Sounds to me as if Carrier wants to have his cake and eat it. On the one hand he wants us to think that this is a valuable tool that will work better that the alternatives and on the other he wants to say that this is what we already do anyway. If criticism of your new tool is really criticism of all human reasoning then you haven't added anything new.

logical bob wrote:Sounds to me as if Carrier wants to have his cake and eat it. On the one hand he wants us to think that this is a valuable tool that will work better that the alternatives and on the other he wants to say that this is what we already do anyway. If criticism of your new tool is really criticism of all human reasoning then you haven't added anything new.

Part of his thesis is that all valid reasoning can be stated in Bayesian terms.

If that is correct, then it's true junking Bayes' Theorem is junking valid reasoning.

Carrier is suggesting making use of this powerful tool to help historians avoid the common mistakes scholars have noticed.

So what's 'new' is using a template of valid reasoning instead of the muddle in place now.

Mus Ponticus wrote:Since Byron isn't answering ...

This isn't an interrogation. I answer exactly if and when I want to. You didn't offer any examples before, so I had no interest in answering.

... I would like to quote Richard Carrier's answer in an interviewfrom last year:

RICHARD: That’s what the book is for, it addresses tons of them, and I’m sure your audience is listening now like: “Ah, what about this? No, it can’t work because…”

Well, what are the big ones? Let’s take one that even scientists debate even among scientists in applying Bayes’ theorem: is the problem of subjective probability estimates.

Obviously, I mean obviously, in history especially – but even in science this is often the case – we don’t have hard, scientifically verified statistical data. We don’t, we can’t poll – we can’t take a scientific phone poll of ancient Roman populations, right? Things like that, you don’t have that kind of data.

There are few cases where you do, very few, and it’s very limited what we can learn from them. Most history doesn’t have access to that data. So you have to give a sort of subjective probability estimate, because people would say “you’re just making shit up, or you’re guessing, or something”.

What I point out in the book and demonstrate in detail, is that: this is how we reason all the time anyway, so if this is a valid objection to Bayes’ theorem, it’s a valid objection to all of human reasoning.

So Carrier is of course not claiming that his method can give some "objective" probability numbers for various historical questions (like the meaning of the "brother of the lord"). In fact, here he says that one of the most common objections to using Bayes' theorem is that you just have to pull the numbers out of thin air ("subjective probability estimate"). His response to that objection isn't "Oh... these numbers aren't subjective!", like one would expect from reading Byrin, but "Right, that's what we do all the time when we reason."

So, Byron, maybe you should try to put your personal dislike of Carrier aside, and listen to that interview, so you would at least understand what he's saying.

I agree that I don't understand what he's saying, but that's because his position is so muddled I doubt the Bletchley Park folks could fully decode it.

I've said before that, to my knowledge, I have no personal dislike of Carrier. I don't like his methods or approach.

From what I can tease out, Carrier is saying that historians aim to produce objective probabilities even if they're unaware of it:-

... it's like, in history we don't need exact probabilities of anything, even the subjective probabilities are really objective because if you're going to make an argument to another historian and say, "You should agree with me on this," even if you don't do it based on Bayes Theorem, you're still saying "you should agree that these things are likely."

If you do it with Bayes Theorem it's exactly the same, "Well you should agree that the probability is at least sixty seven percent," and if you agree with that, you have to agree with the conclusion, 'cause it follows necessarily by deductive logic."

And that's the way I think historians should be arguing amongst each other. And a lot of disagreements could be resolved, even to the point where both sides agree we can’t really know what the answer is.

In the pdf of his lecture, one section's titled, "Advancing to Increasingly Objective Estimates," and elaborates, "The fact that we can improve the certainty of our conclusions by improving the certainty of our estimates of likelihood and unlikelihood is precisely what historians need to learn from Bayes' Theorem."

So Carrier thinks that historians can use math to infuse objectivity into assessments of historical probability. The problem comes in his failure to demonstrate how the formula transfers to the evidence.

It sounds to me like Carrier is saying that when historians argue

"You should agree with me on this," even if you don't do it based on Bayes Theorem, you're still saying "you should agree that these things are likely."

based on an informal sense of "likelihood" what they really ought to be arguing is a faux formal sense of likelihood with percentages of likelihood arbitrarily assigned to each underlying premise.

What I point out in the book and demonstrate in detail, is that: this is how we reason all the time anyway, so if this is a valid objection to Bayes’ theorem, it’s a valid objection to all of human reasoning.

If we're reasoning this way "all the time" (and I'll concede that we often do) what does imposing Bayes Theorem upon these arguments actually add to the argument. Particularly if, as he seems to concede here

Obviously, I mean obviously, in history especially – but even in science this is often the case – we don’t have hard, scientifically verified statistical data. We don’t, we can’t poll – we can’t take a scientific phone poll of ancient Roman populations, right? Things like that, you don’t have that kind of data.

there is no way to assign these probabilities without being arbitrary.

The point I'm making is thus: if we have no way of being any more precise than "likely" or "unlikely" then what good does arbitrarily affixing a number to these assessments do?

Absolutely none as far as I can see. It's a pointless endeavor.


None of this, by the way, objects to the validity of Bayes Theorem. It objects to it's usefulness in the case of historical arguments. If we can't define the probability of any of our underlying assumptions with a reliable and applicable data set then there's simply no way we can usefully apply Bayes Theorem to the question being examined.

If we actually did have access to all the relevant data for all of our underlying assumptions then, sure, we could apply Bayes Theorem. As it is it would be an exercise in futility to attempt to do so.

logical bob wrote:Sounds to me as if Carrier wants to have his cake and eat it. On the one hand he wants us to think that this is a valuable tool that will work better that the alternatives and on the other he wants to say that this is what we already do anyway. If criticism of your new tool is really criticism of all human reasoning then you haven't added anything new.

You're not being very logical, bob. Carrier is merely saying that this specific objection is as valid against Bayes' (I will from now on abbreviate it as BS ), as it is against normal reasoning.

E.g. next time Byron says: "It's most probable that the meaning of 'brother of the lord' is biological.", the same objection would be: "Probable? That's a mathematical term. Show me the math or stop using that term! Are you just pulling 'most probable' out of thin air?"

Byron wrote:This isn't an interrogation. I answer exactly if and when I want to. You didn't offer any examples before, so I had no interest in answering.

You made a specific claim regarding Carrier's case, and I asked you to back that claim up. I don't know why I should give examples when asking you to back up your claims.

Byron wrote:I agree that I don't understand what he's saying, but that's because his position is so muddled I doubt the Bletchley Park folks could fully decode it.

Blame the messenger

In the pdf of his lecture, one section's titled, "Advancing to Increasingly Objective Estimates," and elaborates, "The fact that we can improve the certainty of our conclusions by improving the certainty of our estimates of likelihood and unlikelihood is precisely what historians need to learn from Bayes' Theorem."

So Carrier thinks that historians can use math to infuse objectivity into assessments of historical probability. The problem comes in his failure to demonstrate how the formula transfers to the evidence.

I'm not sure what you think he's saying there. Look at this quote from that same section:

In other words, the question must be asked, “How do you get those values? Do you just pull them out of your ass?” In a sense, yes, but in a more important sense, no.

You aren’t just blindly making up numbers. You have some reason for preferring a low number to a high one, for example (or a very low one to one that’s merely somewhat low, and so on), and the strength of that reason will be the strength of any conclusion derived from it—by the weakest link principle, i.e. any conclusion from Bayes’ Theorem will only be as strong as the weakest premise in it (i.e. the least defensible probability estimate you employ).

Byron, imagine this. If you take up a book on the HJ, would you think it absurd if the author, instead of saying "probably true", "very probably true" and "almost certainly true", used numbers instead? E.g. "70% probability of being true", "80% probability of being true" and "90% probability of being true"?

TheOneTrueZeke wrote:None of this, by the way, objects to the validity of Bayes Theorem. It objects to it's usefulness in the case of historical arguments. If we can't define the probability of any of our underlying assumptions with a reliable and applicable data set then there's simply no way we can usefully apply Bayes Theorem to the question being examined.

If we actually did have access to all the relevant data for all of our underlying assumptions then, sure, we could apply Bayes Theorem. As it is it would be an exercise in futility to attempt to do so.

Yup. As I said a few pages back, values have to be assigned to questions without a dataset ... and Carrier offers no formula to make that assignment.

The sad thing is, he could write a whole book on this question without spotting this glaring error in his argument. This is precisely why academia employs the peer-review process. It's not, as Carrier seems to think, there to block his brilliant theories. It's there to save him from making an arse of himself.

Mus Ponticus wrote:Byron, imagine this. If you take up a book on the HJ, would you think it absurd if the author, instead of saying "probably true", "very probably true" and "almost certainly true", used numbers instead? E.g. "70% probability of being true", "80% probability of being true" and "90% probability of being true"?

You didn't ask me but I would think it very silly indeed. I would be hard pressed to keep my eyes frontwards.

Mus Ponticus wrote:Byron, imagine this. If you take up a book on the HJ, would you think it absurd if the author, instead of saying "probably true", "very probably true" and "almost certainly true", used numbers instead? E.g. "70% probability of being true", "80% probability of being true" and "90% probability of being true"?

I'd think it was eccentric, but not absurd. I would however think it was absurd if they claimed that those numbers conferred any objective advantage to their subjective assessment.

Not that books on the historic Jesus (or any historical subject) tend to obsess about probability in the way Carrier is. He's blown this way out of proportion. His Bayes material may be useful in socio-economic history, but that's a specialized field, and he's not limiting himself to it. If he did want to go that route, he'd need to frame his argument in an article and submit it to a journal in socio-economic history.

TheOneTrueZeke wrote:You didn't ask me but I would think it very silly indeed. I would be hard pressed to keep my eyes frontwards.

I'd find it odd if any book gave so much focus to probability. When Dale Allison examined whether Jesus was an ascetic, he didn't trawl through the gospels making probabilistic assessments: he examined what the material was saying and offered conclusions. It was a qualitative, not a quantitative exercise.

Likewise, when, say, a historian makes a study of 14th Century land ownership, they examine the records and give conclusions. There's no need to abstract it in the way Carrier is.

Byron wrote:Yup. As I said a few pages back, values have to be assigned to questions without a dataset ... and Carrier offers no formula to make that assignment.

The sad thing is, he could write a whole book on this question without spotting this glaring error in his argument. This is precisely why academia employs the peer-review process. It's not, as Carrier seems to think, there to block his brilliant theories. It's there to save him from making an arse of himself.

This isn't a "glaring error in his argument". And there is no "formula" to make that assignment, there doesn't have to be.

Byron wrote:I'd think it was eccentric, but not absurd. I would however think it was absurd if they claimed that those numbers conferred any objective advantage to their subjective assessment.

"Objective" advantage? Do you mean that it would be absurd if anybody claimed that using "70%" were more objective than "probably"?

Mus Ponticus wrote:"Objective" advantage? Do you mean that it would be absurd if anybody claimed that using "70%" were more objective than "probably"?

Again, you didn't ask me but...yes, it would be emphatically absurd in the case where one cannot empirically demonstrate how that "70%" figure was derived.

Mus Ponticus wrote:

Byron wrote:Yup. As I said a few pages back, values have to be assigned to questions without a dataset ... and Carrier offers no formula to make that assignment.

The sad thing is, he could write a whole book on this question without spotting this glaring error in his argument. This is precisely why academia employs the peer-review process. It's not, as Carrier seems to think, there to block his brilliant theories. It's there to save him from making an arse of himself.

This isn't a "glaring error in his argument".

Apparently all the peer review in the world hasn't stopped virtually every historian from making the 'glaring error' of arbitrarily saying "X is probable" without being able to define what the hell it's supposed to mean.

And there is no "formula" to make that assignment, there doesn't have to be.

This seems to beg the question 'what's the formula now for assigning the term 'probable' to something we know nothing about the probabilities of?'

Byron wrote:I'd think it was eccentric, but not absurd. I would however think it was absurd if they claimed that those numbers conferred any objective advantage to their subjective assessment.

"Objective" advantage? Do you mean that it would be absurd if anybody claimed that using "70%" were more objective than "probably"?

The way things seem to stand now it's a hopeless muddle of subjectivity.

At least if someone says "X is 75% probable" the reader would have a better idea of how likely the writer intends to be than "X is very probable".

proudfootz wrote:Apparently all the peer review in the world hasn't stopped virtually every historian from making the 'glaring error' of arbitrarily saying "X is probable" without being able to define what the hell it's supposed to mean.

I'm reminded of the courtroom farce in the Australian drama Joh's Jury, where a biased holdout juror baffles both judge and jury by wasting the court's time with demands for "clarification" on reams of plain-English phrases.

"Probable" in this context means "in my subjective judgment, this conclusion is more likely than not." There's no mystery to it. It's overturned not by abstracted theorizing, but by engaging with the evidence and showing where the assessment has gone awry.

proudfootz wrote:

Apparently all the peer review in the world hasn't stopped virtually every historian from making the 'glaring error' of arbitrarily saying "X is probable" without being able to define what the hell it's supposed to mean.

It's not defined in quantitative terms. It has been defined in qualitative ones. You can comb back through the HJ thread to see that.

If you have a problem with qualitative assessments of "probable" then you have a problem with history in general.

At least if someone says "X is 75% probable" the reader would have a better idea of how likely the writer intends to be than "X is very probable".

Which would be useful if you're a bookie otherwise...not so much.

Mus Ponticus wrote:This isn't a "glaring error in his argument". And there is no "formula" to make that assignment, there doesn't have to be.

So how the fuck does Bayes' add anything?

Perhaps there should be a massive gathering of historians wherein they each subjective assess and assign a percentage value to each and every minute question of history within their field of study. Hundreds of thousands, possibly MILLIONS of tiny little possible points of fact evaluated and given exacting and excruciating assessments with numbers plucked from the sky and affixed to them. Said numbers are then crunched and statistically analyzed within in inch of their lives!

Producing?

What?

A bunch of numbers.

How useful.

proudfootz wrote:
Part of his thesis is that all valid reasoning can be stated in Bayesian terms.

If that is correct, then it's true junking Bayes' Theorem is junking valid reasoning.

Carrier is suggesting making use of this powerful tool to help historians avoid the common mistakes scholars have noticed.

So what's 'new' is using a template of valid reasoning instead of the muddle in place now.

Show, don't tell. I've already shown you how Carrier's paper
(a) fails to illustrate the use of Bayesian reasoning to justify any propositions that aren't trivially obvious anyway
(b) makes liberal use of precisely the vague and subjective approaches he purports to criticise.
It would appear you have no answer.

Mus Ponticus wrote:If you take up a book on the HJ, would you think it absurd if the author, instead of saying "probably true", "very probably true" and "almost certainly true", used numbers instead? E.g. "70% probability of being true", "80% probability of being true" and "90% probability of being true"?

I concur with m'learned friend. That would indeed be absurd. If you want to be taken seriously when you say that something is 90% probable you need to be able to show your working. 90% is roughly the probability that 9 people chosen at random will all have different birthdays. That's a fact. Using the numbers if you can't deliver the goods is just trying to look objective by disguising what you're doing as probability. You might as well try to weigh the evidence in kilograms. Seems that's what Carrier's all about here.

TheOneTrueZeke wrote:Perhaps there should be a massive gathering of historians wherein they each subjective assess and assign a percentage value to each and every minute question of history within their field of study. Hundreds of thousands, possibly MILLIONS of tiny little possible points of fact evaluated and given exacting and excruciating assessments with numbers plucked from the sky and affixed to them. Said numbers are then crunched and statistically analyzed within in inch of their lives!

Producing?

What?

A bunch of numbers.

How useful.

And then someone cries "hegemony" ...

Who does history like this anyway? How about questions like this from an A-level paper?

Was military superiority the main reason for the expansion of British influence in India in the period 1757 to 1785?

Who would expect the answer here to be that its 60% likely that this is the case? You just can't bring this method to bear on actual historical questions. But of course Carrier's not interested in history in any general sense. He's fixated on that single yes/no question.

logical bob wrote:Who does history like this anyway? How about questions like this from an A-level paper?

Was military superiority the main reason for the expansion of British influence in India in the period 1757 to 1785?

Who would expect the answer here to be that its 60% likely that this is the case? You just can't bring this method to bear on actual historical questions. But of course Carrier's not interested in history in any general sense. He's fixated on that single yes/no question.

Testify.

You'd answer that question by assessing what's meant by "military superiority," then probing whether it can be separated from other factors. If you were after one of those shiny A*s, you could examine whether the chronological cut-off distorted the inquiry. Evidence would be adduced, and conclusions drawn. Nowhere would you start churning probabilities.

As you say, Carrier, on the available evidence, doesn't give a rat's ass about history. He's never held a research position, and I've yet to see a single bit of post-doctorate research from him. He's a professional anti-theist (although I'm still baffled by what exactly it is he does for a living) and history is a means to that end.