- Review of Minas Papageorgiou’s Jesus Mythicism: An Introduction by Tormod Tobiassen, Norwegian theologian at Bergen University College. The original in Norwegian is at Tormod Tobiassens weblogg; a Google translation at Current Books About Jesus Mythicism.
- I’m not sure what the current status of the Kickstarter for a debate between Bart Ehrman and Robert M. Price is. I have been slow to post on this in part because I did not like the idea of paying Ehrman $5000 (even if he does give the money to charities — I’d rather he be assured by receipts others have already donated to charities of their own or better still just speak for no other motive than the public interest). My other part reason was that from what I have read by Ehrman on the subject and from what I have seen of his manner in some videos when addressing the topic, I really can’t see him making any genuine effort to bring due diligence or seriousness to the debate. I’d rather see a debate with a scholar who undertakes a more professional approach vis a vis the public interest. Others may disagree, however.
- Tim Hendrix published on Scribd a review of Richard Carrier’s earlier book, Proving History, in which he questions Carrier’s use of Bayes’ theorem for historical argument. (I understand that Tim’s research field is Bayesian methods for machine learning.)
- Jerrel Arkes, a 30 year old atheist from the Netherlands, has opened a new site, www.science-vs-religion.com [link no longer works: Neil Godfrey, 24th July, 2019], on which people can vote (with social shares) for Science or Religion. The intent (as I understand it) is to start conversation through social media.
The following two tabs change content below.
Neil Godfrey
Neil is the author of this post. To read more about Neil, see our About page.
Latest posts by Neil Godfrey (see all)
- Jesus Mythicism and Historical Knowledge, Part 2: Certainty and Uncertainty in History - 2024-11-18 01:15:24 GMT+0000
- Jesus Mythicism and Historical Knowledge, Part 1: Historical Facts and Probability - 2024-11-16 01:05:37 GMT+0000
- Palestinians, written out of their rights to the land – compared with a new history - 2024-10-15 20:05:41 GMT+0000
If you enjoyed this post, please consider donating to Vridar. Thanks!
“My other part reason was that from what I have read by Ehrman on the subject and from what I have seen of his manner in some videos when addressing the topic, I really can’t see him making any genuine effort to bring due diligence or seriousness to the debate.”
Which is sad. The fundamentalists and others in the historicist choir will say Bart Ehrman won the debate, but for anyone sitting on the fence, or in either the mythicist or agnostic camp, or haven’t formed an opinion on the matter will see Robert M. Price expose him as an idiot.
Agree with your assessment of Ehrman and thus lack of excitement about the proposed debate.
What I want see is a debate (or better extended correspondence) between Mark Goodacre and Richard Carrier. Goodacre is clearly one of the most rational and open-minded mainstream scholars around. He would provide serious counter arguments to mythicism, not straw-mans. His NT pod has been a joy for me to listen too and his previous interactions with Carrier were fascinating. I think Carrier deserves a response from a major scholar within the academy, and Goodacre would be a fine choice.
J
My opinion of Robert M Price has taken a turn for the worse in recent years; he seems unwilling to consult scholars from other fields (e.g. historical linguistics) when he doesn’t know whether a claim in that field is worth taking seriously or not, and just handwaves it away – “I am no historical linguist, so clearly I cannot know”. This is an excuse he has hid behind, so as not to have to say outright that e.g. DM Murdock is presenting shoddy linguistics, but instead can present her as a serious scholar.
I think that’s a fair critique, but not a damning one. It fits in with something Carrier once said about Price: that Price is good at coming up with hypotheses about what might have happened, but much less good at determining what is most likely to have happened. I’m still a fan of Price’s, but my approach to his ideas is that they are always interesting and always possible, but I have to judge for myself how likely I think they are.
Robert M. Price should be appreciated for what he’s good at. As far as i understand he respects Murdocks research when it comes to topics he is familiar with, and why should he ‘denounce’ her on a topic he’s not an expert in? What’s the relevance of his opinion on that anyway given his non-expertise on historical linguistics, which i’m sure he’d freely and happily admit to?
Full-disclosure; Long time listener of Price’s bible geek podcast.
Big fan of the the bible geek price…but also agree with the Greg’s comment. He has an amazing encyclopedic and open mind (which he generously offers to the lay public). Incredible teaching skills. But I don’t find he is the most convincing of the mythicists (and on other questions). Does not put the case in the most persuasive way in my humble opinion. But I love the geek…
J
The Critique of Richard Carrier’s use of Bayes’ is pretty damning to my my eyes. My math isn’t good enough to evaluate properly in full, but the logic of the argument seems pretty sound. I hope Carrier responds.
What did you see as particularly significant in his critique?
The author suggests that Carrier’s definition of probability and his additions and subcategories of it(Hypothetical frequencies, True frequencies etc.) are poorly defined and applied in such a way that is either useless or antithetical to Bayesian reasoning. Carrier never represents himself as being anywhere other then the mainstream on Bayesian math, and if the author is correct that he isn’t, then we probably safely throw Carrier’s Bayes based historical method out the window. As it stands, I neither know enough math nor know what sort of weird, esoteric arguments are going on in Mathland to be sure.
I hope this doesn’t come off as an argument from authority. To me, it would be very much like if someone came in here and stated: “As all Biblical scholars agree, Luke was written first. Therefor…”. I would cease listening at that point. Either this person has not read what biblical scholars have to say, or so misunderstood their arguments that he or she might as well not have. Any additional arguments placed on top of such a position would almost certainly be wrong(even if in principle, such arguments could still be correct) Hendrix Places Carrier in a similar position Re: Bayes.
Lastly, let me say that I don’t think this affects Carrier’s arguments on the historicity of Jesus. I think those arguments are persuasive and can stand alone.
Carrier is not the first or the only historian to apply Bayesian analysis in historical inquiry. I’d be interested in know if critics of Carrier can point to any comparable criticisms of other historians using Bayes. I don’t see the problems, myself. Bayesian analysis is what we all do all the time anyway when we are serious about tracking down unknowns and probabilities. The maths is not the point and is really an extra — as Carrier himself points out. It’s the pattern and method of reasoning that is the crux of the argument and that I can’t fault. http://vridar.org/category/book-reviews-notes/carrier-proving-history/
Hi Neil,
Thanks for putting a link to my review on your blog, I have been a long time reader and I am really exited to see myself on the front page. According to scribd no less than 160 people have opened the document, considering this is a math article I think that shows a lot about how dedicated your readers are!
Regarding the use of Bayes and history, like I wrote in the review, it is an interesting idea to consider how Bayes theorem can be applied to historical questions, and I think Bayes theorem has it’s use outside textbook arguments. Concerning critism of other uses of Bayes theorem, I must admit I am not aware of how others are using Bayes theorem for history, and I thought it was a novel idea in “Proving history” since I didn’t find any other cited work. If you have a link I would be happy to read it.
I think it is important to be carefull what we mean by “using Bayes theorem for history”. Suppose we try to evaluate a hypothesis of any sort in light of a piece of evidence, I think it is fairly uncontroversial to say people intuitively understands the relevance of
(1) how likely the hypothesis is in the absence of the evidence (from physically impossible to highly probable) and
(2) how likely we are to observe the given evidence if the hypothesis is true in relation to how likely we are to observe the evidence if the hypothesis is not true.
Notice (1) and (2) map closely to Bayes theorem qualitatively, just as we would expect.
However as I read “Proving history” (I wrote the review before receiving “On the historicity of Jesus”) Dr. Carrier propose that we should determine historical questions by estimating specific probabilities and then plug these probabilities into Bayes theorem and use the computed probability to judge if a proposition is true or not. I think this view is implied both in the title and the text, and it is what Dr. Carrier does in “On the historicity of Jesus”.
I think there are serious *practical* difficulties with this program that I would not know how to solve myself, both relating to the estimation of probabilities and formation of hypothesis and probability spaces. I stress I do not consider these to be academic concerns but problems with practical implications. I think a major weakness of the book is that it does not give examples of how these questions should be settled in practice, but relies on general arguments which I don’t think addresses these questions. I was not aware of the arguments by Ian untill this week, however I think it is interesting Ian and I as best as I can tell considers exactly the same problems regarding the use of Bayes theorem for history.
I try to give examples of why I think this discussion is incomplete in relation to the criteria of embarassment and the library-example, however I actually removed a larger section from the review discussing these issues because I did not know how Dr. Carrier was going to address them in OHOJ and the discussion would therefore be speculative. I have only read parts of OHOJ, but these problems are certainly there, along with other things. Notice Dr. Carriers response to Ian in the linked article either misses Ians points or does not explain how these problems should be solved in practice, but again relies on general arguments from proving history Ian is well aware of.
I am hoping to explain these issues in detail later, however one of the basic issue with the use of Bayes theorem is very simple: Forget about Bayes and suppose I show you a large bowl with Blue (B) and Red (R) M&M’s and ask you to guess how many red M&M’s there are. So you look at the bowl and guess there are 1000 red M&M’s. Now I tell you that you are doing it wrong, what you should really do is to notice that if T is the total number of M&M’s you can compute R by the formula:
R = T – B
and I then discuss why this formula is mathematically true at some length. What is being missed here is that you still have to estimate numbers, but instead of just estimating R, you now have to estimate T and B. Moreover, if you make an error of the same magnitude in your estimates, the error in T-B will be larger than the error in R. *the mathematical truth of the formula is irrelevant*. The same problem exists when applying Bayes theorem to history, which Ian and I both point out. Dr. Carriers response is as best as I can tell just affirming the truth of the equation without addressing the actual problem.
This is not to say this is the only or most serious problem, but it should be addressed. I think the discussion of hypothetical frequencies in chapter 6 and the proof in Chapter (iirc) 3 is trying to address this issue and I discuss both sections in my review.
Finally I completely agree with the posters who point this does not affect the historical case against Jesus and I am actually greatly enjoying this aspect of OTHOJ. I hope *someone* will do a review of the historical argument at some point.
Cheers,
/t
Regarding using Bayes’ theorem to explain historical reasoning there is a book by
Aviezer Tucker: “Our Knowledge of the Past A Philosophy of Historiography”.
Hi,
Thanks for the reference. I haven’t had a chance to read the relevant chapters, however skimming them it seems the book argues that Bayes can be used /in principle/ (this follows by simply accepting the standard Bayesian interpretation of probabilities as beliefs, see my review), but does not offer precise suggestions for estimating probabilities, and I believe the author propose Bayes theorem more as a way to structure an argument than the computation of probabilities. Notice neither Ian or I has any objections to this use, but see problems with the later.
As I said, I haven’t read the text closely, but some of the equations are odd. For instance the equation on p.112 is wrong as can be seen by setting n=1.
Cheers
/t
Hi Tim,
I haven’t looked into the math so you might as well be right. I want to read your review (and Ian’s) carefully before deciding. As a general thought though I think it’s good that somebody is actually try to cast historical reasoning in rigorous ground. Even if Carrier’s approach is not perfect it’s a first step. Any field that uses the word “probably” should go through the trouble of understanding how probabilities work.
But as I said I haven’t actually looked into this carefully so I’ll shut up now.
Real Historians Do Bayes!
Hi Nikos,
I think historical reasoning by large is already very rigorous and carried out by very smart and careful people! I think the most important issue is if using Bayes theorem (in the quantitative manner, see my other posts) is actually an improvement. A wrong (imprecise) use of Bayes theorem will *most likely* lead to overconfident assertions, but with an extra layer of mathematical complications — I don’t think this is what Jesus studies need.
Part of understanding what probabilities is is also understanding it is only one of several theories (and semantics) of uncertainty, as well as some complications in applying probability theory in practice. I think PH focuses to little on these aspects.
Cheers,
/t
1) I think the main and most important thrust of the whole application of Bayes thing is that you are complaining that in practice, assigning a probability to X where X is “this evidence would exist given Y” is really hard/arbitrary/unclear.
But actually, the whole thing were historians already do that, but just use vague terms instead of math, is worse. I have read sentences that go “X is [probably/likely/more likely/indicated by] because Y” in which those words were used to mean every possible probability from 20% to 99.9999%, because historians don’t actually have the vocabulary to distinguish between those things at all. So any assigned probability number, however arbitrary or hard to make, is an improvement on first principles because it means that you have to at least be clear about what alleged arbitrary probability you are making. And if you look at OTHJ, you can see how that works. It doesn’t matter if the probability (or range) of 4/5ths is arbitrary or not, it is way the fuck more clear than the standard practice amongst at least biblical historians to just replace all numbers in carrier’s book with “probably” thus depriving the term of any meaning.
The assigning of relatively arbitrary numbers is a huge step up, because it gives us a reasonable point to actually start talking to each other about the usefulness of any given evidence for the conclusion. Some people may disagree whether something is 2/3s or 4/5ths, but since the current terminology is just people disagreeing whether things are 0-.80 or .20-1, that is a huge step up.
2) You also say that R = T-B is “bad” because by separating it out, you introduce the error range multiple times and therefore might be less accurate.
But the entire point is that by forcing people to actually separate out the formula you can actually get them to make statements with a lower error rate. The current error magnitude of “Jesus existed” is that some people say he probably didn’t, which means something between 80%-0%, and some people say he probably did, which means something between 20% and 100%. Even if you fix the problems that people are using the word probably wrong, and even if you make them use actual numbers, the error magnitude is 100% of the possible range.
A statement that individually analyzing the pieces of evidence and then computing the likelihood based on that might cause you to introduce the error range multiple times is meaningless. The error range is already 100% of the probability space when you ask people to estimate the conclusion. You can’t possibly get worse. So we should damn well look and see whether separating out our evidence allows us narrow the range.
Hi Kaelik,
Re 1, actually I am not sure I agree that replacing a word like probable with an (somewhat) arbitrary probability is an improvement. But I think we need to discuss these problems in a specific context to arrive at a convincing argument. If you look at Dr. Carrier’s treatment of the criteria of embarrassment and my treatment, how would you determine who are the more correct and what numbers we should use? Notice we arrive at qualitatively different answers.
Re 2, you wrote: “You also say that R = T-B is “bad” because by separating it out, you introduce the error range multiple times and therefore might be less accurate.
But the entire point is that by forcing people to actually separate out the formula you can actually get them to make statements with a lower error rate”
I am not sure I follow this argument. If we agree the formula tends to compound the errors we are making, how will it lead us to make statements with a lower error rate?
Regarding the following: “A statement that individually analyzing the pieces of evidence and then computing the likelihood based on that might cause you to introduce the error range multiple times is meaningless. The error range is already 100% of the probability space when you ask people to estimate the conclusion. You can’t possibly get worse.”
Leaving aside the issue how we determine what the error range currently is, isn’t the real issue how we know we are doing better by using Bayes theorem? I think this discussion suffer from not having a particular context. I hope to review OHOJ later and hopefully it will be more clear what sort of problems crop up.
Cheers,
/t
grung0r,
While I agree that Hendrix offers a good critique of the section of the book dealing with hypothetical frequencies, true frequencies, etc, I’m not sure it is that consequential. In fact I wondered what that section was even doing in the book. Carrier may have offered some reason for it to be there but to me it seemed out of place, perhaps better suited for a philosophy journal (though I doubt it would be accepted there). It was attempting to reconcile different philosophical views of probability (frequentist vs Bayesian), a side issue at best. That doesn’t really affect whether or not Bayes’ Theorem can be used in historical inquiry.
More consequential was where Hendrix stated in his conclusion:
He demonstrated this shortcoming with his own analysis of the Criterion of Embarrassment. I think this is a general problem and even carries over into Historicity. Often Carrier will analyze some document or set of facts, draw his conclusions about them and then (I think) use those conclusions as established when calculating his probabilities.
I say “I think” because he is general pretty vague about it and I often wasn’t sure if we were supposed to be taking his preliminary conclusions as established or not. Like are we supposed to taken it as a given (by Carrier) that there was a Jewish Christian group preaching Jesus had lived a hundred years earlier? Or, were we supposed to factor in that that interpretation of the evidence could be wrong. Ditto with claims about whether early Christians would use a more specific title for a brother in the flesh. Certainly those are things that could be disputed, or unable to reach a firm conclusion.
Whenever one of those points come up where there could be more than one answer the model should reflect that. Generally Carrier didn’t do that. I suspect it is because, as Hendrix states this will lead to very complicated models that are extremely hard to provide defensible numbers for.
If there is one solid critique of the Bayesian approach I think it is that. Yes, in principle, reasoning about history can be modeled with Bayes’ , but in practice the models will quickly become too cumbersome.
I never got the impression Carrier concluded the various propositions in the train of reasoning were “settled”. It was all about probabilities, not absolutes. Again, the maths are actually superfluous to Bayes’ reasoning really.
The reason I brought up the hypothetical frequency issue was that Hendrix stated:
“What is a probability in Proving History? To the best of my knowledge, probability is being equated with hypothetical frequencies, however this suggestion is definitely non-Bayesian…”
While I agree that the philosophical arguments about probability do not affect whether Bayes’ Theorem can be used as a historical method, it is absolutely central to Carrier’s application of it if he really is using “Hypothetical frequencies” as a synonym for “probability”. What the hell is a hypothetical frequency in terms of a singular historical event? Are we using the many worlds hypothesis(and if we are, wouldn’t the answer always be a probability of 1 or 0)? I would even suggest that if Carrier is doing this, it could explain the various vagaries and inconsistencies you note in his calculations. Hypothetical frequencies in terms of historical events sound an awful lot like “hypothetical documents” or “magical boxes in which we can stuff whatever we want into and get whatever we want out of”
Perhaps I’m misunderstanding what’s being gotten at though. I I cannot over-emphasize how much I suck at math.
You may have already seen this; Jerry Coyne strikes again.
https://whyevolutionistrue.wordpress.com/2015/07/23/readers-beefs-im-a-new-atheist-embedded-in-the-political-right/
Thanks. I hadn’t seen that. Pity Jerry didn’t tell me about it. Have I ever denied that terrorists claim religious motives? Remind me if I have, but I do know I have posted on why and how they claim religious motives — and the relative force of these motives in the larger context of what scholarly research has demonstrated (based on other motives the terrorist ALSO espouse.) A pity I have seen this only after 100 responses have already been made there so how will any reply of mine be seen?
You’ll notice this one.
“What “scholarship” that people like Godfrey and Robert Pape have mentioned or produced has completely ignored what the terrorists say about their own motivations”
This suggests to me he hasn’t actually read Pape or any of the scholars you’ve cited. Either that or he is willfully distorting matters.
What’s comical is that in one of the articles he cites to refute Pape, we can read:
“He (Pape) is right that Islam doesn’t cause suicide terrorism.”
Take a bow, Jerry.
Also this gem:
“I would maintain that this “scholarship” we ignore (and, in fact, I’ve read it) is tendentious and ideologically motivated”
No evidence for this claim rom Coyne, of course. It’s also funny considering that he accuses others of smearing.
I just dashed off this in reply to Jerry’s post. He doesn’t normally let any comments of mine through so I was surprised to see him post on this one.
Hi Jerry, My attention has only now been drawn to your post. I do hope you will allow me a reply.
First, I have never said that political motivations are the only ones for terrorists. Quite the contrary. See my various posts at http://vridar.org/category/terrorism-politics-society/ — I include research identifying a range of motivations but the one constant is political.
Second, I have never ignored what the terrorists themselves say, but I do think you seem to be selective in what you quote from them. Perhaps you are only quoting what you hear in mainstream media but even so I find it hard not to understand how you can not be aware of their oft-stated geo-political aims, as well as their religious hopes.
Third, I have cited more than one specialist. Pape is only one. Others I have addressed from the fields of anthropology, political science, sociology and psychology include Havez, Jackson, Esposito, Atran, Rahim, Hassan, Burke, Hage, Asad, and others. In fact my Pape references are dated. Terrorism has entered a new phase since then. But the political motivations are just as loudly shouted by the terrorists themselves as have their religious affiliations.
What concerns me most is that by “blaming Islam” for terrorism we are in fact essentializing Islam — imputing to it a mythical essence that is quite out of place. Islam does not exist except as a range of practices and beliefs by communities. Each community believes it has the only true faith, but the researcher steps above all that and understands the sociological construct that make up all faiths.
We bypass a critical and informed understanding of what we are up against when we essentialize Islam and blame the religion per se. Of course certain beliefs are evil, and many people are evil, and it is those we need to target with smarts, not populist ignorance.
You are a great scientist and I love your work in that area. But I beg you to inform yourself of what your scholarly peers say about topics outside your specialist area. And if you find a debate, then do justice to both sides before making up your mind. I have had to change my own views on some major political and religious issues myself. I used to have very similar opinions as yours. That was before I began to read the research and get behind the mass media images.
As for Islamophobia, I have explained exactly what I mean in http://vridar.org/2013/05/02/islamophobia-origin-and-meaning-of-the-word/
……
In another comment a Marella accused me of posting “vicious” responses to those who disagreed with my politics. I asked her for evidence of that. But again my comment got whisked away into ether…. so far.
Do I really sound “vicious”?
Vicious? Ha.
It seems only fair to add to that, as I did suggested in the above response, that Pape has certainly looked into the statements of suicide terrorists. In Dying to Win and Cutting the Fuse this can be seen on numerous pages. Some of the other scholars you referenced have also done this. In addition they’ve also interviewed several terrorists and , contrary to Coyne’s claims, looked at the role of religion in the path to terrorism. Their conclusions are based on research and evidence and are not ideologically driven as Coyne baselessly declares.
One final point on the ideologically driven research accusation: Coyne previously endorsed Nicolai Sennels as an expert on Muslims and jihad. For Coyne to accuse others of being ideologically driven after backing Sennels is laughable.
https://whyevolutionistrue.wordpress.com/2015/01/23/the-unctuous-and-dangerous-karen-armstrong/
Sennels, by any stretch of the imagination, is a stonewall anti-Muslim idealogue. Aside from the fact that Sennels is involved with all kinds of far right causes such as PEGIDA, his ‘research’ is a joke. Sennels claims he did research on Muslim youths, yet he did not keep statistics of any kind. Sennels essentially uses pseudo scientific rhetoric to call for Muslims to be removed from Europe.
http://www.newenglishreview.org/custpage.cfm/frm/63122/sec_id/63122
I hadn’t heard of Sennels until now, but checked the “New English Review” website which struck me personally as varied in content and often interesting. How “pseudo” or “rhetorical” are statements based on psychological assessments, historical facts about cultures, and opinion polls on terrorism taken among young Muslims in Europe, is a matter for judgement. If people find some Muslim settlements and specific customs difficult to integrate into western society, they are entitled to suggest that Muslims return to the House of Peace (aka Submission) in accord with a traditional practice that, while traders and missionaries are desirable, permanent settlement, without exclusive sharia autonomy, inside the infidel House of Conflict is improper.
“O believers, take not Christians or Jews as friends/allies.”
Hi, I am the commenter known on WEIT as “Marella”. I have posted an apology on WEIT for using the word “vicious”, it was a bit over the top, I’m sorry.
Thanks, Marella. Very much appreciated.
I know my posts on certain issues are very hard to stomach for many readers. I hope my recent post about understanding evil helps explain where I am attempting to come from.
I really do prefer in depth discussion and mutual understanding to hostility and if/when I do use inflammatory language or innuendo I’d prefer to be pulled up for it — as I sometimes am — than for it to pull the discussion downwards.
Jerry has denied me an opportunity to reply to either him or any of the comments on his blog. Along with his comments he posted my email without replying to me or informing me that he was using it this way. He knows he is welcome to comment here if he feels I have misrepresented him but has not done so.
I really had hoped for a professional and cordial response, to be able to disagree without being disagreeable. But that appears not to be Jerry’s style when he is speaking outside his area of expertise.
If that was your hope you haven’t been reading much of him or his commenters, perhaps lately, but more likely for at least several years. Coyne is touchier with dissent than McGrath and has a much larger ego. I’d put him in as a replacement for Dennett in the Four Horseman and call what they produce on Islam a pile of horse sh*t.
But he’s a scholar. A professional. He understands how research works and the rules of public debate. Yes, I know my optimistic streak is still with me and it does keep getting hit for a six.
Coyne, McGrath, Hurtado, Keith, Crossley, Le Donne, Joe Hoffmann, Casey, “N.T. Wrong”, West (well, Okay, West is only a pretend scholar with sham qualifications so I guess we can excuse him). . . .
(You’re correct, I have rarely read his non-evolution posts (whether about cats, trips or Islam) except to keep in touch with enough info to know he’s still not learning anything.)