our disagreement, it'd be:

* I think A1 is not completely balanced, and I haven't had time to continue the other thread. I like g's statement about it in this thread.
* A2 is probably larger than the -40dB you give it if you consider each element individually (me and g), especially if you toss in all other arguments from philosophy that you don't mention
* I think A3 is balanced out by the fact that there's a pretty major religion that we're discussing.
* We all agree on B
* C is 0dB (me)
* D and E together can't be bigger than 40dB (me) or 10dB (g)
* g thinks F might be negative evidence; I think it's at best neutral.
* I think there's some prior improbability due to the fact that we don't observe resurrections ever; I've suggested -90dB for this.

Add that up and I think you get a number less than -100dB; as I've mentioned elsewhere I'm pessimistic about our ability to get such small numbers without making mistakes; this is small enough that it starts to make me uncomfortable.

48. imho says:

Hi g,

I've understood your beliefs the entire time. I just happened to address my meandering rant to you... Sorry. Anyway, I'm a tad puzzled by your response ... It's absurd because we are discussing magically invisible sky fairies, without a single shred of proof, whose mere existence would directly conflict with the single greatest foundational pillar of science... If that's not absurd then I don't know what is... Let me restate that pillar for everyone's benefit. The laws of Physics are immutable. That is, they exist everywhere and for everyone... no exceptions. If Zeus can bend natural law, then we as scientists can no longer trust the results of our experiments... we should go home immediately. Right now two separate experiments disagree slightly as to the mass of the Higgs... But no need to explore this further... The ambiguity is obviously proof that Ra intervened in one of the experiments... pure lunacy.

49. imho says:

Hi lavalamp,

The universe as a simulation is a mainstream physics hypothesis, well within accepted natural law. So well known in fact, that several popular physicists like Brian Green and Steven Wolfram have discussed the issue extensively. Respectable names have even calculated that this is indeed a highly probably scenario. With much of modern physics suggesting that the universe consists of pure data this is completely plausible...

It's also perfectly within natural law, to believe that Robot Spock visited the earth 10,000 years ago, built a pyramid on the exact geographic center of all landmass on the planet, taught the resident savages the rules to civilized society, and promised to return much later and judge our progress... With more planets than stars in the galaxy this is more than plausible.

No rational physicist would disagree with either of these things.

50. lavalamp says:

imho, for certain values of "god" and "simulator" those hypotheses are isomorphic, so it is extremely amusing to see those posts of yours back-to-back. :)

51. Mom the linguist says:

I think there is a difference between being in a state which is in itself tortuous and having someone actively torturing you, though they might feel equally nasty. Jesus's descriptions of Hell owe a lot to the image of the trash burning area outside of Jerusalem---being thrown out with the trash is what he's talking about. Fire that never goes out isn't exactly synonymous with what goes it burning forever. The majority opinion in Christianity seems to be that people last forever whatever they do and can't be utterly destroyed, but there is a largish minority opinion that extinction does happen, which has been held by Christian thinkers throughout history. Part of the difficulty comes from the fact that Hell isn't really talked about much in the Bible, and part of it arises from philosophical issues about time, which really don't belong in this thread...

The natural assumption is of course that you have to be "good enough" to get into Heaven (or stay out of Hell) but this is far, far from the Christian position. Nobody is good enough to get into Heaven. We can only get there by the work of God in Jesus. What exactly you have to do to qualify for this has always been somewhat a matter of debate, especially in the edge conditions. If you go by the words of Jesus in Matthew 25 (which is a traditional New Year scripture--Happy New Year, everyone), then there are going to be a lot of people who are pretty surprised to find themselves in Heaven because they didn't even know they knew Jesus.

Christians who believe in free will (which is most of them) all pretty much believe that everyone has a choice to avoid Hell. There are Christians who don't believe in free will, but I can't figure out how you argue with those people....

Most of us are convinced, however, that just going about your life in an ordinary way, kind of selfishly pursuing your own interest, but kind of trying to be nice, isn't going to cut it in the end.

52. lavalamp says:

MtL, as a consequentialist, I don't distinguish between those two scenarios.

Also, I'm well aware of the gehenna story, and of the protestant story of how one is saved.

53. g says:

MtL,

Just by way of saving you some effort, let me remark that both lavalamp and I were Christians for many years. In particular, I'm very well aware that it's not standard Christian doctrine to say e.g. that nice people go to heaven and nasty ones to hell -- that it's something much, much crueler and more arbitrary than that. (Of course you wouldn't put it that way, but that's how it looks without the Christian glasses on.) And I'm well aware that the prospects of earning one's way into heaven are supposed to be pretty hopeless.

I agree that it makes a lot of sense to see the fires of Gehenna as destructive rather than tormenting, but I don't see how you can take that view since you say you do believe in eternal torment.

Also: yes, I appreciate that eternal and everlasting may be not merely different things but different kinds of thing. But, again, I think that anything that counts as an instantiation of the Christian tradition about hell had better be at least about as bad as everlasting torture. Just add what's in store for the blessed had better be at least about as good as everlasting happiness.

I think someone who sets up a system in which billions get eternally tortured is in about the same place morally (according to my values and also to those found in the "nicer" bits of Christianity) whether they do the torturing personally, or delegate it to others, or contrive for the victims somehow to do it to themselves.

54. Mom the linguist says:

I realize I haven't said what my opinion is, which was asked for. I suspect that the question of whether eternal destruction seems to be everlasting is some kind of undecidable problem, like whether when you drop something in a black hole it falls forever or not. This is because I hold that we experience as time is part of our physical universe, just like space. So what happens when, outside of our physical universe, is always kind of asking the wrong question. Now I brought in the time problem, which I wasn't going to--oh well. This is also a standard Christian position, though most unscientifically educated Christians don't really hold it, like most other unscientifically educated people. When I teach this in church I often get the "Wait a minute, that can't be right!" reaction, until I show them the scripture passages which have led well-known theologians of many eras to hold very similar views, even before Einstein.

Really most Christians don't care about these details very much, and why should they? If you have the time and smarts to pursue it, philosophical theology is very interesting, but most of them spend more time worrying about the moral choices they are making and the useful work they are doing. Nevertheless I don't feel like I ought to have to defend the naive popular ideas of people who may be trying to be close to God but haven't had the education or leisure to develop their own nuanced theological positions.

55. Aron Wall says:

Dear imho,

I'd say welcome to my blog, but I don't exactly like your modus operendi of arguing by means of insults rather than by means of substantive argument. Despite your expressed shock, I imagine you are aware that there have existed some Christian physicists both in the past (e.g. St. Maxwell & St. Faraday) and the present (e.g. Sts. Don Page or William Daniel Phillips (a recent Nobel laureate)).

Breaking the laws of Physics "a bit", anywhere in the universe, DOES seem obviously absurd to me. THE fundamental pillar of all science is that the natural world can be explained by laws that hold for everyone everywhere. So as not to offend, I won't talk about God. Instead, lets imagine an invisible sky fairy that does magic.

You must be aware that the classic atheist-forum rhetorical trick of replacing "God" with "invisible sky fairy" is offensive. Not because it's rude to criticize other people's beliefs, but because it's rude to deliberately misunderstand them. Believing that the Ultimate Reality is more like a mind than a set of equations is not at all the same thing as believing that there's little people living in flowers. Most of your arguments against God are based, not on anything technical or precise, but on making analogies to other beliefs that everyone thinks are silly. This is a lazy form of argumentation because it allows you to avoid grappling with the details of the other person's actual beliefs. There's a reason why some grownups believe in God, and none of them believe in Santa Claus or the tooth fairy. That's why there aren't any "Asanta" internet-forums where people make fun of Santa Claus believers by comparing them to Theists.

Anyway, your "fundamental pillar" of Science, that the laws of physics are the same everywhere, isn't any such thing. The fundamental pillar of science is observation: if we discovered that the laws of physics were slightly different in the Andromeda Galaxy we'd just have to cope. We wouldn't throw our hands up in despair and say, "That tears it, I guess there's no point in thinking about anything ever again." Besides which, although a Theory of Everything which applies in all situations is a nice aspiration to pursue, we haven't gotten there yet. Our actual theories are all incomplete and only apply in certain patches of reality. So a better "pillar" is the recognition that all our physical theories are approximate models which only apply in certain situations. As the other atheists on this thread all recognize, exceptions to generally applicable laws in special circumstances is hardly a logical contradiction in terms. In fact, it's what always happens when there's some extra causal factor that one's previously formulated laws don't take into account.

But this whole discussion would go better in the comments thread of the future articles in which I will evaluate Christianity relative to Pillars of Science II-IV. If you think that you can avoid bluster and make actual substantive arguments, then I would be happy for you to participate in those conversations. Otherwise, not so much.

g and lavalamp,
I've got some replies I want to make to you, but I won't have the time for at least another few days. Until then, this is my last comment (except as moderator if needed).

56. Mom the linguist says:

Dear g:
I actually wrote my last comment before seeing the last thing you wrote, though I seem to have accidentally spoken to some of your points anyhow, at least re the everlastingness of punishment. I think it really does matter what the awfulness of Hell consists of. If it consists of the horribleness of being left out, not being in the giant party of love that would have perfectly willing to have you in it if you hadn't refused to come in the door, the horribleness of being a nasty person that hates and rages against good, the horribleness of realizing that you have done really really bad things (and surely we all know there are people who have done really bad things), I don't think it's quite fair to characterize that as God "making people torture themselves". Yeah, Hell is horrible--maybe it needs to be. If I thought Christianity required me to believe in the eternal torment of the undeserving without a choice, I wouldn't be a Christian either.

On the other hand, I don't think this is very relevant to the arguments Aron was making about why he thinks Christianity is true.

57. imho says:

Hi Aron, Sorry if I insulted you, but IMHO sometimes the brutal bluntness is the best medicine.

As requested, this will be my last post as conversations like these generally aren't that productive, and I don't want to (figuratively) poop on your (figurative) couch in your (figurative) house.

As far as substantive arguments go... I can not state it any simpler: Science has no room for magic! That's manifestly correct, and nothing more technical or precise needs to be said.

Remember, the first step is the hardest.

58. Aron Wall says:

imho,
I just deleted a comment of yours on the grounds that it was gratuitiously blasphemous. I would have sent my rejection as a private email, but I discover that the yahoo email account you submitted to my blog is not valid. Please submit a valid email address with any future comments, or you will be banned permanently.

59. lavalamp says:

As far as substantive arguments go... I can not state it any simpler: Science has no room for magic!

imho, if you're still reading, may I suggest you read the first two chapters of this: http://hpmor.com/chapter/1

Science is a way of figuring out how the universe you happen to find yourself in functions. It works whether the universe runs on math or magic.

I happen to agree that our particular universe doesn't run on magic. I rather doubt that magic can even be made coherent. But I strongly disagree with your statement here. You can do science on anything that follows a causal structure.

Also, it might not be a bad idea to note that Aron seems to be a pretty bright guy and probably (to him) the average atheist seems a little dumb. He's probably not used to having atheists keep up with him in an argument. You're not helping this impression. (Here I am perhaps generalizing from my own experience, so Aron will have to forgive me if I presume too much...)

MtL, yeah, I think the best theories on hell maintain that existence there must still be positive. I personally think the annihilationist view is still pretty bad. But at that point it becomes somewhat difficult to square with the symbology in the bible. I'm not asking you to defend the popular christian view, just to recognize that there's some substantial distance between it and yours. If a sufficient quantity of your theological views provoke surprise from an average Christian, I begin to think it might be fairer to them if you started calling yourself something else.

60. g says:

lavalamp, while I agree with your general point I don't think holding a slightly nonstandard view on hell is close to being enough to stop someone describing themselves as a Christian.

61. lavalamp says:

... I don't think holding a slightly nonstandard view on hell is close to being enough to stop someone describing themselves as a Christian.

Oh, neither do I. I was trying to say that there's some N such that if you hold N different theological views than the average practitioner of your religion, maybe you shouldn't consider yourself to be practicing that religion. I wasn't trying to claim that the correct value of N is 1. (Probably this should be weighted by importance, as certainly there are a few positions most Christians would consider essential. Probably I weight the issue of hell much higher than the average Christian would.)

62. Mom the linguist says:

Christianity is surprising! Anyone who deeply looks into it will find lots of surprising points that they didn't at first get or understand. Science is surprising too. Any scientist with deep knowledge about his subject will have particular views with surprising features to those only familiar with the popular conception of his field. The truth is surprising! This wow-I-didn't-know-that factor is why people buy books about science, theology, history, anything. Come on, you guys, this is (at least partly) a physics blog! And physics is downright weird. Lots of what physics experts tell us is deeply surprising to even scientific experts, let alone scientific amateurs. None of my views on Christianity are that idiosyncratic. They are within a range of well-known views where there are differences of opinion, analogous to whether you are a string theorist or a quantum gravitationalist (?) within the physics community. Someone with exactly my beliefs could teach in many Christian colleges or be ordained in many denominations.

63. lavalamp says:

Someone with exactly my beliefs could teach in many Christian colleges or be ordained in many denominations.

Yes, if you define christianity as the average beliefs of seminary professors, I'm sure you're not that atypical. However, I'm not so sure that it shouldn't be defined by its average practitioner, in which case you would be at least somewhat atypical, if my personal sample of christians is anything to go on.

Fields in science don't have this issue to the same extent because you can usually just do an experiment to see who's right. There's no experimental way to determine which of several parties is reading the bible with the right theology.

64. g says:

I don't think the term "Christianity" should be defined by the average beliefs of any particular group, but by (something like) the range of beliefs of people who consider themselves Christian and are generally considered Christian by others. It really is a very varied thing, and I think any definition that has the effect of excluding a lot of people-generally-considered-Christians is a bad definition.

I do, none the less, strongly agree that there are limits (though probably fuzzy ones) and that it's possible for the label "Christian" to be inappropriate for someone even though they call themselves Christian. But I've not seen any sign that anyone here is in that category or close to it.

There's a different way in which divergence from "historically typical" Christianity can be relevant, though, which gets back towards the discussion we were having earlier. On the face of it Christianity is very unlikely to be true merely because it makes lots of fairly specific claims and the more claims you make the more likely you are to be wrong about some of them. (Important but totally tangential note: of course Christianity isn't simply a body of doctrines; it's also a set of attitudes and practices, an institutional affiliation, maybe an (actual or imagined) personal relationship, etc., etc., etc. "Christianity" above should be taken as a shorthand for something more like "Christian doctrine". Just as it is in the title of this very blog entry. End of digression.) To some extent that improbability -- which on the face of it might be extremely large -- is cancelled out by the fact that this same rather specific body of propositions is in fact a major world religion. Because, e.g., one possible explanation for the prevalence of belief in those propositions is that they're true and this fact has somehow been vouchsafed to the followers of that religion. But now, suppose that you're defending Christianity, and it turns out that as you refine the particular version of Christianity you're defending in order to meet various objections, you also move some way away from the particular forms of Christianity that are most strongly represented in the tradition. Well, then, the probability of those "God ensured that his people knew the truth" explanations goes down accordingly, because it turns out that actually what his people thought they knew contains lots of mistakes, maybe important ones. You have to retreat (implicitly or explicitly) to explanations that permit lots of mistakes to creep in, but that somehow ensure that the whole thing isn't substantially wrong. Which means you get correspondingly less "cancellation" of the improbability Christianity gets from its specificity.

65. Aron Wall says:

For the most part, I'm going to try to address the issues raised on this thread elsewhere rather than adding yet more comments to an article already bloated with them. But there are a couple of comments which I find sufficiently outrageous, that I don't think I can let them pass without comment:

lavalamp:

To accept more modern stuff, at a very minimum, I'm going to require video evidence that professional stage magicians can't explain. The fact that this stuff isn't on youtube with hundreds of millions of hits is evidence to me that it doesn't exist.

First of all, as a Bayesian, you should know that it's not kosher to make arbitrary demands about what sort of data you are, or are not, willing to accept. The only thing that matters is what the likelihood ratios are, not whether it meets some sort of pre-declared test restricting the data to one very limited genre.

To be sure, if Christianity is true, and God performs some miracles (mostly healings) today as he did in the NT, then it would be surprising if this never happened when there was a camera around. So one would expect to have some videos of alleged miracles, and my understanding is that there are some videos. However, I see no reason to believe that the most convincing miracles must necessarily be the ones that are filmed. Most films only capture things from one angle, don't include the medical records before & after, and can easily be manipulated using special effects, as anyone who watches movies knows. The combination of demands you are making seems designed to be pretty much impossible to meet, even if miracles do occur. It seems better and more accurate to interview the patient, witnesses, doctors, and medical records before and after the alleged naturalistically-inexplicable healing.

lavalamp:

D and E together can't be bigger than 40dB (me) or 10dB (g)

Only a factor of 10 for (the apparent existence of) multiple historical testimonies is absurdly low, and I think does justify my statement that g has to be a skeptic about ancient history in general to be consistent. 10dB isn't even enough to take you from being 20% certain to being 80% certain of some proposition! Basically this would require one not to believe in any past events which seem even the least bit improbable. I can only think that there must be some smuggling in of prior probabilities against Christianity into the evaluation of this liklihood ratio.

Also, if those were really the odds, one would expect evidence as good as Christianity for Resurrection-like experiences, allegedly seen by dozens of people in multiple sources, 1 out of every 10 times a charasmatic religious leader, say, (of which there have been tens of thousands, at least) dies. But we don't see this by any means.

I also think that lavalamp's four orders of magnitude doesn't really do justice to all of the things that would have to go wrong, if Naturalism were true, to get written testimony like that in Paul's letters and the Gospels/Acts. But at least lavalamp is responsive to the historical evidence to some degree, rather than putting almost all of his argumentative weight on philosophical considerations.

By the way, I thought the "apologia" posted by lavalamp was a pretty good read. Obviously I don't agree with his assessment of the cumulative weight of the evidence, primarily because I'd put a whole lot more weight on historical considerations. But I was most fascinated by footnote 122, which states that:

I've heard a couple [stories from individuals] that are difficult to explain without accusing my first-hand source of lying or misinterpreting something it seems unlikely to have misinterpreted. I prefer to leave them as “unexplained” in my head because the people I heard them from seem to be being honest. Don't forget—we're talking about things that absolutely must be supernatural here. For merely statistically unlikely events, I can disbelieve a supernatural explanation is required without needing to think ill of the person reporting the experience.

It seems to me that lavalamp (or should I call him D. Smith in this context?) is in the possession of more than one first-hand (to him) miracle reports, which are impossible to explain naturalistically, such that he has good reason to believe that they are truthful, and unlikely to have made a mistake. He cannot even think of a good naturalistic explanation, so he lets them remain unexplained, even though 1) lies, 2) mistakes, 3) alternative naturalistic explanations, and 4) supernaturalism are, as far as I can see, pretty much logically exhaustive (although not exclusive) explanations for such claims.

Why not simply admit the obvious? That whether or not Christianity is true, Naturalism (despite working pretty well in many contexts) is, when taken as a Theory of Everything, an excessive prejudice for a tidy metaphysics, held with unjustified fervor by its adherents, which is abundantly contradicted by the honestly reported experiences of a fairly large fraction of the population.

Or do you think he works from outside the laws of physics, and may change the state of the universe at his whim (as I thought while a Christian)?

yes.

66. lavalamp says:

First of all, as a Bayesian, you should know that it's not kosher to make arbitrary demands about what sort of data you are, or are not, willing to accept.

I think it's good practice to state up front what sort of evidence is going to have high enough likelihood ratios to do anything to my current posterior. Generally, I have a sense that if there were a god that wished me to believe, he/she/it would have no problem supplying the evidence it would require.

...in the possession of more than one first-hand (to him) miracle reports, which are impossible to explain naturalistically, such that he has good reason to believe that they are truthful, and unlikely to have made a mistake. He cannot even think of a good naturalistic explanation, so he lets them remain unexplained...

My opinion there was understated out of charity, even at the time-- and since writing that, I've come across information that made me stop accepting one individual as a trustworthy source. Unfortunately, I now think mistakes, exaggerations, and lies are completely sufficient explanations.

Only a factor of 10 for (the apparent existence of) multiple historical testimonies is absurdly low...

My understanding is that g doesn't accept that there are multiple historical testimonies. I tend to agree with him that your 140dB is too high. 10dB seems a bit low to me, but if you believe that basically everything went through the bottleneck of Paul and maybe one or two other people on its way to us, lies become a pretty reasonable explanation; I'm not certain enough about that sort of explanation to give only 10dB.

I'll try to let this thread go now, but since you believe in a simulator god-- well, I have an argument against this that I will try to save for a more appropriate thread.

67. dog ears says:

It (nearly) all comes down to the prior probabilities, I suppose. (Happily not quite all, else I'd have to find honest work.) I am, therefore, curious about a different facet of people's priors: Suppose that something broadly like the Christian story were true: a messiah is born, performs miracles, teaches, makes disciples, dies. How much evidence would you expect there to be? In worlds with a messiah, would you expect more evidence than ours has for Christianity? Less? About this much? More to the point, how does this world fall, percentile-wise?

The answer to this question doesn't establish anything about whether or not Christianity is true, of course, but I find that such meanderings do have epistemological value. The thing is, if you'd expect roughly this sort of evidence in worlds with messiahs, then either you are a Christian or it would be rather difficult for you to be a Christian. If the latter, you might more easily be able to believe in a different God (one who would create a world without suffering, perhaps, or one who would declare His existence through rearranging the stars), but a God who so loved the world that He gave His only begotten Son––a much harder sell.

If you're a non-Christian, then either you think that there's less evidence for Christianity than you'd expect were there a messiah or you're comfortable being wrong in the majority of worlds in which there is a messiah. Both possibilities seem interesting, perhaps even pertinent.

68. g says:

Let me say a few words about that 10dB figure.

1. Let's divide the NT evidence concerning the resurrection into two bits. (a) The mere fact that the NT says it happened. (b) All the other stuff. I agree that (a) gives quite a lot of evidence, merely by its specificity. But that's already been counted when you say that (E) includes "factors which cancel out (A1) and (A3)". (You may recall that I agree that that happens to some extent, but don't think the cancellation is complete.)

So, if you include (a) here and assume that "Christianity says that Jesus died and was raised from the dead" is not part of the background information we have when evaluating the NT as evidence, then I will happily agree to a larger figure than 10dB. But for every dB you gain there, you lose the same amount when we come to reckon how improbable Christianity is merely on account of being a conjunction of lots of claims, because you can no longer offset the specific claim "Jesus was raised from the dead" against the specific evidence "the NT documents say that Jesus was raised from the dead".

2. For the exact same reason, my assessment of (b) at only 10dB doesn't mean I should expect 10% of all charismatic religious leaders should have comparable documentary evidence of having been resurrected. It means I should expect 10% of all that claim their founders were resurrected to have comparable documentary evidence for that claim. Which seems pretty reasonable to me.

3. And, for the same reason again, I don't see that I am in any way obliged to be a total skeptic about ancient history in general. What I am skeptical about is the claim that the NT documents offer more than about 10:1 evidence for the resurrection beyond what they give merely by saying it happened.

Butting in briefly to Aron's and lavalamp's discussion of naturalism, miracles, etc.: It is true that naturalism is abundantly contradicted, at least prima facie, by the reported experiences of a fairly large fraction of the population. (Honestly? Probably, but it's really hard to be sure; not because I think many people are flat-out lying, but because all sorts of self-deception and exaggeration are terribly easy.) But it's also true that in those cases where it's possible to investigate those reported experiences properly and someone does so, it generally turns out that the experiences aren't after all good evidence of supernatural happenings. People are very easy to fool, very good at fooling themselves, and very bad at evaluating evidence.

69. g says:

dog ears, I really don't think it does all come down to people's priors. (Though clearly they are of some importance.) But, to try to answer your question: I think there's quite a wide range of possible worlds with Christs, and what I'd expect about them depends on just how "like the Christian story" I'm supposed to be assuming. Because some bits of the Christian story are things I decidedly wouldn't expect an actual messiah, or an actual God Incarnate, to do, and some of those have an impact on what sort of evidence I'd expect there to be.

For instance, Christians like to say things like this: "The Jews expected a messiah who would come and liberate them from the Romans, and set up an eternal reign of justice and peace. In Jesus, God confounded their naive expectations and did something much more surprising and subtle." And it seems to me that in so far as the idea of a messiah makes sense at all, and in so far as they really were expecting that, they were right to have that sort of expectation, and the fact that that isn't what Jesus did is evidence against his actually having been The Messiah.

So. I think the "naive" expectation that a messiah -- still more an incarnate God -- would make more of a visible mark on the world, is a perfectly reasonable expectation. And in those possible worlds, I would expect to see a lot more evidence. But it's always possible that the Christians are right when they say that's too simplistic an expectation, and that it's more fitting for God to show his glory through obscurity, his riches through poverty, his strength through weakness (though that still looks to me like rationalization), and in those (perhaps rather few) possible worlds where God has some sort of commitment to never doing anything that gets him clearly noticed, it wouldn't be that surprising for a messiah or incarnation to leave little evidence. But note that in those worlds, a whole lot of the Old Testament had better be fiction, because the OT's god clearly has no compunction at all about showing his power dramatically, at least some of the time.

I don't see any obvious incompatibility between a god who is more visibly active in the world, and a god who so loved the world that he gave his only begotten son.

70. lavalamp says:

So what you're saying is that when we accept christianity's existence as background information, we've already incorporated most of the evidence that the bible would otherwise have given us? So would it be fair to say you don't think Aron's 140dB is necessarily wrong, just double-counting?

71. g says:

140dB seems too much to me even if it weren't double counting. But yes, I think most of the evidence the Bible gives for the resurrection is what comes from the mere fact of saying it happened.

72. lavalamp says:

I put together a spreadsheet to collect my thoughts. I'm not entirely happy with it. But it looks like, to have a shot at convincing me, you'll need to convince me that I shouldn't count the background rate of resurrections as evidence against Christianity and that the fine-tuning argument is an argument for Christianity.

73. Aron Wall says:

lavalamp:

But it looks like, to have a shot at convincing me, you'll need to convince me that I shouldn't count the background rate of resurrections as evidence against Christianity and that the fine-tuning argument is an argument for Christianity.

Well, it looks like I have a chance then! These are really epistomology/philosophy issues, so they don't really require going much into the nitty-gritty of historical details. Let me briefly try to make a case for the 2 things you mention, plus one other maneuver which I think I can get some significant traction out of.

1. Let us define Christianity' (note the prime) as being the proposition "At least 98% of the bits in Christianity are true" (where we arrange the theory-space so that religions related by a small number of bit flips make very similar claims). Now suppose I only wish to convince you of Christianity'. (An adherent of Christianity' could probably reasonably be characterized as a Christian, although their theology might be viewed as "liberal" or "heretical" by more conservative types, depending on which bits ended up being different.) Using the central limit theorem on your .01 error rate, plus the rule that the error bar on a Poisson process goes like the square root of the expected count, I find that the result is a Gaussian distribution centered around 99% with 98% located 5 sigma away. Now I'm not going to claim this as a $10^6$ liklihood ratio for Christianity', since the error rate of the bits aren't independent of each other (if the whole thing is made up, then the whole thing is made up). What I do claim is that this eliminates your factor of 72 dB.

2. It isn't fair to count the background rate of Resurrections against Christianity. Consider by way of analogy this proposition: "The fastest human being can run a mile in under 4 minutes." Let us suppose it has been established--either by inductive evidence or from basic facts about human physiology--that fewer than $10^{-6}$ individuals can do this. If your next door neighbor Joe boasts that he can run the mile in under 4 minutes, the unlikelihood of his claim is relevant when deciding whether or not to believe him. Then you hear on the news that some Olympic athelete just ran the mile in under 4 minutes. In considering whether or not to believe this report, the factor of $10^{6}$ isn't relevant since the Olympic athlete is not a randomly selected human being; they are already known to be one of the fastest people in the world.

In the same way, Jesus is one of the "Messiahest" people who have ever lived. That is, on the basis of the data in (D), he is one of a small handful of people who are most likely to be the Messiah, if anyone is. This by no means proves him to be the Messiah, any more than the fact of someone being an Olympic athlete would prove that he can run the mile in under 4 minutes. But it does make the background rate irrelevant.

3. There's a lot of questions surrounding the fine-tuning argument: (a) what is the appropriate set of prior probabilities to place on the parameter space of constants? (b) how much fine-tuning is there really (i.e. what are the odds of this occuring by chance alone)? (c) is it possible for there to exist a physical mechanism to adjust the parameters to the right values? (d) is it possible that life of a different sort than ours could exist anyway? (e) could it be explained by a multiverse? and (f) should this sort of thing count as evidence at all, since if we didn't exist we wouldn't observe anything?

I don't have time right now to go into depth on any of these particular issues, I hope to make a series of posts on them, several months from now. There's several different kinds of fine-tuning claims, and some are a lot more serious than others; the philosophical literature doesn't do a good job separating the wheat from the chaff. In my opinion the cosmological constant is the most serious, but there are some others such as the Higgs mass, the triple-alpha process, and some other stuff needed to get stars to work right.

If you'll accept my professional expertise as a working physicist on issues (a-d), the best answers in light of our current knowledge are (a) assuming something like reductionistic naturalism, where the laws of physics at short distance scales can't depend on whether they produce life or other interesting effects at longer distance scales, the prior probabilities can be estimated using something called the "renormalization group" which is a standard tool in QFT used for other purposes; (b) cosmological constant: $10^{-120}$, Higgs mass: $10^{-30}$, triple-alpha: $10^{-2}$; (c) cosmological constant: extremely unlikely in our current state of knowledge, our experience in QFT is that cancellations like this only happen when they're forced to by some special symmetry, and we can mathematically classify all possible symmetries of QFT's: the only one which could help is supersymmetry, but supersymmetry if it exists is broken in Nature and therefore can only mitigate the fine-tuning to $10^{-60}$, not eliminate it entirely. One could always postulate some unknown quantum gravity mechanism, but this would have to operate at energy scales which are both much smaller than the Planck scale, and which can also be probed by existing experiments. Higgs mass: here supersymmetry or possibly certain other ideas could fix it, although the current LHC data isn't looking good right now. Triple alpha: hard to see how this could work since the values of the constants needed to produce carbon aren't special in any other way, besides that they produce carbon. (d) For the cosmological constant, if it were anywhere close to its expected value, no complex information processing of any sort would be possible, Higgs mass: not sure I'll have to think about it, Triple-alpha: Conceivable that life could exist without carbon or any heavier elements, but seems unlikely.

With respect to the more philosophical questions: (e) $10^{120}$ universes each with their own laws of physics is logically possible, but surely one pays an Occam's razor probability cost for postulating this sort of thing without any other evidence! Unless you would be certain that there was a multiverse even if there was no fine-tuning, which seems totally crazy to me.

(f) I think the "firing squad" cases make the answer to this one obvious. If I'm about to be executed, 100 marksmen shoot at my head, and they all miss or fire blanks, this still "calls for an explanation" even though I would not be there to ask the question if I had died.

74. Aron Wall says:

lavalamp,

First of all, as a Bayesian, you should know that it's not kosher to make arbitrary demands about what sort of data you are, or are not, willing to accept.

I think it's good practice to state up front what sort of evidence is going to have high enough likelihood ratios to do anything to my current posterior. Generally, I have a sense that if there were a god that wished me to believe, he/she/it would have no problem supplying the evidence it would require.

If you want to say that a loving God, if he existed, would probably have made his existence totally obvious to everyone, then you don't need to go so far as the absence of video evidence. It's already obvious that God hasn't made his existence blindingly obvious; if you think he should have, then that's more of an Argument from Evil kind of a discussion than a Historical Evidences kind of discussion.

However, if God wants you to have to use your brain to discover that he exists, that doesn't necessarily place restrictions on what kind of evidence God ought to use. What I'm trying to address here is the question "Assuming that God exists and in some contexts performs miracles at roughly the same rate as in e.g the early church as claimed by Christianity, what are the odds that we would get video evidence of a kind that would satisfy James Randi." And I think the odds of this are actually not all that high. Why would there be millions of hits on youtube? Why wouldn't all the atheist websites dismiss it for exactly the same way they dismiss every other kind of miracle claim, without doing any sort of careful investigation to see if the circumstances are special? Anyway, I was trying to argue that there are other kinds of evidence that are better than video evidence is likely to be.

My opinion there was understated out of charity, even at the time-- and since writing that, I've come across information that made me stop accepting one individual as a trustworthy source. Unfortunately, I now think mistakes, exaggerations, and lies are completely sufficient explanations.

Well, I don't know these people so I'll have to defer to you about this. When it comes time to discuss modern miracles, I'll try to supply some better instances, although obviously these won't be people you already know.

BTW, I wouldn't use the term "simulator" to refer to the classical theism picture where God is the fundamental entity, and he decides what else exists. To me simulator suggests the existence of a physical computer or somesuch somewhere.

g:

But it's also true that in those cases where it's possible to investigate those reported experiences properly and someone does so, it generally turns out that the experiences aren't after all good evidence of supernatural happenings.

I think there's a decent number of counterexamples to this, where actually it looks just like it really happened, and then the skeptics have just said some lame things and moved on. But we'll have to save this for a later discussion. In the meantime you could read the Keener book if you want to look into some specifically documented cases.

Let's divide the NT evidence concerning the resurrection into two bits. (a) The mere fact that the NT says it happened. (b) All the other stuff. I agree that (a) gives quite a lot of evidence, merely by its specificity. But that's already been counted when you say that (E) includes "factors which cancel out (A1) and (A3)". (You may recall that I agree that that happens to some extent, but don't think the cancellation is complete.)

What I actually said was that (D) cancels out (A3), while (E) cancels out (A1) with some left over. I don't think it's been shown that the cancellation isn't complete. The only argument I recall for this is lavalamp's statement that one expects some fraction of the bits of data to be corrupted, but one can respond to this by incorporating a margin of error into the definition of "Christianity". The reason why there is some left over afterwards is to account for the implausibility of whatever conspiracy or mistake is needed to get the Resurrection claims which we see.

As lavalamp says:

My understanding is that g doesn't accept that there are multiple historical testimonies. I tend to agree with him that your 140dB is too high. 10dB seems a bit low to me, but if you believe that basically everything went through the bottleneck of Paul and maybe one or two other people on its way to us, lies become a pretty reasonable explanation; I'm not certain enough about that sort of explanation to give only 10dB.

Actually, I think lies are the most reasonable explanation, but I think that the lies have to have involved a significant number of early people besides Paul. If you read Paul's letters it's clear he just doesn't have the status in the early Christian community to invent its most important doctrine; there's lots of other church leaders (e.g. Peter, John, and James brother of Jesus) who are more respected than him, and he's constantly dealing with Christians who are dismissive of his apostolic claims. So I think it really qualifies as a conspiracy.

Imagine coming into an organization founded by 10 disciples of some great man. You come into this community as a former enemy of the great man. You have to persuade all 10 of these people, not just that you have seen him come back from the dead, but that they have. Failing that, to persuade (say) 90 out of 100 next-generation leaders selected by those 10 people, that those 10 people in fact made this claim. Your chances of success are clearly low. And this is taking for granted that you have some reason to want to do this, besides it being true.

g:

2. For the exact same reason, my assessment of (b) at only 10dB doesn't mean I should expect 10% of all charismatic religious leaders should have comparable documentary evidence of having been resurrected. It means I should expect 10% of all that claim their founders were resurrected to have comparable documentary evidence for that claim. Which seems pretty reasonable to me.

You expect that 10% of such religions would have a major enemy converted by having a vision of the supposedly resurrected religious leader? Really?

And to claim that the leader appeared to dozens or hundreds of different people? And not (so far as we can tell) to have evidence of people recanting afterwards, as in the case of Joseph Smith's plates? And to be (as far as we can tell) willing to die for their claim? And to have multiple documentary sources which have to be individually impeached? etc etc. I wouldn't expect this at all.

Judging from my experience of other religious claims, I would expect to see that the large majority of Resurrection claims wouldn't be supported by any better historical evidence, than say the doctrine of the "Assumption of Mary" is. (Sorry to pick on Catholics, but this was the first example to come to mind). I would not expect to find the pattern of evidence which I see in the NT.

75. lavalamp says:

#1: I need to read some (mathy) stuff to see if I agree with your suggestion or not. I'm not completely happy with the way I did the complexity penalty, but I have this strong intuition that you need evidence to make up for complexity; otherwise, you could make stuff true just by saying it. I'm not sure that accepting 98% of the bits as correct would be good for you; I feel like that would just concentrate the errors in the least likely to be correct bits, like those coding for the resurrection, etc. (Also, my 2400 bits is probably an extreme underestimate of the length of the Christian hypothesis.)

#2: I already accounted for that, it's what the "world religion bonus" is for. :) As you can see, it almost completely cancels the -90dB resurrection penalty. (I put it down as 70 and not 90 because I'm arbitrarily saying that atheism has something like a +20dB world belief system bonus.)

#3: I'll happily agree to anything you say on (a)-(c) and probably (d). For (e), I go by Kolmogorov complexity-- I have to pay for extra rules; if the rules predict tons of universes, I don't have to pay for each one (just like modern physics doesn't have to pay for each star individually). I tend to think the MWI makes sense, and it seems to give us as many universes as you could want. If you don't accept MWI, I can find a paper by Tegmark in which he argues that you still get all the multiverses just from the fact that space seems to be infinite (flat or open). (Fair warning: my true belief here is that something like the MUH must be true, so there's three layers at which I think there could be/probably are multiple universes.)

Finally, (f): Bayseian evidence involves (as you know) comparing what the world would look like if the hypothesis is true with what it would look like if it were false. Observers for whom naturalism is true must observe a universe in which they can exist. Observers for whom theism is true must also observe a universe in which they can exist. The fact of one's existence therefore isn't bayesian evidence one way or the other. The question becomes, "What's the relative frequency of naturalism-dwellers vs. theism-dwellers?" And there it becomes difficult; I'm not sure one can admit a way of mechanically enumerating all possible worlds and still believe in something that counts as a god.

In other words, firing squad survivors will invariably find a bizarre and bewildering array of malfunctions in the equipment of the soldiers shooting at them, or some weird conspiracy, etc. That's par for the course. Anthropics is weird.

One last problem with fine tuning, it's not an argument for the christian god specifically. It's an argument for theism (or simulatorism). I think if it were an argument at all, it would mostly bump up the probability of deism and only add 1 or 2 dB to Christianity.

Finally, if you can somehow overcome all that, I think the philosophical evidence makes (chirstianity+god is evil) much more probable than (christianity).

76. lavalamp says:

I have to run, but I just want to point out that you can't actually believe this and simultaneously come up with the very high probability that you have for Christianity's truth. Your numbers, if true, mean that god's existence is blindingly obvious.

77. g says:

(I am not lavalamp.) On #1, I don't see that this eliminates lavalamp's 72dB at all. For that matter, if he's anything like me, the proposition he's considering is probably more like Christianity' than like Christianity-sensu-strictu to begin with. (At least for the more "generous" of the three sets of figures in his spreadsheet. But maybe I haven't quite understood your argument here, which is quite terse; if that seems like a good explanation for my not seeing why you think you've eliminated that 72dB term, perhaps you might like to expand on it.

On #2, I think this is again a matter of correct accounting for information that (so to speak) locates Christianity in hypothesis-space, and balancing it against various kinds of prior uncertainty. I haven't thought through exactly where Aron and lavalamp are allocating the various bits of evidence and improbability here, but my initial reaction is that Jesus's being rather "messiah-y" does probably mean that 10^-9 isn't the right prior probability to use for his resurrection, but that it by no means eliminates that improbability altogether. So far as I know, before Jesus there was no expectation that the messiah would die and promptly be resurrected; nor that, conditional on his dying at all, he would promptly be resurrected. And among somewhat-messiah-y figures in history (of whom there's quite a supply), the number of credible resurrections is ... small. And there are pretty good reasons for thinking that the rate of resurrections is actually zero (or at least way, way, way lower than 10^-9) -- but this may not be the right place for them in the accounting.

On 3(e), I don't think Occam's razor is much concerned with the size of the universe (by which I mean here "absolutely everything there is", which might turn out to be what's sometimes called a multiverse, or something more exotic still) but only with the complexity of the laws that govern it. (Exactly what one should penalize in Occam's razor and why is a difficult philosophical question and doesn't have a universally accepted answer. The sentence before this parenthesis describes my opinion, which I know is not only mine, but I don't claim that everyone else is obliged to agree.) Multiverses of various sorts have been proposed for plenty of reasons other than fine-tuning, which suggests but obviously doesn't prove that you don't need anything like a 10^120 Occam penalty for the idea.

On 3(f), I think a slightly deeper analysis is called for, so let me briefly sketch some bits of one. Suppose 100 expert marksmen shoot at you from close up and all miss. What do we mean when we say that this calls for an explanation? That we think "they just all happened to miss" is not likely to be the best explanation. Why? Because there are lots of candidate explanations with much higher prior probability. The marksmen will all miss just by chance, very very occasionally. But elaborate practical jokes, political conspiracies, and hallucinations all happen much more often. (That is not meant to be an exhaustive list of explanations, of course.) And because our experience says that when very weird things happen, there usually turns out to be some non-chance explanation. Which isn't surprising, because again there are lots of less-random ways for very weird things to happen, so (almost?) all instances where sufficiently-weird things happen have actually been the result of less-random factors.

Now, assuming arguendo that the best contemporary naturalistic theories make the universe as fine-tuned as Aron says, how (if at all) does the above apply? It does seem that we have some candidate explanations with higher probability, and maybe creation by the Christian god is one. (Though it's not entirely obvious; lavalamp and I both estimate several tens of dB of evidence against the Christian god, after all.) But here is another kind of candidate explanation: "Our current understanding of the laws of physics is wrong in some way that explains most of the apparent fine tuning". This is of course rather lacking in detail -- but so is "The Christian god did it". And, really, would any reasonable person bet against it at more than (say) 10^6:1? Is our understanding of physics so good that there isn't even a 1-in-1,000,000 chance that when we find the right physical theory it will turn out to explain why the cosmological constant is what it is, or do away with the need for it to be so exactly tuned at the start of time, or something? I say: duh, no, of course it isn't. And since this is more a question of the philosophy of science than one of physics as such, I think Aron ought to say the same if he is to be consistent with his general view, stated earlier, that philosophical arguments never deliver more than a few orders of magnitude of evidence.

78. g says:

Note: when I wrote the above I hadn't read the three comments immediately preceding it. In fact I still haven't, but I have at least now seen that they're there. I'm going to look at them shortly.

79. g says:

So, first of all, my apologies for writing "(E)" where I meant "(D) and (E)". That was a screwup.

No, of course I don't "expect that 10% of such religions would have a major enemy converted by having a vision of the supposedly resurrected religious leader". That's far too specific, and again if you try to make that argument actually work you'll find you're double-counting the specificity of the Christian testimony. I expect that 10% of such religions will have comparably impressive documentary evidence. Clearly we disagree about how impressive that documentary evidence is. In particular, I don't think the fact that Paul was converted and said he had visions of Jesus is much evidence that Jesus was actually resurrected, and I don't really understand why you think it is.

Speaking of Paul, I'm not sure what lavalamp's position is but I certainly don't think it's likely that the idea of Jesus's resurrection originated with Paul making it up. (I don't have a specific favourite theory, but here is one possibility: Jesus's body was put in the wrong tomb by mistake; later some people went back to the right tomb and of course found no body there; after that, the very common experience of hallucinating encounters with recently dead loved ones led to the idea that Jesus had actually risen from the dead; most of the further details found in the NT are embellishments of the sort that naturally happen to any intriguing story over a few decades of oral transmission. Given that the NT specifically says Jesus was buried in someone else's tomb, the first and most improbable step here seems to me to have probability at least 10^-3. Given that, the rest seems very, very plausible. And this story seems to account for just about everything the NT says about the resurrection unless one assumes for no particular reason that the people involved were exceptionally skeptical and reliable. -- All of this, I repeat, is just one possibility, and I don't mean to imply that it's the only viable alternative to the Christian story.)

So, getting back to what I'd expect in (say) 10% of religions that claim their founder died and was raised from the dead. I would not expect to see the exact same claims as the early Christians make about Jesus, nor the exact same pattern of documentary evidence. I would expect to see similar sorts of claims and similar sorts of evidence. I do not find the documentary evidence we have in the Christian case nearly as impressive as you do. If (as I expect) there is going to be a later post (or more than one) specifically on that evidence and why you find it so impressive, perhaps we should defer detailed discussion of this point until then? A couple of very brief comments before I drop it, though: (1) Is there, in fact, much evidence that any of the "eight witnesses" or the "three witnesses" later recanted? Wikipedia (which is always right, except when it's wrong) seems to indicate that they didn't even though a number of them broke with the LDS, at least for a time. At the most, Martin Harris seems for a while to have said he saw the golden plates only "with a spiritual eye" -- but later he strongly denied having said any such thing. (2) The evidence that any of the apostles faced death and could have avoided it if they'd recanted is very, very thin. (3) Supposing it to be true, the most it could show is that they sincerely believed their claims, not that they were actually true. That would (almost) rule out deliberate deception by the apostles in question, but not the sort of scenario I described above, nor one involving deception by other people (e.g., some other disciples besides the ones who were willing to die for their beliefs stealing the body of Jesus -- which I don't think especially likely but wouldn't bet against at more than, say, 10^3:1).

Incidentally, Aron, your blockquoting in that comment is messed up; if you are able to fix it retroactively, you might want to do so.
[done --AW]

80. g says:

Oh, one other remark about fine tuning. (Though again we should perhaps leave this until later, if Aron plans to devote future posts to that topic.)

In order for "God did it" to be a good explanation for a finely tuned universe, it's not enough to suppose that God would have wanted there to be intelligent life -- because what you need to explain is not merely "a universe suitable for intelligent life" but "a universe suitable for intelligent life but only barely", that being the whole point about fine tuning. On the face of it, given a god who wants to create intelligent life and who has a preference for a universe that largely runs according to rules, you would expect that god to choose a region of possible-universe space that gives plenty of scope for "tweaking" to get whatever sort of outcomes he wants with as little intervention as possible; and you would have no reason to expect that he would pick a point in possible-universe space that is surrounded only by a teeny-tiny region of other viable universes. And, on the face of it, I'd expect such regions to account for only a small fraction of the portion of possible-universe space in which intelligent life (or religiously-inclined life, or just life, or just complex stuff, or whatever God might be supposed to want) is possible.

Therefore, for God to be a good explanation for fine tuning, one of two things needs to be the case. Either you need to have a reason why God would have preferred a finely tuned universe; or you need to have a reason why actually the universe had to be finely tuned (i.e., why there simply isn't a bigger region of parameter space, even with different physical laws, permitting the emergence of intelligent life, or whatever else his goal may have been). And these reasons, whatever they are, have to provide something like the 10^120 (or whatever) probability factor that you're trying to explain, because otherwise your theistic hypothesis hasn't succeeded in making our universe much more probable than it is on an atheistic hypothesis.

I don't know of any such reasons, nor do I know of any reason to think there are any such. So I think the theistic explanation is left with some vague handwave like "maybe God has some good reason for preferring such a universe", which indeed might be right but seems absolutely no better than "maybe the true laws of physics explain (or explain away) the fine tuning after all".

(All the above is assuming for the sake of argument that all the usual claims about how finely tuned the universe is, and how unable present physical theories are to account for the fine tuning, are correct. I have not so far been convinced that those claims are right, though that may be partly or wholly because I haven't put a lot of effort into investigating them.)

81. Aron Wall says:

#1: I need to read some (mathy) stuff to see if I agree with your suggestion or not. I'm not completely happy with the way I did the complexity penalty, but I have this strong intuition that you need evidence to make up for complexity; otherwise, you could make stuff true just by saying it.

I don't think your "strong intuition" can possibly correct. I've explained why enough times (e.g. the Farmer Jones case) that I'm not sure what else I can say, so let me just point out that we all have some fairly specific beliefs. If some fraction such as 1% of the bits of complexity carried over into the implausibility of the conjoined hypothesis, in a way that was there even after "priming" to accomodate an error rate in the source, I don't think you'd be able to believe anything complicated. Try it with something else besides Christianity, and see what happens.

I'm not sure that accepting 98% of the bits as correct would be good for you; I feel like that would just concentrate the errors in the least likely to be correct bits, like those coding for the resurrection, etc.

Well, that would have more to do with other sorts of a priori unliklihoods, not just specificity, doesn't it? Anyway, it would be rather strange if a core doctrine like the Resurrection was encoded in only a few bits. Besides the Resurrection is also the part which has a greater share of the evidence for it.

(Also, my 2400 bits is probably an extreme underestimate of the length of the Christian hypothesis.)

Doesn't matter. No matter how many bits there are, you can do the same thing.

#2: I already accounted for that, it's what the "world religion bonus" is for. :) As you can see, it almost completely cancels the -90dB resurrection penalty. (I put it down as 70 and not 90 because I'm arbitrarily saying that atheism has something like a +20dB world belief system bonus.)

I'm confused by your motivation for sorting the probabilities this way; the meaning of the phrase "world religion bonus" isn't very clear. Why not just say you're assigning a 20 dB penalty to Christianity, on the grounds that it requires additional entities that do things such as Resurrections, which seems weird to you? Perfectly respectable, that.

For (e), I go by Kolmogorov complexity-- I have to pay for extra rules; if the rules predict tons of universes, I don't have to pay for each one (just like modern physics doesn't have to pay for each star individually), and it seems to give us as many universes as you could want.

It's one thing not to have to pay for extra stars when we know there exist other stars, and we know there exist mechanisms to produce other stars. It's another thing if we had to invent gazillions of stars with no evidence in order to "explain" something about our own solar system.

I understand the intuitive appeal of Kolmogorov complexity, but I don't think it actually works well when you try to apply it in practice. The simplest possible hypothesis which includes us is indeed that every logically possible universe exists (which has complexity approximately equal to 0). At that point we're stuck, because we've now explained everything (e.g. the sky is blue because all possible sky colors exist, etc.) without explaining anything. Keep going and you'll run into "the measure problem"--I think you'll probably have to conclude we're a Boltzmann brain--and all sorts of horrible paradoxes.

I think the right answer is to go back to square one and admit that Occam was right: "Entities should not be unnecessarily multiplied". In other words, you pay some price for postulating extra entities, as well as for algorithmically complicated hypotheses. I forsee only trouble if I abandon this elementary rule of common sense.

I tend to think the MWI makes sense

It doesn't make any sense to me at all. I cannot force my mind into the shape required for MWI to be true. "Every possibility happens" is logically equivalent to "Every possibility happens with probability 1" which is inconsistent with making probabilistic predictions about the world. If many worlds is true, then by definition Jesus did rise from the dead, in some universe. And that universe is just as real as our own. I don't think that different universes can exist more or less than each other, so there's no hope of recovering the Born rule. Insanity!

If you don't accept MWI, I can find a paper by Tegmark in which he argues that you still get all the multiverses just from the fact that space seems to be infinite (flat or open).

Not good enough, you also need the laws of physics to be able to vary. Also, inflation means that the universe would probably look infinite even if it was finite.

In other words, firing squad survivors will invariably find a bizarre and bewildering array of malfunctions in the equipment of the soldiers shooting at them, or some weird conspiracy, etc. That's par for the course. Anthropics is weird.

As g correctly points out, the firing squad survivors will almost invariably discover that they were the victim of a conspiracy or practical joke (fake executions have occured), not "a bizarre and bewildering array of malfunctions in the equipment of the soldiers shooting at them" (something requiring 100 unlikely events occuring simultanously). God (and other candidate explanations such as a physics mechanism) are like the former, "we just got lucky" is like the latter. The former is to be preferred to the later.

One last problem with fine tuning, it's not an argument for the christian god specifically. It's an argument for theism (or simulatorism). I think if it were an argument at all, it would mostly bump up the probability of deism and only add 1 or 2 dB to Christianity.

Huh? That's not how Bayesian updating works. If the universe is fine-tuned, all possible explanations for that fine tuning (natural or supernatural) are raised by the same factor (assuming the liklihood ratios are the same, i.e. all explanations are equally good aside from prior probabilities.) Christianity is a kind of Theism, so arguments for Theism are equally well arguments for Christianity, relative to Naturalism.

to include a comment from g:

But here is another kind of candidate explanation: "Our current understanding of the laws of physics is wrong in some way that explains most of the apparent fine tuning". This is of course rather lacking in detail -- but so is "The Christian god did it". And, really, would any reasonable person bet against it at more than (say) 10^6:1? Is our understanding of physics so good that there isn't even a 1-in-1,000,000 chance that when we find the right physical theory it will turn out to explain why the cosmological constant is what it is, or do away with the need for it to be so exactly tuned at the start of time, or something?

The cosmological constant is (so far as we know) a constant, so that it takes the same value now as it did at the start of time. (It involves the dynamics, rather than to the initial conditions of the universe. Otherwise I would think it was a lot weaker since we don't know what determined the initial conditions of the universe.)

Other than that, I agree completely. That's why I only claimed 2-4 orders of magnitude from the fine-tuning argument.

Finally, if you can somehow overcome all that, I think the philosophical evidence makes (chirstianity+god is evil) much more probable than (christianity).

I'm not sure I want to convince you of that, but it would be interesting to find out what would happen if I did! But surely if the Gospels accurately portray God as having the character of Jesus, there is significant positive evidence to be had there that God is (at least in some sense) benevolent, i.e. he cares about ethics and has compassion and wants us to be saved? Are you that sure that suffering isn't important, or that free will can't necessitate something similar to Hell, so sure that even if you came to accept the other Christian claims about Jesus you couldn't come to believe this too? Plus there's always the Universalist interpretation of the Bible, which isn't by any means the most plausible reading (some probability hit there) but isn't impossible.

82. Aron Wall says:

g,

I suspect that if you slap together a random set of laws of physics some random way (e.g. using a cellular automaton), the odds are you won't get life, or complicated structures on many different orders of magnitude. In that sense I find the generic Argument from Design plausible, I suspect that God had to do some fine-tuning if he wanted simple mathematical laws of physics + life. But the degree of fine tuning necessary may depend on the kind of laws.

Suppose it turns out there's one kind of laws of physics A in which life is nearly inevitable no matter what the constants, and our kind B where it requires fine-tuning. On Theism, we just need a reason for God to prefer universes of type B to type A (it doesn't have to be because he wants fine-tuning, it could be for some other reason. This is unmotivated, but it's hardly impossible, there's only 2 choices in this hypothetical after all. If you assume that God has a 1/3 chance of preferring A, a 1/3 chance of preferring B, and a 1/3 chance of not caring and making the selection randomly, it only lowers the probability of (Fine-Tuned Universe | Theism) by a factor of 3.

To the extent that it's a problem, some of the other candidate explanations seem to be in the same boat, e.g. if there's a multiverse with universes of type A and B, with overwhelming probability you'd expect to live in A rather than B, contrary to observation, unless you restrict the multiverse to universes of type B.

83. Aron Wall says:

No, of course I don't "expect that 10% of such religions would have a major enemy converted by having a vision of the supposedly resurrected religious leader". That's far too specific, and again if you try to make that argument actually work you'll find you're double-counting the specificity of the Christian testimony. I expect that 10% of such religions will have comparably impressive documentary evidence.

Well, I don't expect this. It's not just the specificity of Paul's conversion, it's the improbability.

In particular, I don't think the fact that Paul was converted and said he had visions of Jesus is much evidence that Jesus was actually resurrected, and I don't really understand why you think it is.

If it's more likely to happen if Christianity is true than if Christianity is false, then that's a likelihood ratio.

Jesus's body was put in the wrong tomb by mistake; later some people went back to the right tomb and of course found no body there; after that, the very common experience of hallucinating encounters with recently dead loved ones led to the idea that Jesus had actually risen from the dead; most of the further details found in the NT are embellishments of the sort that naturally happen to any intriguing story over a few decades of oral transmission. Given that the NT specifically says Jesus was buried in someone else's tomb, the first and most improbable step here seems to me to have probability at least 10^-3. Given that, the rest seems very, very plausible.

When you multiply in the inability of whoever actually buried Jesus to identify the correct tomb, plus the fact that it is not usual for these sorts of hallucinations to happen to lots of different people at roughly the same time (and even if they did, they would probably conclude they were seeing a ghost instead), plus Paul for whom Jesus was not a "loved one", the odds start looking considerably lower than $10^{-3}$. And of course, it doesn't help to explain the other reported miracles in the Gospels and Acts, so you need some independent hypothesis for how these got started. In any case though, unless you were thinking that you can find 100 other hypotheses as good as this one, you seem to be conceding that 10 dB was too low.

It's also interesting how statements about the apostles for which there is some historical testimony, get to be "very, very thin", while conjectures based on no specific historical evidence are "very, very plausible".

PS today is the last day which I will be able to devote to this conversation for at least a week. While it would be rude to ask you not to reply to any of my comments, this will probably be my last comment, except for one more thing--a spreadsheet of my own.

84. lavalamp says:

g:

...if he's anything like me, the proposition he's considering is probably more like Christianity' than like Christianity-sensu-strictu to begin with.

Yes. That 2400 bit thing sorta assumes you can state Christianity in about two paragraphs.

Aron:

Huh? That's not how Bayesian updating works. If the universe is fine-tuned, all possible explanations for that fine tuning (natural or supernatural) are raised by the same factor (assuming the liklihood ratios are the same, i.e. all explanations are equally good aside from prior probabilities.) Christianity is a kind of Theism, so arguments for Theism are equally well arguments for Christianity, relative to Naturalism.

Er, yes, that was very poorly phrased on my part. What I was trying to communicate is that if Christianity is beating atheism by less than 40dB, then Deism is going to beat Christianity, because Deism is basically a version of atheism supported by the fine-tuning argument.

I don't think your "strong intuition" can possibly correct. I've explained why enough times...

OK. Given my understanding of your theory, I don't see why a historian can't make something "true" just by writing it in a book. This is what my intuition is rejecting.

As g correctly points out, the firing squad survivors will almost invariably discover that they were the victim of a conspiracy or practical joke...

Yes, I originally said something to that effect and edited it out for brevity, thinking that it was obvious that practical jokes would be among the leading causes of simultaneous 100-weapon failure. I stated it that way because I wanted to point out the observations of the firing squad survivor immediately upon surviving.

I don't think that different universes can exist more or less than each other...

That's logically equivalent to some being more probable than others, no? You have to think like this to establish background rate of atheist vs theist universes, no? (Do we all agree that this is the question that the fine tuning argument reduces to?)

"Every possibility happens" is logically equivalent to "Every possibility happens with probability 1" which is inconsistent with making probabilistic predictions about the world.

That sounds so wrong that I feel like I must be misunderstanding you (infinities can have different sizes; "every possibility happens" doesn't give information on the relative frequency. It's trivial to make a Turing machine that does both X and Y an infinite number of times but does X twice as much as Y). I think I will wait for another thread for this topic, though; hopefully you will make one eventually. I also think your take on Kolmogorov is not right but this thread is too long for that, too.

No need to respond to anything above, I mostly just wanted to clarify some poorly worded statements. (There appears to be a comment of mine above with your writing in it? I don't care if you respond that way, but I think something happened to the formatting which could confuse people.)

I added a column to my spreadsheet. I still feel completely unmoved by your arguments for fine-tuning, but I took off the complexity penalty. I don't think what I was doing is correct, but I also don't think 0 is correct, so I reserve the right to increase that in the future. My +70/-90 works out to a -20, which you think is reasonable (perhaps using different reasoning), so I'll leave it.

85. g says:

lavalamp, since you ask, I agree that the ratio background rates of universes with and without gods (whatever exactly that means!) is a large part of what the fine-tuning argument ends up being about, but I think that's by no means the end of it. There's the question of how likely it is (given what we know) that the real laws of physics turn out to explain, or refute, the fine tuning; the question of how much of possible-universe-space (not just the local bit we happen to be in) permits life; the question of how confident one can reasonably be that a god would want to make a universe that more-or-less-naturally supports intelligent life; the question of whether, if so, they'd likely prefer a universe in a "bigger" viable region of possible-universe space (e.g. because that gives more scope for the god to choose a particular set of parameters that lead to the right sort of outcomes); etc., etc., etc. Lots of tough questions of physics, theology, and philosophy here.

86. g says:

Aron, it appears that my comment about tuning the cosmological constant "at the start of time" was unclear, for which I apologize. I wasn't trying to imply that the cosmological constant varies (though so far as I know there's no fundamental obstacle to a slowly varying cosmological "constant"). What I was trying to get at is that (as I understand; I am not a cosmologist) if Lambda isn't very close to zero, you get an exponential increase or decrease in (in this case) the density of energy in the universe over time, so that the universe quickly becomes uninhabitable one way or another; so it has to be very small; but there's no currently known reason why it should be (and indeed there's a heuristic argument suggesting a ludicrously large value for it). But yes, what I wrote was a lousy way of alluding to all that. Sorry.

If there are two completely different sets of physical laws, one of which makes life "very easy" to produce and one of which makes it "very hard", I think I would fairly firmly expect a god to prefer some variant of the "very easy" laws, on the grounds that whatever else it wanted to do would be more likely to find its optimum in the universe-family with weaker constraints. This depends a bit on what we mean by "very easy"; for instance, if there's some simple set of laws with no free parameters at all that makes the emergence of intelligent life very likely, should we consider that "very easy" (if you're in that sort of universe, you probably get life, no tuning required) or "very hard" (there are very few different universes of that sort)? The fine-tuning argument is full of this sort of question, and I think very few of them have clearly correct resolutions. (On this particular one, I am inclined to go with something Kolmogorovish, so that a theory with one free parameter should be considered approximately equiprobable with an otherwise similar theory with no free parameters, and that probability distributed accordingly across the parameter space. But there are all sorts of fiddly details here that might turn out to make it a bad idea.)

I still don't see why you find Paul's conversion so monstrously improbable (beyond mere specificity). For the avoidance of doubt, of course I do understand that if it's monstrously improbable on not-Christianity and quite likely on Christianity then that makes it evidence for Christianity. What I don't understand is why you think that ratio so large.

On skeptical explanations for the phenomena you take as powerful evidence for the resurrection:

First, a meta-note. When I describe one possibility and go out of my way to say, twice, that I am not saying it's what actually happened or the most likely thing to have happened, but merely giving it as an example of the kind of thing that might have, I wish you wouldn't quote my description as if I were claiming it as fact.

The inability of whoever actually buried Jesus to identify the correct tomb is the main source of the improbability I gave. Hallucinations of recently dead loved ones are extremely common (according to the article I linked to, more people have them than don't); it wouldn't take many such -- perhaps in fact no more than one -- to spread the idea of a resurrection, if Jesus's tomb had recently been found empty; and after that, (a) you'd naturally get more people thinking they'd seen Jesus (think about what happens when a house gets the reputation of being haunted, etc.) and (b) even if the number of people making that claim were rather small, the usual processes of exaggeration and Chinese whispers seem eminently capable of producing the sort of accounts we have in the NT after a few decades.

Other miracle stories: miracle stories accrete around famous figures, especially in religious contexts. They just do. You can find them in plenty of other religions. If you think the particular stories in the NT are impressive in some way that requires more than this to be said, what way is that?

I don't see the incompatibility between reckoning only 10dB of evidence for Christianity in the NT documents and saying that there are (let's say) 10 skeptical theories of prior probability (let's say) 10^-3 each. Did you perhaps misinterpret the 10^-3 figure as an estimate of how likely the NT reports are given the theory I described, or something? That's not what it was meant to be. The other thing you may be missing is that I'm not sure Pr(NT documents | real resurrection) is as high as you're implicitly taking it to be.

"Thin" and "plausible" are not opposites. If something (1) is the kind of thing that could well have happened but (2) has very little positive evidence, it could be correct to say both that the evidence is very thin and that the thing is very plausible. For instance, I would willingly say both about the proposition that the "gospel of John" and the "second letter of John" have the same author. (Very roughly, "evidence very thin" means "probability not very close to 1" and "very plausible" means "probability not very close to 0", though of course there are all sorts of caveats.)

Final note: Aron, I saw your comment to the effect that you've little more time for this discussion at present; that's fine and I promise not to take any absence of response as a concession that I'm right about everything.

87. Aron Wall says:

I have to run, but I just want to point out that you can't actually believe this and simultaneously come up with the very high probability that you have for Christianity's truth. Your numbers, if true, mean that god's existence is blindingly obvious.

I meant blindingly obvious before doing a critical investigation into the historical evidence. There are lots of things that are nearly certain, but not obvious, because the process of becoming certain requires investigation. Obvious refers to how easy it is to know, not just to credences.

88. Aron Wall says:

lavalamp:

Er, yes, that was very poorly phrased on my part. What I was trying to communicate is that if Christianity is beating atheism by less than 40dB, then Deism is going to beat Christianity, because Deism is basically a version of atheism supported by the fine-tuning argument.

I'm not sure which 40dB ratio you're referring to here, but whatever I pick I don't see how to make sense of this comment. Are you assuming a version of Deism which woudn't be vulnerable to the Argument from Evil? In any case, Christianity is supported by the Historical Argument, modern day miracles, and certain types of religious experiences, which don't apply to Deism.

OK. Given my understanding of your theory, I don't see why a historian can't make something "true" just by writing it in a book. This is what my intuition is rejecting.

Of course a historian can't make something true just by writing it in a book, if he does so without regard to whether it is true! On the other hand, if he does careful historical research to see what is true, and only writes down those things, then he can make a statement like "At least 98% of what I just wrote down is true" probably true. What's so counterintutive about that?

Yes, I originally said something to that effect and edited it out for brevity, thinking that it was obvious that practical jokes would be among the leading causes of simultaneous 100-weapon failure. I stated it that way because I wanted to point out the observations of the firing squad survivor immediately upon surviving.

The point is that God, like the practical joke, is an actual causal explanation for our existence. It wouldn't be acceptable for the survivor of a firing squad to say "since in any case I exist, there is no need to explain my existence, and therefore the 100-weapon failure theory is just as good as the practical joke theory". Similarly, in the case of fine-tuning, you can compare God to other possible explanations (such as the multiverse), but it's not acceptable to say that it doesn't provide any evidence for any explanations at all, as I think you originally tried to claim.

I don't think that different universes can exist more or less than each other...

That's logically equivalent to some being more probable than others, no? You have to think like this to establish background rate of atheist vs theist universes, no? (Do we all agree that this is the question that the fine tuning argument reduces to?)

I don't agree at all. As a monotheist, I believe in only one God---that's why we write it with a capital "G", because in our system it's a unique proper title rather than a category of which there can be more than one. So if there are multiple universes which all exist, they would all have to share the same God. There's no such thing as a "theist" versus an "atheist" universe; either all universes have a God or none of them do. Unless you use "universe" sloppily to mean "possible set of views about the universe", but that's not what the multiverse idea is about.

I don't think that different universes can exist more or less than each other...

That's logically equival"Every possibility happens" is logically equivalent to "Every possibility happens with probability 1" which is inconsistent with making probabilistic predictions about the world.

That sounds so wrong that I feel like I must be misunderstanding you (infinities can have different sizes; "every possibility happens" doesn't give information on the relative frequency. It's trivial to make a Turing machine that does both X and Y an infinite number of times but does X twice as much as Y).

First, in standard set theory it's true that some infinities have different sizes, but 2 times an infinity is always equal in size to the original infinity, since you can put the members in one-to-one relationship. Secondly, I'm a Bayesian not a frequentist, so I deny that frequency has the same meaning as probability. To me, probabilities only make sense when you have multiple alternative possibilities of which just one is true. When all the possibilities exist, you are not dealing with probabilities but something else.

Suppose that the world contains 9 tigers and 1 bear. The frequency of bears among the animals is 1/10, but I would never say that the probability of a bear existing is 1/10. The probability of a bear existing in this hypothetical is 1, i.e with certainty, a bear exists.

Now, quantum mechanics assigns to each possible outcome of an experiment a probability value, and the probabilities aren't all equal, they can be pretty much any real number beteween 0 and 1. Suppose we consider a universe X, specified in full detail, which is more probable than a universe Y. What is the MWI person supposed to say about this situation? That there are actually infinitely many copies of X (all identical) and infinitely many copies of Y (all identical), but that there are more copies of X than Y? This runs into the problem that the set theoretical cardinalities will actually be equal, and it also contradicts my "Identity of Indiscernables" intuition that 2 identical universes can be treated exactly the same as 1 universe for all intents and purposes. Or do we say that there's just 1 copy of X and 1 copy of Y, but that X exists more intensely? But I don't know what that would mean either. It would be like saying an elephant is more probable than a mouse because it is bigger.

There appears to be a comment of mine above with your writing in it? I don't care if you respond that way, but I think something happened to the formatting which could confuse people.

89. Aron Wall says:

g:

If there are two completely different sets of physical laws, one of which makes life "very easy" to produce and one of which makes it "very hard", I think I would fairly firmly expect a god to prefer some variant of the "very easy" laws, on the grounds that whatever else it wanted to do would be more likely to find its optimum in the universe-family with weaker constraints.

This assumes that God doesn't otherwise care which of the two kinds of physical laws get implemented. But it's easy enough to imagine that one of the kinds of laws would be intrinsically more conducive to something or other that God wanted. If I go to the store and there's 1 hammer and 100 different kinds of wrenches, that doesn't necessarily make me 100 times more likely to want the wrench for whatever household chore that needs doing.

"Thin" and "plausible" are not opposites. If something (1) is the kind of thing that could well have happened but (2) has very little positive evidence, it could be correct to say both that the evidence is very thin and that the thing is very plausible. For instance, I would willingly say both about the proposition that the "gospel of John" and the "second letter of John" have the same author. (Very roughly, "evidence very thin" means "probability not very close to 1" and "very plausible" means "probability not very close to 0", though of course there are all sorts of caveats.)

Of course we have to consider the context in which a particular piece of data fits into the argument. In the particular case of Papias and Irenaeus, even if you think they are each only 90% likely to be reliable--and I think the supposedly "thin" evidence I presented is plenty to clear that bar---the testimony would still increase the probability of the Gospels being authentic by a factor of up to 100 (to the extent that they're independent of each other). That's significant enough to be worth pointing out.

I still don't see why you find Paul's conversion so monstrously improbable (beyond mere specificity). For the avoidance of doubt, of course I do understand that if it's monstrously improbable on not-Christianity and quite likely on Christianity then that makes it evidence for Christianity. What I don't understand is why you think that ratio so large.

Well, there's a reason why I compared it to Hitler being struck by lightening, coverting to Judaism, and becoming a respected Rabbi. That seems like a good example of a monstrously improbable event. It's not just that he converted, it's that naturalistically speaking he seems to have converted as a result of some sort of brain malfunction, sufficiently potent to provide a hallucination of Jesus speaking to him, despite his former extreme hostility to the religion. These sorts of things just don't happen every day. On the positive side, note how well Paul's conversion fits in with the basic Christian message that any person, no matter how bad, could conceivably be forgiven and be used by God.

Now about that "besides specificity" caveat. In a previous comment it was suggested that there's lots of different possible signs that might support a religion, that Christianity happens to have Paul's conversion but other religions will have something different. The idea seemed to be that one gets multiple shots at miracles, so if (say) there's e.g. $10^6$ possible kinds of things which might go funny in the vicinity of a new religion, you'll expect about one of them to provide a $10^{-6}$ event.

Here's the thing, though. The Gospels and Acts claim that the birth of Christianity was accompanied by many different kinds of miraculous signs of a great many kinds. But your skeptical filters are very efficient. Most of these signs are such that you find it easy to dismiss them by saying they must have been made up somewhere along the line. But in the case of Paul, this won't work because it's just so obvious from Paul's own writings that he really did used to be hostile, and that his conversion really happened. It's only because of a rather rare combination--the miracle itself being closely tied up with the existence of an early writer--that Paul's conversion can get through the skeptical filter enough that we're even discussing it. This is itself a very special feature of Paul's conversion, unlikely to be present for most signs and wonders.

Here's an analogy to illustrate the probabilities. Suppose there are $10^6$ different buckets, and each bucket contains either a white ball or a black ball. On Theory 1, the odds of a bucket containing a white ball are $10^{-6}$. Theory 2 starts off like Theory 1, but then some person deliberately replaces some significant fraction of the black balls with white balls. If you look at all the buckets and see only one white ball, this is not very strong evidence for Theory 2. But suppose that you only get to look at one randomly selected bucket, and you see it has a white ball. (Other people look at the other buckets, and tell you they see a bunch of white balls, but you don't believe them.) In this case, that one white ball is strong evidence for Theory 1.

When I describe one possibility and go out of my way to say, twice, that I am not saying it's what actually happened or the most likely thing to have happened, but merely giving it as an example of the kind of thing that might have, I wish you wouldn't quote my description as if I were claiming it as fact.

I don't think that's a fair accusation to make. My phrase "unless you were thinking that you can find 100 other hypotheses as good as this one" was an explicit recognition of the fact that you were proposing this as one possible hypotheses among many. The purpose of the quotes is to reference which thing I'm trying to respond to, not necessarily to include the full context in which the statement was made. I think we can presume that anyone who reads my comment probably read yours as well.

Indeed, I should hope you aren't proposing it as the most likely theory, because it's outrageously implausible in multiple respects. But seeing as it's the most speicific theory you've given, it's perfectly fair to criticize it on its own terms.

The inability of whoever actually buried Jesus to identify the correct tomb is the main source of the improbability I gave. Hallucinations of recently dead loved ones are extremely common (according to the article I linked to, more people have them than don't); it wouldn't take many such -- perhaps in fact no more than one -- to spread the idea of a resurrection, if Jesus's tomb had recently been found empty; and after that, (a) you'd naturally get more people thinking they'd seen Jesus (think about what happens when a house gets the reputation of being haunted, etc.) and (b) even if the number of people making that claim were rather small, the usual processes of exaggeration and Chinese whispers seem eminently capable of producing the sort of accounts we have in the NT after a few decades.

Don't think so. First of all, there's a reason why each of the 4 Gospels describes the presence of other witnesses to the burial, besides Joseph of Arimathea. The existence of multiple witnesses to the burial makes it less likely that all of them would have been mistaken together.

Secondly, your hallucination theory is totally incapable of explaining practically any of the specific details of the Gospel texts. We don't call it "usual processes of exaggeration" when a story is replaced with one which has almost no details in common with the original event. On the theory you mention, the disciples would have had a brief sensory feeling of the presence of their beloved master Jesus, looking as he always did. Such a person may feel like they've had a spritual encounter, but no one would confuse this experience with actually interacting with a person who is physically present in the ordinary way. It's the exact opposite of a stanger physically showing up, whom you don't recognize until they've talked to you for a while. Nor is it like a person who displays his wounds to a large group of people to be inspected, or who provides teaching to multiple people at the same time.

Two of the Gospels report that Peter went to the tomb and saw the burial cloths lying by themselves. This is not really a natural "exaggeration" of the original story, more like a complete fabrication. The same goes for the angels (and guard) at the tomb, as well as pretty much every detail of every specific Resurrection appearence recorded. The degree of distortion in the narratives would have to be pretty close to 100%.

And note, this distortion is supposed to have occured during a time interval, for most of which the original apostles were still living! Why didn't they correct these false stories? It's not like they weren't in charge. I think it's possible to be misled here by the phrase "oral tradition", as well. Getting these stories from AD 30 to AD 70 only requires approximately one re-telling of the story. It's as if you think those entire decades was spent doing nothing but passing the story from one person to the next, with each person dying just after the next person receieves the message!

Other miracle stories: miracle stories accrete around famous figures, especially in religious contexts. They just do. You can find them in plenty of other religions. If you think the particular stories in the NT are impressive in some way that requires more than this to be said, what way is that?

Sure, there are other miracle stories for other famous figures. Of course, most of them aren't corroborated by multiple historical sources, are invented centuries rather than decades later, are not present in historical sources which provide sober testimony to boring fatual details and have people react in realistic ways, or are in sources which were clearly intended to be fictional, and so on. Seldom is there any suggestion that people risked death for making the original claim. (Plus, the shear quantitity of miracles in the Gospels and Acts is rather surprising, considering the comparative shortness of the length of time, and the number of details which don't seem to be fictional.) If you think there's a parallel example with these same features in another religion, why don't you name it specifically?

90. lavalamp says:

I'm not responding to everything in a (mostly failed) attempt to keep things short.

Er, yes, that was very poorly phrased on my part. What I was trying to communicate is that if Christianity is beating atheism by less than 40dB, then Deism is going to beat Christianity, because Deism is basically a version of atheism supported by the fine-tuning argument.

I'm not sure which 40dB ratio you're referring to here, but whatever I pick I don't see how to make sense of this comment. Are you assuming a version of Deism which woudn't be vulnerable to the Argument from Evil? In any case, Christianity is supported by the Historical Argument, modern day miracles, and certain types of religious experiences, which don't apply to Deism.

Say for example that, before updating on the fine-tuning argument, your probabilities for (christianity, deism, atheism) are (-20, 0, 0). Then, if one thinks the fine-tuning argument is 40dB, after updating on it, your probabilities are (20, 40, 0). (Numbers chosen only to illustrate my point.)

... Similarly, in the case of fine-tuning, you can compare God to other possible explanations (such as the multiverse), but it's not acceptable to say that it doesn't provide any evidence for any explanations at all, as I think you originally tried to claim.

My point all along (perhaps poorly expressed) is that surviving a firing squad gives you no additional information on the relative likelihood of pratical jokes and 100-weapon-misfires; that's background information, i.e., your prior. The fact that we exist gives no additional information on the relative frequencies of the processes that could cause intelligent agents to exist. The argument therefore is about the background rates for all of those possible processes. I'm not claiming it doesn't need an explanation, I'm claiming that it's not evidence one way or the other about the explanation.

... Secondly, I'm a Bayesian not a frequentist, so I deny that frequency has the same meaning as probability. To me, probabilities only make sense when you have multiple alternative possibilities of which just one is true. When all the possibilities exist, you are not dealing with probabilities but something else.

Suppose that the world contains 9 tigers and 1 bear. The frequency of bears among the animals is 1/10, but I would never say that the probability of a bear existing is 1/10. The probability of a bear existing in this hypothetical is 1, i.e with certainty, a bear exists.

Now, quantum mechanics assigns to each possible outcome of an experiment a probability value, and the probabilities aren't all equal, they can be pretty much any real number beteween 0 and 1. Suppose we consider a universe X, specified in full detail, which is more probable than a universe Y. ...

My understanding, which may or may not be correct, is that those who hold MWI believe that we're in an epistemic state of "indexical uncertainty". To continue your bear/tiger analogy, it's like every morning you wake up as either a bear or a tiger, but you don't know which until you open your eyes and look at the mirror in your room. Hence the relative frequencies are also our subjective probabilities (specifically, they're the background rate and should be our prior). Or in other words, "universe X exists with probability 1" can be true at the same time as "I will experience existing in universe X with probability < 1". Or do you think I misunderstand MWI proponents? Things get weird when the multi/universe is constantly generating copies of you...

I don't agree at all. As a monotheist, I believe in only one God---that's why we write it with a capital "G", because in our system it's a unique proper title rather than a category of which there can be more than one. So if there are multiple universes which all exist, they would all have to share the same God. There's no such thing as a "theist" versus an "atheist" universe; either all universes have a God or none of them do. Unless you use "universe" sloppily to mean "possible set of views about the universe", but that's not what the multiverse idea is about.

Hm. This strikes me as incoherent. Your "God" is/has, among other things, a mind capable of thought processing. Minds are information-processing devices and are necessarily made of smaller pieces working together by rules (recall you granted earlier that there's probably no souls). Therefore "God" must exist in a substrate of some sort. We could probably call this substrate a universe, or say that it exists in one. Hence you have an infinite regress problem. (I expect you to disagree at some point in this chain, but I don't know where so I'll leave everything as bald assertions and supply evidence/reasoning where needed later.)

I'm not sure that I can use "universe" sloppily until we have a non-sloppy meaning for it. :) If it has some precise definition within your field, I don't know it. If I'm talking about "possible universes" or "logically possible universes" I'm meaning something like "coherent states by which some sort of reality could possibly be arranged, even if it's not our reality", and this is probably an idea as much or more from philosophy than physics. I see "multiverse" as a generic term, e.g., MWI describes one sort of multiverse, and MUH describes another. MUH seems to preclude any additional levels on top of it.

91. Aron Wall says:

lavalamp:

Say for example that, before updating on the fine-tuning argument, your probabilities for (christianity, deism, atheism) are (-20, 0, 0). Then, if one thinks the fine-tuning argument is 40dB, after updating on it, your probabilities are (20, 40, 0). (Numbers chosen only to illustrate my point.)

Sure, if those are the priors I agree.

My point all along (perhaps poorly expressed) is that surviving a firing squad gives you no additional information on the relative likelihood of pratical jokes and 100-weapon-misfires; that's background information, i.e., your prior. The fact that we exist gives no additional information on the relative frequencies of the processes that could cause intelligent agents to exist. The argument therefore is about the background rates for all of those possible processes. I'm not claiming it doesn't need an explanation, I'm claiming that it's not evidence one way or the other about the explanation.

Right. It doesn't tell you anything about which explanation to pick, so long as we assume that they all are explanations (more accurately, that their likelihood ratios are equal...). It is not at all clear to me that the multiverse should count as an explanation: normally we explain anomalies in the data by postulating theories which account for the data, not by postulating that the data is part of an even bigger set of unobservable data, in which the anomalies happened by chance. But this whole subject is controversial since different people have different ideas about how to handle observer selection effects.

Suppose I concede for the sake of argument that, if there is a multiverse M with more than $10^{120}$ worlds, plus some mechanism for giving them each different laws of physics, then this would count as an explanation for the fine-tuning of the cosmological constant. It still doesn't follow that fine-tuning gives no evidence for G, God's existence. It depends on the prior probabilities of M and G (and any other explanations you might want to consider). Supposing fine-tuning to be the only consideration, you'll end up accepting whichever explanantion you assign the highest prior probability. But that doesn't mean fine-tuning isn't evidence for the other explanantion. Since fine-tuning is equal evidence for all candidate explanations, it has to raise the probabilities of all of them equally. For example, suppose that I assinged M a prior of $10^{-2}$ and G a prior of $10^{-6}$. Then after I discover that fine-tuning is necessary, M goes up to nearly 1 while G goes up equally to $10^{-4}$. In this way fine-tuning does provide positive evidence for the existence of God.

The only way for fine-tuning to give zero evidence for G, is if you were already convinced of a rival explanation (such as M) before taking into account evidence of fine-tuning. Personally I think it's crazy to be certain of M without any evidence at all, but your taste may differ.

I'm not sure how you get 20-40dB of evidence from the fine-tuning argument. I don't think it's bayesian evidence either way, since P(we'll find ourselves in a universe we can live in) is near 1 by definition, no matter what. But if I hadn't thought of that, I'd expect you to [do an analysis involving the string theory multiverse]

You are the expert in what your beliefs were. But it sure looks to me like you originally thought that fine-tuning shouldn't count as evidence even if there is no multiverse, since you bring up the multiverse as an independent argument.

My understanding, which may or may not be correct, is that those who hold MWI believe that we're in an epistemic state of "indexical uncertainty".

Yes, I agree that's what they think. It just doesn't make any sense to me that one should handle the probabilities in the way that they do. If there's a library with every possible book written in it, I don't think that any of them are really quite real since none of them are privileged in any way. If I went to the library and went to the right floor to check out Hamlet, my work in finding the right place would be tantamount to writing Hamlet. If everything exists equally, than nothing really exists at all.

Of course, MWI'ers do believe some worlds are more privileged than others, the ones which are more probable. If you're David Deutsch, you explain this by saying that each world exists in an infinite number of identical copies, but that some of the infinities are bigger than others. From a set theory point of view this doesn't make much sense to me.

Other MWI adherents say that the whole phrase "many worlds" is a misnomer, and the real way to interpret QM is just that the wavefunction is the only thing that exists, and it never collapses. But then you still have to figure out how to get the Born rule to come out of the wavefunction.

I'm not sure that I can use "universe" sloppily until we have a non-sloppy meaning for it.

I was only trying to distinguish between 1) multiverses, where one postulates that there's a bunch of universes that ALL exist simultanoeusly, and 2) "universes" in the sense of rival conceptions of the world of which at most ONE is really the case. Confusingly, the entire multiverse in sense (1) is a single universe in sense (2).

Hm. This strikes me as incoherent. Your "God" is/has, among other things, a mind capable of thought processing. Minds are information-processing devices and are necessarily made of smaller pieces working together by rules (recall you granted earlier that there's probably no souls). Therefore "God" must exist in a substrate of some sort. We could probably call this substrate a universe, or say that it exists in one. Hence you have an infinite regress problem. (I expect you to disagree at some point in this chain, but I don't know where so I'll leave everything as bald assertions and supply evidence/reasoning where needed later.)

I agree that our minds are, or at least involve, information-processing divices containing smaller pieces working together by rules (though I side with Chalmers rather than Dennet, in that I don't think pure logic is sufficient to deduce that an information-processing divice would necessarily be conscious.) But that doesn't mean that all minds (let alone things like God which are only analogous to minds) must work in the same way. I denied the existence of what you call souls, not because I think they are logically incoherent, but because there's a lot of specific scientific evidence that we think with our brains. That doesn't automatically rule out the existence of another, immaterial and immortal component of our selves, but I don't think this additional component solves Chalmer's "hard problem" of consciousness nor is necessary in my opinion to explain the afterlife, and therefore I see no specific reason to believe in it. But that's because of Occam's razor, not intrinsic logical impossibility. So although I disbelieve in what you call souls, I probably don't find them quite as implausible as you do.

A brief note on terminology. In my language, souls do exist and I would continue to believe in them even if I were an atheist. To me the term means "the core aspects of who I am, the pattern of life and intelligence within me". The ancients thought the same way, which is why they said that animals also have souls ("anima" is Latin for soul, so "animal" literally means something with a soul!) I think if you tried to convince an ancient Hebrew, Greek, or Roman that my cat had no soul, they would reply, "What do you mean? It's alive, it moves around and breathes and has intentions, of course it has a soul. The question is what kind of soul." I simply deny that this soul is immaterial or inherently immortal (in my theology immortality is by grace rather than nature).

But now, back to God. There are a number of highly relevant differences between God's mind and my own. First of all, my mind is limited and only knows some things. What I know is contingent and can change with time. I process information by constructing in my own mind "representations" of the thing I'm thinking about, which differ from the original object by 1) existing in a different medium, 2) from one particular angle or perspective, 3) having limited resolution, and 4) including "errors" when my senses or reasoning are mistaken. All of these features suggest the existence of something like a "brain" which processes the information in question.

But in the case of God, his knowledge of the world is complete, unchanging, and perfect. It takes an enormous amount of information to specify which things I know, but it takes almost no information to specify which things God knows, since the rule is he knows everything. I also think that the divine mind is probably non-representational. For suppose that God knew things by storing them in his "mind" somewhere. Since his knowledge is perfect, his mental images would have to be exactly the same as the real world. But then there would be two complete copies of the universe, identical in every way. This seems like unnecessary duplication; it makes more sense to say that the divine ideas and the real world are one and the same thing.

Similarly, it does not take a "brain" to specify which things God desires and wants, since for us classical theists what God wants is strictly identical to the ethically good. You have to accept the existence of an objective ethics, but after that one simply says this objective ethics is the exact same entity as God's desires, neatly sidestepping the Euthyphro dilemma of which came first. (You could argue in reverse and say, if there exists an objective ethics, that's already very similar to saying that the fundamental nature of reality is analogous to a mind, to the extent that "it" regards certain things as good or bad.)

In other words, God is an eternal mind whose subjective thoughts are strictly identical to objective reality. This is sufficiently different from you and me that I don't think it's wise to generalize from the fact that our thought is in brains. We are but the "image of God", minds transposed into the iherantly limited medium of information-processing systems.

Note also that classical theism is quite simple. Once you're willing to get past the initial hurdle that God's nature is specified using "personalist" language rather than some mathematical algorithm (which I think is the primary sticking point for those who have accustomed themselves to naturalist thinking), the classical theist God is just about the simplest kind of "mind" one can conceive of. Hence it seems suitable as a candidate theory for the fundamental nature of reality. (Of course, Christian theology adds additional layers of complexity on top of the classical theist notion, but these were forced on us by the historical data rather than by philosophical reasoning.)

I'm not responding to everything in a (mostly failed) attempt to keep things short.

I do appreciate this, even if I show my appreciation in a somewhat strange way by writing long treatises on theology in response. :-)

Regrettably, I do have limited time and---what's actually more limiting---emotional energy to devote to this (it would be less burdensome to my "soul" if I could carry on this argument without caring about the truth and whether I convince you or not. But for better or for worse, this is not the case). I may write one or two more posts in the current series (and respond to comments on those threads), but after that I'll probably need to blog about less controversial subjects for a couple months, before returning to the arena to discuss e.g. modern-day miracles. Not that you are forbidden from replying to this comment, but I may or may not respond to your reply.

For whatever it's worth to you, you and g are in my prayers. Normally I think it's a sign of a superficial spirituality to pray solely for people's conversions (saving their "souls") without caring about the rest of their life, but one of the limitations of this medium is that I don't know anything about the rest of your life! So you'll have to take that as a stand-in for everything else, I guess. Whatever conviviality would have been expressible if we'd been debating in a coffee shop instead of online, let's pretend that it was virtually present. Aron.

92. lavalamp says:

First, a short response to fine tuning which hopefully will end that particular thread.

I'm not sure how you get 20-40dB of evidence from the fine-tuning argument. I don't think it's bayesian evidence either way, since P(we'll find ourselves in a universe we can live in) is near 1 by definition, no matter what. But if I hadn't thought of that, I'd expect you to [do an analysis involving the string theory multiverse]

You are the expert in what your beliefs were. But it sure looks to me like you originally thought that fine-tuning shouldn't count as evidence even if there is no multiverse, since you bring up the multiverse as an independent argument.

Let me clear this up. The fact that we exist is powerful evidence against all theories that say we shouldn't, and tells us nothing about the remaining theories. I'm pretty sure that we agree on this, given what you wrote just above what I quoted. If we consider three hypotheses:

* M: One or all of: Multiverse/MWI/Everything that can exist, does
* R: Our universe is randomly the only one and just happens to support life
* G: God did it

The prior probabilities of these positions don't change, relatively speaking, after conditioning on our existence, because they all predict that we'll exist. In terms of prior probability, I think R beats G (see below), but M is vastly more probable than either (on account of vastly lower complexity). The fact that we exist doesn't change this assessment. Hence that thing I said earlier about fine tuning really being an argument about priors, which I guess we really do agree on after all?

(Final note: G is strictly less probable than R if you go by Kolmogorov and expand it to "[who] did [what]"; R expands only to "[what]". "You can't compress the laws of physics by adding in a god," would be another way of saying this. But of course this is an argument about priors...)

(Final final note: I just wanted to say that if MUH happens to be true, I do agree that this gives very little insight as to which particular mathematical structure we inhabit.)

I haven't read the rest of your comment in detail yet, because I got excited about the thought of putting this to rest. :)

93. g says:

(Blog comment threads really are a pretty poor venue for this sort of discussion, aren't they?)

But it's easy enough to imagine that one of the kinds of laws would be intrinsically more conducive to something or other that God wanted.

Yeah, it's easy enough to imagine, but "easy to imagine" is a poor surrogate for actual probability. (Hence the representativeness and availability heuristics, the "conjunction fallacy" where Linda is "more likely" to be a feminist bank teller than a bank teller, etc.) In the absence of any actual knowledge about these various hypothetical possible sets of laws, you don't get to say "no, a god wouldn't be more likely to choose the 'easy' set because he might have had reasons to prefer the 'hard' set".

In the particular case of Papias and Irenaeus, even if you think they are each only 90% likely to be reliable -- and I think the supposedly "thin" evidence I presented is plenty to clear that bar

I don't. And I don't see any reason to think they're close to independent, either. The approach you're taking here seems to me to be one of wholesale credulity. That may be appropriate in many cases -- much of the time, no one has much incentive to lie or stretch the truth -- but it just won't do here. We know that early Christians of various sorts produced pseudonymous gospels. There are obvious motives for producing such things. All the positive things we know about Papias put together give no actual grounds for thinking he'd know whether Matthew was written by Matthew or not. (So maybe he talked to the apostle John. That's cool and all, but why would that make him an expert on the authorship of Matthew? Why would John necessarily know that? What reason have we to think that Papias asked him?) Irenaeus was a century or so after Matthew was written; why should we be 90% confident about his opinions about its authorship? Especially given how strange what he says about Matthew is ("a written gospel in their [the Hebrews'] language", when Matthew as we actually have it shows no sign of a Hebrew or Aramaic origin).

Everything we know about Paul's conversion comes from Paul via Luke (taking the assumption, which is surely the most optimistic one, that Luke's account in Acts 9 is based on information from Paul). So what we know, at most, is that Luke said Paul said he had an encounter with Jesus. He seems to have described it in inconsistent ways on two occasions, which may or may not matter. (I don't think it matters much.) Did he actually have such an experience? Who knows? It's not unknown for people to make such things up, I'm afraid.

Everything we know about Saul's previous persecution of the Christians also comes from Paul (most of it via Luke). Anyone with any experience of evangelical Christianity is familiar with the tendency of converts to exaggerate their pre-conversion badness; I wouldn't place any large bet that Saul was anywhere near as big a persecutor of Christians as Acts makes him out to have been. Even if he was, he's hardly an analogue of Hitler. (Perhaps if Caiaphas or Pilate had converted they would have been.) So, more like some random concentration camp official converting to Judaism.

It seems to me that if you look at Jesus's message, as reported by the gospels, rather little of it has anything to do with how "any person, no matter how bad, could conceivably be forgiven and be used by God". That's a much more prominent theme in the teaching of ... Paul. Which, if you think about it, is hardly a surprise.

it's just so obvious from Paul's own writings that he really did used to be hostile, and that his conversion really happened

So obvious? I'm unconvinced. I am, as it happens, inclined to think that he really was hostile and really was converted. But if I weren't, the obvious alternative hypothesis would be that for some reason he lied about those things, and I don't see anything in his writings that particularly rules that out. What his writing shows is that he wanted to present himself as a formerly hostile convert. The most likely explanation is of course that he actually was one, but if you see evidence that goes beyond the fact of his saying so then I'd be interested to know what.

But suppose that you only get to look at one randomly selected bucket, and it has a white ball.

I agree that this is much stronger evidence for Theory 2 than simply finding a white ball somewhere in the buckets. But I don't see how the story of Paul is much like this. Where's the miracle? (There are some miracles in the story as found in Acts, but those no more pass a reasonable "skeptical filter" than any of the other miracle stories in the gospels and Acts, or in the origin stories of any number of other religions.)

The fact that an opponent of Christianity converted to Christianity is not good evidence of a miracle. Opponents of things change sides all the time. (Random example: one prominent atheist is a chap called Dan Barker, who used to be a Christian preacher. He's now the head of an organization called the Freedom From Religion Foundation.) The fact that said opponent of Christianity became a prominent Christian is not good evidence of a miracle. (Dan Barker is pretty prominent too; probably not by coincidence; some cachet attaches to that former-enemy label.) Where's the miracle?

there's a reason why each of the 4 Gospels describes the presence of other witnesses to the burial, besides Joseph of Arimathea

The gospels' conspicuous disagreement about what other witnesses were present is interesting. (The Synoptics say it was a couple of Maries, or in Luke's case some not-so-precisely specified women. John says it was Nicodemus. All seem to me to suggest that they've given a full list. I'm sure Mary Magdalene and Nicodemus were easily confused with one another.) Anyway: yes, no doubt there's a reason. The reason may be that saying there were witnesses was a useful way to contradict skeptical theories like the one I mentioned. Unfortunately, saying there were witnesses is not quite the same thing as there actually having been witnesses. (Though actually my proposal is not that JoA buried Jesus alone, but that JoA procured the body and then left it to others -- perhaps the Maries -- to do the actual burying, and that they got the place wrong. So it's the presence of JoA that I'm calling into question, more than that of the alleged other witnesses.)

your hallucination theory is totally incapable of explaining practically any of the specific details of the Gospel texts

That's OK, because practically all of those specific details are different (and in many cases outright contradictory) from one gospel to another. There's nothing particularly awful about that -- observation and memory are startlingly unreliable, as it turns out -- but it does mean that we can't be very confident that any details at that level are actually correct.

Anyway: yes, I am proposing that over a few decades what started out with a few disciples' ambiguous experiences turned into much more elaborate stories involving more people and stranger happenings. Tales grow in the telling, especially this sort of tale. (That is: stories about wonders wrought by key figures of one's religion.) You can see a little -- alleged -- example of this even in the NT itself: according to John's gospel it was widely believed that Jesus had said in so many words that John would not die until the Second Coming, whereas in fact he had merely made an offhand dismissive remark.

It's as if you think those entire decades were spent doing nothing but passing the story from one person to the next, with each person dying just after the next person receives the message!

What is? For the avoidance of doubt, of course I don't think any such stupid thing and I don't know why you suggest it.

If you think there's a parallel example with these same features in another religion, why don't you name it specifically?

I don't think there's an example with those features in Christianity! The resurrection-appearance stories are generally not corroborated by multiple sources; each source has its own set of stories, and in the cases where they overlap generally nothing but the barest kernel of the story is consistent from one source to another. The sources don't "provide sober testimony to boring factual details" any more than, I dunno, Grimm's fairytales do; that is, of course there are some boring details in there, but boring details can be found anywhere, from good history to outright fantasy and all points in between; the sources we have are hardly "sober" in the sense of showing any sort of skepticism about miracle stories. As I have said before, the evidence that the early Christians risked death for claiming what they did is pretty poor, and in any case it's only evidence that they believed it, not that it was true. I'm afraid I don't understand why you think the fact that there are lots of miracles in the gospels and Acts is evidence for their veracity.

94. g says:

(This is in response to Aron replying to lavalamp.)

If you're David Deutsch, you explain this by saying [...] some of the infinities are bigger than others. From a set theory point of view this doesn't make much sense to me.

What about from a measure theory point of view? That would seem more appropriate.

I do agree that where the Born rule comes from> is a bit of a mystery, but it's just as much a mystery on any interpretation of QM. And, for what it's worth, handwavily the Born rule seems pretty natural: given that you need to put a measure on your Hilbert space, it's a very natural measure to use. Deutsch and his followers have a way of getting the Born rule decision-theoretically; it feels a bit like a sleight of hand somehow, but I think there may be some insight there.

one simply says this objective ethics is the exact same entity as God's desires, neatly sidestepping the Euthyphro dilemma of which came first.

The Euthyphro dilemma isn't about which was chronologically first. It's about logical priority. Does God want X because X is good, or is X good because God wants X? This question can still be asked if God's desires and what's good are the exact same thing. (One answer, for instance, might be that "good" really means "in accordance with God's desires", in which case things are good because God wants them. Another might be that God is morally perfect, so that of course he desires exactly what's good, in which case God wants things because they're good.) Of course there's a third answer: neither is because of the other, it's just a brute fact that they're the same. But that seems very unsatisfactory.

I have another concern about the notion of classical theism you're proposing here: it seems awfully close to making God a sort of epiphenomenon. God's believing that X is simply the same thing as X being the case; God's wanting X is simply the same thing as X being good; it seems like a universe having this sort of God is indistinguishable from one in which those things happen, and those other things are good, without reference to God. The missing factor, I think, is God's agency; to make a difference he must actually do something. But as soon as you introduce that, I think the claim of supreme simplicity goes out of the window.

(I can't speak for lavalamp, but you're welcome to pray for me if it pleases you to do so. It seems unlikely to do much harm and of course if you're right it might do much good. My expectation, of course, is that the main thing it will do is to waste a bit of your time :-).)

95. g says:

(Ecch. Apparently blockquoting automatically italicizes here, so my attempts to get italicized blockquotes by nesting an i inside a blockquote of course did the exact opposite of what I intended. Never mind.)

96. lavalamp says:

OK, on everything else (mostly written before I saw g's comments). To start at the end: of course you are not obligated to respond to this. This sort of conversation stopped costing me so much emotional energy right about the time I deconverted, but I understand how taxing it can be. I haven't seen you claim the free will defense yet, so it doesn't seem inconsistent for you to pray for the souls of others. :) And it seems that in about a week I will find myself living not so terribly far from you, as I've been absorbed by a certain large tech company-- if you ever find yourself in the Bay Area, I would be thrilled to continue this in an actual coffee shop. :)

Other MWI adherents say that the whole phrase "many worlds" is a misnomer, and the real way to interpret QM is just that the wavefunction is the only thing that exists, and it never collapses. But then you still have to figure out how to get the Born rule to come out of the wavefunction.

This is sorta my understanding, I guess, although both things you mention sound the like they describe the same thing to me. It's also my understanding that no one else has a better explanation for the Born probabilities, please tell me if that's not correct. I'm not quite sure if I should interpret your comments about MWI as "Aron thinks Aron is confused" or "Aron thinks MWI is confused"-- I know I'm confused about QM in general, so I'm not really going to be able to contribute here. :)

... There are a number of highly relevant differences between God's mind and my own. ... in the case of God, his knowledge of the world is complete, unchanging, and perfect. It takes an enormous amount of information to specify which things I know, but it takes almost no information to specify which things God knows, since the rule is he knows everything.

Your god is the MUH?? (I don't mean to score cheap points here, but that's the entity that takes the least amount of information to specify...) :)

More seriously, if god knows everything, then (recall your library analogy) it still takes many bits of information for him to think about anything in particular, because he has to locate it in the vast space of things he knows.

I also think that the divine mind is probably non-representational. For suppose that God knew things by storing them in his "mind" somewhere. Since his knowledge is perfect, his mental images would have to be exactly the same as the real world. But then there would be two complete copies of the universe, identical in every way. This seems like unnecessary duplication; it makes more sense to say that the divine ideas and the real world are one and the same thing.

So for your god, the map *is* the territory. Can your god do hypothetical reasoning? If not, how can he plan anything? If so, do billions of sentient creatures die every time he thinks about something else?

Similarly, it does not take a "brain" to specify which things God desires and wants, since for us classical theists what God wants is strictly identical to the ethically good.

Naming his desires "that which is ethically good" doesn't let you avoid specifying them. (Also, I'm inclined to think that the argument from evil is -60dB against this god.)

You have to accept the existence of an objective ethics, but after that one simply says this objective ethics is the exact same entity as God's desires, neatly sidestepping the Euthyphro dilemma of which came first.

Ethics is (in my view) just multi-agent game theory, which is objective once every agent's values are specified, but doesn't help you in this argument since there's no universally valid set of values.

In other words, God is an eternal mind whose subjective thoughts are strictly identical to objective reality. This is sufficiently different from you and me that I don't think it's wise to generalize from the fact that our thought is in brains. We are but the "image of God", minds transposed into the iherantly limited medium of information-processing systems.

It's not clear that your god has any dynamic components. In fact it sounds to me like a relabeling of naturalism, where "physical laws" means "god's desires" and "state of universe" means "god's knowledge". It definitely seems incompatible with the Christian idea that god is a person. It sounds more like deism or animism.

I'd also argue that, even without solving the hard problem of consciousness, we can be quite certain that minds at least involve processing information, whatever else they may do, and that this requires a computing device. IOW, minds might be more than computers, but they can't be less than computers.

I'll also argue that if you think god's thoughts form our reality, that still implies that there's something bigger than our reality that god must exist in, so you haven't solved the regress problem this way. If your god is identical with our reality (there's nothing left over after you remove the thoughts concerning us), then I don't think it's fair to describe what you believe in as "god".

97. Aron Wall says:

Dear g and lavalamp,

This will be my last reply in this particular conversation. I suspect that further discussion would not result in our coming to agreement, for various systematic reasons that I believe have to do with your biases. (Doubtless you would want to say the same thing about me, but naturally I believe that I'm the one who is right!) Your a priori philosophical priors are too low and your cynicism about testimony is too high, so I can't make any progress here.

Leaving aside more general philosophical considerations such as the argument from evil, I still claim that the specific historical record looks more or less as one would expect it to, if the Resurrection and certain other early Christian miracles had really happened. And conversely, it does not at all look to me the way I would expect if it had all been made up (see below).

lavalamp,
I didn't say that God takes no information to specify (in fact I do not believe that I have fully specified him!). I'm saying that in the case of an omniscent being there is no need for a "brain"-like entity specifying which things God knows, since he knows everything that exists. (Not the same as the MUH which says that everything exists.) And in this sense he is a lot simpler than other persons. Nor am I arguing that God has no reality above and beyond the universe. Yes, we are thoughts in the mind of God, but there's something else that exists on top of the phyiscal universe, namely God himself, who was there before the physical universe began. Hamlet is one of Shakespeare's thoughts but Shakespeare has a reality above and beyond the reality of Hamlet and his other characters.

I suppose God does hypothetical reasoning by simultaneously knowing all possible things that he might create. But I don't really know since I'm not God. I also don't get your point about "thinking about things in particular" since it's not like God has to stop thinking about one thing to start thinking about other things. Anyway, in my view God is the fundamental entity which created everything else. He didn't originate in any kind of other world, he had no beginning or end--I'm sure you know the standard Christian doctrines on this since you used to be a Christian.

lavalamp:

* M: One or all of: Multiverse/MWI/Everything that can exist, does
* R: Our universe is randomly the only one and just happens to support life
* G: God did it
...G is strictly less probable than R if you go by Kolmogorov and expand it to "[who] did [what]"; R expands only to "[what]". "You can't compress the laws of physics by adding in a god," would be another way of saying this. But of course this is an argument about priors...)

The odds of the cosmological constant being as small as it is randomly (as opposed to by some putative physical or supernatural mechanism that selected a specific value near 0) is about $10^{-120}$. You can't actually think that God is less probable than that.

By defining R to include the statement that the universe "just happens" to support life, all you're doing is moving improabability from the P(life | Naturalism) to P(R). It doesn't get rid of the huge number.

Your argument that G is strictly less probable than R is specious; if that were a logically valid mode of inference you could use it against every kind of causal agent, not just God. The point is that Naturalists aren't allowed to say things like "the laws of nature were selected in order to produce life", while Theists are allowed to say that.

g:

The Euthyphro dilemma isn't about which was chronologically first. It's about logical priority. Does God want X because X is good, or is X good because God wants X? This question can still be asked if God's desires and what's good are the exact same thing.

Here's some more questions like that one: Does a triangle have three sides because it has three angles, or does it have three angles because it has three sides? Is a particle charged because it has an electric field, or does it have an electric field because it is charged? Is killing an innocent person without good reason wrong because murder is wrong, or is murder wrong because killing an innocent person without good reason is wrong? Which is prior, logic or math?

I don't think the question of which of two identical things is logically prior necessarily has a well-defined answer. I believe that God's desires and ethics are one and the same entiity: neither preceeds the other either chronologically or logically. As part of God's nature, they just ARE and everything else is a consequence of that. (By the way, the original version of the Euthyphro dilemma was written down by Plato, an ethical monotheist, arguing against polytheistic piety. I think it was a lot more cutting in that form.)

lavalamp,
I didn't say that God takes no information to specify (in fact I do not believe that I have fully specified him in any way). I'm saying that in the case of an omniscent being there is no need for a "brain"-like entity specifying which things God knows, since he knows everything that exists. (Not the same as the MUH which says that everything exists.) And in this sense he is a lot simpler than other persons. Nor am I arguing that God has no reality above and beyond the universe. Yes, we are thoughts in the mind of God, but there's something else that exists on top of the phyiscal universe, namely God himself, who was there before the physical universe began. Hamlet is one of Shakespeare's thoughts but Shakespeare has a reality above and beyond the reality of Hamlet and his other characters.

I suppose God does hypothetical reasoning by simultaneously knowing all possible things that he might create. But I don't really know since I'm not God. I also don't get your point about "thinking about things in particular" since it's not like God has to stop thinking about one thing to start thinking about other things. Anyway, in my view God is the fundamental entity which created everything else. He didn't originate in any kind of other world, he had no beginning or end--I'm sure you know the standard Christian doctrines on this since you used to be a Christian.

g:

The Euthyphro dilemma isn't about which was chronologically first. It's about logical priority. Does God want X because X is good, or is X good because God wants X? This question can still be asked if God's desires and what's good are the exact same thing.

Here's some more questions like that one: Does a triangle have three sides because it has three angles, or does it have three angles because it has three sides? Is a particle charged because it has an electric field, or does it have an electric field because it is charged? Is killing an innocent person without good reason wrong because murder is wrong, or is murder wrong because killing an innocent person without good reason is wrong? Which is prior, logic or math?

I don't think the question of which of two identical things is logically prior necessarily has a well-defined answer. I believe that God's desires and ethics are one and the same entiity: neither preceeds the other either chronologically or logically. As part of God's nature, they just ARE and everything else is a consequence of that. (By the way, the original version of the Euthyphro dilemma was written down by Plato, an ethical monotheist, arguing against polytheistic piety. I think it was a lot more cutting in that form.)

In the absence of any actual knowledge about these various hypothetical possible sets of laws, you don't get to say "no, a god wouldn't be more likely to choose the 'easy' set because he might have had reasons to prefer the 'hard' set".

In the absence of knowledge I do what every Bayesian does, namely pick priors. So long as the "hard" and "easy" laws are quite different from each other in other respects, it doesn't strike me as unlikely that God would have a relevant reason for picking from the "hard" set.

The approach you're taking here seems to me to be one of wholesale credulity. That may be appropriate in many cases -- much of the time, no one has much incentive to lie or stretch the truth -- but it just won't do here. We know that early Christians of various sorts produced pseudonymous gospels.

We know that they did so well after the apostles and their first hearers were dead. That doesn't mean they'd be able to do so earlier on, when the apostles were still living or recently died. Furthermore, the historical quality of these other Gospels is quite clearly different even on a superficial look, and they seem never to have been accepted by most Christianity. The question is how hard would it be to produce forgeries which otherwise looks authentic and have other people testifying to them.

All the positive things we know about Papias put together give no actual grounds for thinking he'd know whether Matthew was written by Matthew or not. (So maybe he talked to the apostle John. That's cool and all, but why would that make him an expert on the authorship of Matthew? Why would John necessarily know that? What reason have we to think that Papias asked him?) Irenaeus was a century or so after Matthew was written; why should we be 90% confident about his opinions about its authorship? Especially given how strange what he says about Matthew is ("a written gospel in their [the Hebrews'] language", when Matthew as we actually have it shows no sign of a Hebrew or Aramaic origin).

Because most historical sources are reliable 90% of the time? I'm not using any special pleading here; I'd give the same credit to say a Muslim author talking about the authorship of some Islamic text 2 generations earlier. I've already addressed the Hebrew Matthew thing above so I won't repeat myself.

The sources don't "provide sober testimony to boring factual details" any more than, I dunno, Grimm's fairytales do; that is, of course there are some boring details in there, but boring details can be found anywhere, from good history to outright fantasy and all points in between; the sources we have are hardly "sober" in the sense of showing any sort of skepticism about miracle stories.

It's hard for me to clearly communicate the heuristics I use, which make me feel like a text is "fact" instead of "fiction". But obviously we all have these heuristics since most of the time (leaving religion aside) we don't have much problem deciding which books fall into which category. My phrase which you quote probably does a bad job communicating clearly what these heuristics are, but for me they seem fairly vivid and clear when I apply them in specific cases.

Of course, one of the heuristic which we sometimes use is that stories containing miraculous or fantastic events are less likely to be true. But this can't be the the only heuristic we use since that would beg the entire question. In every other respect, these heuristics positively scream to me that something like Mark or Acts isn't just "made up".

The fact that you cite Grimm as a parallel example shows that there must be a gigantic disconnect between us. I believe in miracles, and yet I also find it totally obvious that the fairy tales in the Brothers Grimm are stylized narratives that were never intended as anything other than fiction. No one in their right mind could possibly take the Brothers Grimm as "sober testimony" the way religious people take the Gospels. If you can't detect the relevant differences in literary style, you're not paying much attention. As C.S. Lewis says in his essay on Modern Theology and Biblical Criticism:

In what is already a very old commentary I read that the fourth Gospel is regarded by one school as a 'spiritual romance', 'a poem not a history', to be judged by the same canons as Nathan's parable, the book of Jonah, Paradise Lost 'or, more exactly, Pilgrim's Progress'. After a man has said that, why need one attend to anything else he says about any book in the world? Note that he regards Pilgrim's Progress, a story which professes to be a dream and flaunts its allegorical nature by every single proper name it uses, as the closest parallel. Note that the whole epic panoply of Milton goes for nothing. But even if we leave our the grosser absurdities and keep to Jonah, the insensitiveness is crass - Jonah, a tale with as few even pretended historical attachments as Job, grotesque in incident and surely not without a distinct, though of course edifying, vein of typically Jewish humour. Then turn to John. Read the dialogues: that with the Samaritan woman at the well, or that which follows the healing of the man born blind. Look at its pictures: Jesus (if I may use the word) doodling with his finger in the dust; the unforgettable nv vuz (13:30). I have been reading poems, romances, vision-literature, legends, myths all my life. I know what they are like. I know that not one of them is like this. Of this text there are only two possible views. Either this is reportage - though it may no doubt contain errors - pretty close up to the facts; nearly as close as Boswell. Or else, some unknown writer in the second century, without known predecessors, or successors, suddenly anticipated the whole technique of modern, novelistic, realistic narrative. If it is untrue, it must be narrative of that kind. The reader who doesn't see this has simply not learned to read.

g:

As I have said before, the evidence that the early Christians risked death for claiming what they did is pretty poor, and in any case it's only evidence that they believed it, not that it was true.

Well, there are several different sources which say that particular apostles did die for being Christians. And it seems pretty obvious to me that they should have expected persecution given what had happened to Jesus, and what they were saying about the Sanhedren and Rome. If that's not good enough, I can't help you. Yes, the fact that they died for it only proves that they believed it, but since we're talking about the original claimed eyewitnesses here, evidence that they weren't lying places some important constraints on skeptical hypotheses.

It seems to me that if you look at Jesus's message, as reported by the gospels, rather little of it has anything to do with how "any person, no matter how bad, could conceivably be forgiven and be used by God".

Except for the bits about imitating God by forgiving one's enemies, dining with sinners and tax collectors, it's not the healthy who need a doctor but the sick, the prodigal son and several other parables, the Son of Man giving his life as a ransom for many etc. Seems like a major theme to me.

The gospels' conspicuous disagreement about what other witnesses were present is interesting. (The Synoptics say it was a couple of Maries, or in Luke's case some not-so-precisely specified women. John says it was Nicodemus. All seem to me to suggest that they've given a full list.

Seriously? None of the gospels say anything like "and there were no other witnesses". I don't see why you think that's implied; to me it would be highly unnatural. If I say "I went to Uncle Joe's funeral; Richards and Aunt Betty were there crying over the body" I don't think anyone would take this to imply "and no one else was there".

These are the sorts of "contradictions" that anyone can find in any set of documents if they squint hard enough. Think about this for a minute. If documents X and Y contradict any time X happens to mention an event or person that Y does not, or vice versa, then the only way to avoid that would be if X and Y both contained an identical outline of details with no differences. But that's not what you would expect real historical sources to look like, and I dare say you wouldn't find the Resurrection Accounts any more convincing if they did look that way.

The resurrection-appearance stories are generally not corroborated by multiple sources; each source has its own set of stories, and in the cases where they overlap generally nothing but the barest kernel of the story is consistent from one source to another.

The degree of difference between the Resurrection narratives has been greatly exaggerated. Suppose we take the following five sources: Matthew, Mark 16:1-8, Luke-Acts, John, and 1 Cor 15. Even if you only count statements which are found in 2 or 3 of these sources, here are some of the events so so supported:

-A set of people including Joseph of Arimathea, Mary Magdalene, and another Mary witnessed Jesus' burial;
-A set of women including the two Maries went to the tomb early on the third day (Sunday morning), when they saw that the stone had been rolled away;
-at least one angel spoke to them saying that Jesus was risen and that they were to report this to the male disciples;
-Mary Magdalene saw Jesus risen from the dead;
-the women reported to the male disciples;
-Peter went to inspect the tomb and saw the linens lying by themselves;
-Jesus appeared to Peter individually;
-Jesus appeared to the group of Twelve on Sunday evening (minus Judas of course, and according to John, Thomas wasn't included until later), and instructed them regarding their future role as disciples;
-Some of the disciples doubted, but Jesus dispelled doubt by inviting them to inspect his body and his wounds;
-Jesus also appeared at least once to disciples in the region of Galilee;
-Jesus eventually ascended to the Father.

What can I say? I believe these statements for roughly the same reason as I believe any other kind of testimony. If I am a fool for doing so, I think it comes down to the prior probability of Christian metaphysics rather than the manner of transmission of the reports being inherently unreliable.

Regarding the supposed implausibility of Christian metaphysics, this statement of yours seemed quite poetic and maybe encapsulates the primary issue:

But it's always possible that the Christians are right when they say that's too simplistic an expectation, and that it's more fitting for God to show his glory through obscurity, his riches through poverty, his strength through weakness (though that still looks to me like rationalization),

In other words, it all comes down to the Cross. The main point of my God and Evil post was that this is basically a core doctrine of Christianity. It's not an ad hoc add-on to Christianity, it is Christianity.

I suppose part of what usually happens when someone converts to Christianity is this. The idea of a Crucified Savior, which looks so absurd from a worldly perspective, suddenly seems to make sense at a deep level, and to connect to everything else we know about the universe, and about sacrificial love. I'm not saying that this by itself resolves the historical issue, but it makes a big difference to ones priors...

and in those (perhaps rather few) possible worlds where God has some sort of commitment to never doing anything that gets him clearly noticed, it wouldn't be that surprising for a messiah or incarnation to leave little evidence. But note that in those worlds, a whole lot of the Old Testament had better be fiction, because the OT's god clearly has no compunction at all about showing his power dramatically, at least some of the time.

That would follow, if you assume that God must either always reveal himself dramatically, or never reveal himself dramatically. That seems like a false dichotomy to me.

One more general comment and then I'm done. In the main article, I've mostly only discussed two kinds of evidence for Christianity: the Historical Argument and Fine-Tuning. I also mentioned modern-day miracles, but I didn't go into detail. I happen to believe that these are excellent arguments, but I don't want to give the impression that I don't think there are any others. You two have written several posts making cumulative cases throwing in every possible argument for Atheism you can think of, so I don't want to give the impression that I've done anything parallel to that. In particular I think that the Cosmological Argument, stuff related to Judaism, certain arguments from Religious Experiences, Ethics, Consciousness, and the human condition more generally, are all pertinent to the analysis. Whereas on the Atheistic side the only things that significantly move me are the Argument(s) from Evil and Occam's razor.

All right, that's it. Signing off.

98. lavalamp says:

OK, I'm trying, but I just can't leave some of what you said without a response, because I don't feel like you understood my arguments.

* M: One or all of: Multiverse/MWI/Everything that can exist, does
* R: Our universe is randomly the only one and just happens to support life
* G: God did it
...G is strictly less probable than R if you go by Kolmogorov and expand it to "[who] did [what]"; R expands only to "[what]". "You can't compress the laws of physics by adding in a god," would be another way of saying this. But of course this is an argument about priors...)

The odds of the cosmological constant being as small as it is randomly (as opposed to by some putative physical or supernatural mechanism that selected a specific value near 0) is about 10−120. You can't actually think that God is less probable than that.

By defining R to include the statement that the universe "just happens" to support life, all you're doing is moving improabability from the P(life | Naturalism) to P(R). It doesn't get rid of the huge number.

You are arguing that the prior for R is very low (and in case it wasn't clear, I agree). You're making this argument based on the space of possible universes (each set of values for the physical constants describing a different "possible universe"). Hence what I have been saying for a long time now: this is an argument about prior probability.

R might be very rare in the space of possible universes, but if we want to compare with G we need to ask how rare G is in the space of possible universes. (Here I use "universe" to mean something that encompasses all of reality.) I'm open to priors other than simplicity if you can come up with one, but to me it seems clear that G takes more bits to write down than R, therefore it's rarer than R.

Your argument that G is strictly less probable than R is specious; if that were a logically valid mode of inference you could use it against every kind of causal agent, not just God.

Not true. If you can compress an event's description by adding an agent in, then the agent-based explanation is, in fact, simpler. So no, it's not specious. Imagine trying to tell a story involving humans, describing their actions without reference to their desires or goals; it would be extremely verbose and confusing. I'm sorry to ask, but since we've gotten to the part where we accuse each other of bias (or possibly wisdom-teeth-extraction pain medication): did you think about this seriously for longer than 30 seconds before responding?

I suspect that further discussion would not result in our coming to agreement, for various systematic reasons that I believe have to do with your biases. (Doubtless you would want to say the same thing about me, but naturally I believe that I'm the one who is right!) Your a priori philosophical priors are too low and your cynicism about testimony is too high, so I can't make any progress here.

I gave you 40dB from history, that's nothing to sneeze at and possibly too much (I suspect similar historical evidence for a religion will be wrong more often than one in ten thousand times). I'm open to arguments that my preference of a simplicity prior is incorrect. I don't believe you can claim that I'm misapplying that prior.

I'm a bit disappointed to be accused of bias. I really do try extremely hard to be critical of every viewpoint evenly. You haven't provided arguments I haven't already taken into account, so I haven't yet needed anything so drastic as this: http://lesswrong.com/lw/ur/crisis_of_faith/

99. g says:

Just a few comments on Aron's latest. No response required from Aron, of course.

further discussion would not result in our coming to agreement, for [...] reasons that I believe have to do with your biases. (Doubtless you would want to say the same thing about me [...])

I would want to say part of the same thing about you: I think you're wrong, as you think I'm wrong; I expect your biases are part of why you are, as you expect mine are part of why I am; but I wouldn't, as you appear to be doing, give up on the possibility of agreement; I don't see any particular reason to think that your biases are incorrigible, that you're so firmly wedded to them that rational discussion is a waste of time. (Saying that about me is in fact just about the only way I know of to offend me, so congratulations.) I do think it's possible, though, that discussion of this particular sort -- which is inevitably low-bandwidth and not very well suited to the kind of tightly coupled back-and-forth that I think offers the best prospects of resolving disagreements about, say, how much one should trust any given bit of historical testimony -- may not be helpful.

And conversely, it does not at all look to me the way I would expect if it had all been made up

Who's claiming that it was "all made up"? No one here, so far as I can see.

Does a triangle have three sides because it has three angles [...]?

Yes, I agree that sometimes such questions have no answer. But consider why that question about triangles has no answer: it's because those two ways of picking out what things are triangles are logically equivalent: the distinction between them has (so to speak) no content at all, at least for a hypothetical perfect reasoner. It seems about as obvious to me as anything can be that this isn't true for "what is good" and "what God approves of" unless either the former is explicitly defined in terms of the latter, in which case Euthyphro, or "God" is just being used as a label for various things about the world, in which case atheism; and if there's some other way in which there can be no answer to the question "which is logically prior?" then I'm currently unable to see it.

Because most historical sources are reliable 90% of the time?

Forget the historical bit for a moment; most people are not reliable 90% of the time -- or, more precisely, they are not 90% reliable on nontrivial questions.

My phrase which you quote probably does a bad job communicating clearly what these heuristics are

In particular, it didn't communicate to me -- the fault is very likely more mine than yours -- that you were intending it to stand for a whole grab-bag of heuristics, rather than just the specific criterion it states explicitly. Sorry about that.

Of course I wasn't intending to suggest that the Grimm fairytales are just like the Gospels in literary style; they're quite different. And, having revisited a few Grimm tales, I find I'd forgotten just how stylized they are, so let me propose instead J K Rowling's "Harry Potter" books. For the avoidance of doubt, again, I am not saying that they are identical in style to the Gospels; I am just saying that the specific heuristic "does it contain sober factual details?" gives completely the wrong answer here -- as, I think, it does so frequently that it just isn't a very useful heuristic. Which was my point.

Some other things I wasn't saying: that stories containing miraculous or fantastic events are never true; that we should use no heuristic other than saying that such stories are less likely to be true; that there are no differences in literary style between Grimm and the gospels.

Well, there are several different sources which say that particular apostles did die for being Christians.

Sure, there are sources that say that. For instance, Aron Wall right here :-). What seem to be in short supply is sources that there's very good reason to believe. I'm not claiming that no early Christians were martyred; quite likely some were. What I am disputing is the line of reasoning that goes like this: Lots of early Christians were martyred; they could have escaped death if they'd admitted that their beliefs were false; therefore they truly believed those things; therefore those things are true. I think every step there, with one probable exception, is dubious: the traditions of martyrdom for people like Matthew, Peter, Philip, etc., don't seem very reliable; it is not at all obvious that recantation would have saved those who were martyred; I do agree that if someone was martyred and could clearly have escaped by recanting, it's good evidence (though not proof) that they were sincere; but being sincere is not the same thing as being right. To repeat something I've said several times: I do not think it terribly likely that all the key Christian doctrines were just made up; I think they are more like mistakes than outright fictions.

Except for the bits about imitating God by forgiving one's enemies, [...]

I agree that Jesus taught something rather like "any person, no matter how bad, could conceivably be forgiven and be used by God" -- though I don't think all the particular things you cite say quite that. I just don't think it looks like "the basic Christian message". (But perhaps I misunderstood that phrase; I took it to mean something like "the central claim of Christianity", but perhaps you only meant "one simple thing taught by Christianity".)

If documents X and Y contradict any time X happens to mention an event or person Y does not

I did not say that. Because it would be stupid. I repeat something I've said before: we will likely have more productive discussions if we all adopt the heuristic of assuming that if the other person seems to have said something absolutely daft then we've probably misunderstood. I apologize if I've failed to do this; it seems to me that you've failed to do it to me again and again and again.

I agree that some ideas along those lines are central to Christianity. I was not saying that the whole idea is an ad hoc add-on (though I think it's at least a bit plausible that it was an ad hoc add-on once, back in the very early days, as the early Christians tried to find a way of making sense of the death of their messiah). But I think that having that as a central idea makes Christianity less likely to be right a priori; I think some applications of the idea are probably convenient rationalizations; and I think that versions of the idea that say that God goes out of his way not to be noticeable make it very difficult for there to be good evidence for his existence. (And, if combined with two further propositions that I think you probably believe -- (1) that belief in God is very, very, very beneficial, and (2) that while really obvious evidence is hard to come by, there's subtler evidence, from things like cosmic fine-tuning, that can be very convincing to those who are clever enough to grasp it -- it seems to me to convict God of a sort of elitism that's quite hard to reconcile with Christianity. Isn't he supposed to reveal himself to the foolish and frustrate the wisdom of the wise, and not the other way around?)

That would follow, if you assume that God must either always [...] or never reveal himself dramatically. That seems like a false dichotomy to me.

The position I was attempting to address here is one that explains the shortage of evidence for God's existence by saying that he has some kind of policy of keeping himself hidden. A consistent policy of hiddenness is very hard to reconcile with the OT, taken at anything even slightly like face value; on the other hand, a policy of hiddenness now together with one of blatant showing-off of divine power in OT times seems awfully implausible and in need of explanation.

You two have written several posts making cumulative cases throwing in every possible argument for atheism you can think of

I suppose maybe lavalamp sort-of has, having posted that apologia he wrote. I assure you that I have not. I haven't by any means thrown in every argument for atheism that I can think of, and I have no idea what would make you think I have.

100. lavalamp says:
You two have written several posts making cumulative cases throwing in every possible argument for atheism you can think of

I suppose maybe lavalamp sort-of has, having posted that apologia he wrote. I assure you that I have not. I haven't by any means thrown in every argument for atheism that I can think of, and I have no idea what would make you think I have.

I think I posted that more for background information about me than anything, and anyway I don't think we've talked about half the stuff in it. I am interested in mapping our disagreement-- I'm not trying to "win" by throwing stuff at the wall until something sticks. This comments-on-blog-post format is a particularly poor one for making a map of an argument, unfortunately. (And if Aron has big guns left to get out, by all means I hope he'll get them out!)

(Saying that about me is in fact just about the only way I know of to offend me, so congratulations.)

In case it wasn't clear from my attempted humor in my last comment: same here.

I repeat something I've said before: we will likely have more productive discussions if we all adopt the heuristic of assuming that if the other person seems to have said something absolutely daft then we've probably misunderstood.

One thing I've noticed about really smart people (and my guess is that everyone in this conversation fits that category) is that they're used to dealing with (what are to them) stupid people; and when they do encounter someone of approximately equal IQ, they find it difficult to seriously think about their arguments due to force of habit. I try really hard not to do this; one thing I specifically do in written comments is to take 30 seconds and try to imagine the response of my interlocutor. I think if Aron did that, it would make the conversation smoother; I don't think half the things I'm saying are arguments he couldn't come up with if he were trying. (Apologies if this is presumptuous of me.)

101. Aron Wall says:

Dear g and lavalamp,

I said I wouldn't continue this particular argument any more. However, since you both seem to have taken offense at me in your closing statements, I feel that I must write something of an apology for the sake of avoiding ill-feeling.

g:

I don't see any particular reason to think that your biases are incorrigible, that you're so firmly wedded to them that rational discussion is a waste of time. (Saying that about me is in fact just about the only way I know of to offend me, so congratulations.) I do think it's possible, though, that discussion of this particular sort -- which is inevitably low-bandwidth and not very well suited to the kind of tightly coupled back-and-forth that I think offers the best prospects of resolving disagreements about, say, how much one should trust any given bit of historical testimony -- may not be helpful.

lavalamp:

In case it wasn't clear from my attempted humor in my last comment: same here.

I do most humbly apologize for giving the impression which offended you two. But I didn't mean it in quite the same way you think I did. I meant that more of this particular discussion (in this medium, centered around these particular topics) is not likely to bring us into agreement, not there is no possible kind of rational consideration which could change your mind. In other words, I meant something a lot closer to what g says in his last sentence. I regard both of you as intelligent people who are trying (in my view unsuccessfully) to weigh the evidence fairly. If I really thought that rational conversation with you was a waste of time, there would not be 100 comments on this thread. When people have been arguing for that long, usually it's better to disengage for a while, and then come at things from a different angle much later.

Obviously, any accusation of bias is to some degree insulting, and I understand that. But I can't change my honest opinion simply for the sake of avoiding giving insult, although I could certainly have avoided expressing my opinion. I tried to lighten the tone by anticipating a tu quoque, but in the long run we just have to deal with the fact that bias on the subject of religion is extremely common, and that we each think we're in the right and the other person is in the wrong. Anyway, rationalists seem to think that bias is as about as ubiquitous as (Christians think) "original sin" is, so it shouldn't be all that shocking of an idea that you (or I) might be guilty of it.

g:

I haven't by any means thrown in every argument for atheism that I can think of, and I have no idea what would make you think I have.

I shouldn't have said that you two were presenting every argument you could think of. This was not something I could have any way of knowing, and I regret having said it. However, please understand that my point was not to criticize you (except for a tendency to defocus the conversation, how could there be anything wrong with presenting extra arguments for ones position?), but rather to compare to what I was trying to do in this thread.

However, if you're curious which comment of yours led me to make this hyperbolic claim, it was the comment a while ago where you mentioned things such as failed prophecy, and the quality of poetry in the Bible. You gave only 10dB to the Historical Argument, but significantly more to several different miscellaneous arguments against Christianity.

lavalamp:

I try really hard not to do this; one thing I specifically do in written comments is to take 30 seconds and try to imagine the response of my interlocutor. I think if Aron did that, it would make the conversation smoother; I don't think half the things I'm saying are arguments he couldn't come up with if he were trying. (Apologies if this is presumptuous of me.)

It is presumptuous of you, or at least a false assumption. On average, I think for considerably more than 30 seconds before responding to any comment of yours. The fact that you happen not to be persuaded by my counterarguments, does not imply that I don't sincerely believe them. In fact, I still stand by my comments even after having read your reply. And sometimes I have anticipated your counter-reply, but I didn't agree with it so I posted my views anyway. :-)

The continued differences between us are not based on me failing to take 30 seconds, they are caused by some radical differences in how we think about epistemology. I don't think we've exhausted the subject of fine-tuning and Komologorov complexity, but I'm too tired to continue this argument right now. Maybe later, when I write some articles on the fine-tuning argument. (Some of what I think about this came up in my earlier conversation with amkglor.)

g quoting me:

If documents X and Y contradict any time X happens to mention an event or person Y does not

I did not say that. Because it would be stupid. I repeat something I've said before: we will likely have more productive discussions if we all adopt the heuristic of assuming that if the other person seems to have said something absolutely daft then we've probably misunderstood. I apologize if I've failed to do this; it seems to me that you've failed to do it to me again and again and again.

Again I apologize for giving the wrong impression. But I think you are also misunderstanding the nature of the argument style I'm using. I wasn't trying to "restate" your position. I was trying for a reductio ad absurdum. In a reductio, you try to get the person to give up a wrong claim by showing that when the claim is consistently applied, it gives absurd results. This should not be regarded as an inherently offensive mode of reasoning.

In other words, I wasn't accusing you of saying the stupid statment in question, expressed in that stupid sounding form. We both agree that the statement expressed in this form is stupid. The statement you did make is that the Gospels have a "conspicuous disagreement" concerning which witnesses are present. So I said to myself, suppose it did count as a "disagreement" if the Gospels don't have the exact same list of witnesses to the burial. And suppose we apply this across the board, not just to the burial but to every other documentary feature. What would it take for two historical documents not to have this sort of "disagreement"? And that's how I came to this general absurd principle, which I don't accuse you of believing. Instead I was hoping that the fact that you don't believe this general principle would cause you to reconsider your statements about the burial. If for some reason you think the burial is a special case to which different principles apply, then that would mean you don't accept my reductio argument. But a failed argument is still no reason to be offended.

And now just two clarifications of my position:

(And, if combined with two further propositions that I think you probably believe -- (1) that belief in God is very, very, very beneficial, and (2) that while really obvious evidence is hard to come by, there's subtler evidence, from things like cosmic fine-tuning, that can be very convincing to those who are clever enough to grasp it -- it seems to me to convict God of a sort of elitism that's quite hard to reconcile with Christianity. Isn't he supposed to reveal himself to the foolish and frustrate the wisdom of the wise, and not the other way around?)

Well, there are quite a few people with low intellectual pretensions who are Christians, and many of them think it is totally obvious that God exists. They have an extremely strong intuitive sense that he does, when they watch a beautiful sunset over the mountains, or think about morality, or read about Jesus in the Bible. In light of this, I don't subscribe to (2).

I don't at all blame you for thinking I did, in light of some of the comments I've said above. But I didn't express my full position very well. I think that God does reveal himself intuitively to ordinary people, but for various reasons, rationalists don't consider this sort of thing to be good evidence. One way to rebut this would be to argue that these intuitions really do correspond to actual evidence; another way is to present more sophisticated arguments, of a sort that rationalists are more likely to find convincing.

If a particular sort of rationalist rejects religious "intuitions" as unreliable, and says that clever arguments are better, and prides himself on carefully examining clever arguments without bias, then I'm not sure he can complain if God chooses to reveal himself to him by the method of these clever arguments, rather than through the intuitions he rejects.

The traditions of martyrdom for people like Matthew, Peter, Philip, etc., don't seem very reliable. I do agree that if someone was martyred and could clearly have escaped by recanting, it's good evidence (though not proof) that they were sincere; but being sincere is not the same thing as being right. To repeat something I've said several times: I do not think it terribly likely that all the key Christian doctrines were just made up; I think they are more like mistakes than outright fictions.

Actually, I agree that many of the martyrdom traditions probably aren't reliable (though I doubt they would have sprung up if the apostles had all died in bed.) I think we can be quite confident however that Peter, Paul, James the apostle, and James the brother of Christ were all really martyred. Though for Peter and Paul, some of the later details of their execution may be questionable.

And if I weren't a Christian, I think I would regard lies as significantly more plausible than mistakes. So that's an interesting difference between us, which may account for some of our talking past each other.

Peace be to both of you.

102. lavalamp says:

It is presumptuous of you, or at least a false assumption. On average, I think for considerably more than 30 seconds before responding to any comment of yours. The fact that you happen not to be persuaded by my counterarguments, does not imply that I don't sincerely believe them. In fact, I still stand by my comments even after having read your reply. And sometimes I have anticipated your counter-reply, but I didn't agree with it so I posted my views anyway. :-)

OK, fair enough, my apologies. I find myself very confused because sometimes we seem so close to agreeing and other times I'm not even sure we're having the same conversation.

And if I weren't a Christian, I think I would regard lies as significantly more plausible than mistakes.

I just saw this on reddit, maybe it will cause you to increase your P(mistakes) a tiny bit: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2201941

:)

103. g says:

Aron, for the avoidance of doubt I'm not in the least offended by the suggestion that I have biases, and if I gave the impression that that was offensive then I screwed up.

You gave only 10dB to the Historical Argument, but significantly more to several miscellaneous arguments against Christianity.

Well, as you may have noticed I think the Biblical documents' evidence for the resurrection is really weak. I didn't list any other particular "miscellaneous arguments against Christianity"; I listed some categories of evidence/argument and said what evidence either way I thought each category provided. I wasn't trying to make any kind of argument against Christianity, I was summarizing my position roughly as you had summarized yours earlier, because I thought it might be useful for mutual understanding. If I'd realised it would be interpreted as "here's my cumulative argument against Christianity, which I hope will be persuasive" then I'd have made it clearer that it was nothing of the kind, but that thought never crossed my mind. I don't think I presented one single argument there; I certainly wasn't trying to.

And, yeah, I mentioned the quality of poetry in the Bible, but (for the avoidance of doubt) that isn't because I think there's much of an argument either way to be made on that basis. I listed a few things that come under the heading of "quality of the Bible", that's all. Why mention it at all? I've heard various rather hyperbolic claims, mostly from Christians, about the literary merit of the Bible; if they were correct, they might be evidence for supernatural influence on its composition. Again, if the thought "Aron will take this as meaning that I think the quality of the poetry in the Bible is a strong argument against Christianity" had occurred to me, I'd have not mentioned it or added a disclaimer; but that thought never occurred to me.

I wasn't trying to "restate" your position. I was trying for a reductio ad absurdum.

OK. So, for the sake of clarity: I do not believe that whenever one document mentions something another doesn't they should be considered contradictory. Nor, so far as I can tell, do I believe anything that implies that. Nor, so far as I can tell, have I said anything that implies it. For a reductio to work, the absurdity you arrive at does actually have to be a consequence of what you're trying to refute, not some uncharitable extrapolation.

(I'm not offended by what seems to me your repeated assumption that it's more likely I believe something really stupid than that you've misunderstood me or extrapolated my beliefs wrongly. Just annoyed.)

In light of this, I don't subscribe to (2).

I wonder whether one of us is misunderstanding the other. Obviously you don't believe -- because that, too, would be really stupid, and you are not really stupid -- that people who aren't very clever never think they have "really obvious evidence" for theism generally or Christianity particularly. Of course lots and lots of people think that. (And lots of people think they have really obvious evidence the other way.) But the impression I had is that you (like the "rationalists" in your next paragraph) "don't consider this sort of thing to be good evidence". Was I wrong? Do you consider that when someone sees a beautiful sunset and thinks "this proves that God exists" they are correct, and that seeing a sunset really is "really obvious evidence" for God's existence?

Er, I notice that I'm asking questions even though you very reasonably want to get out of this discussion. Please feel free not to answer them.