Paul�s basket escape from Damascus (Robert Eisenman)

[Doug Shaver]

3) At this point, given your claim that "Our evidence for the hypothesis is the New Testament", one wonders why you do not even consider the texts of Acts and 2 Thessalonians.

I assume you intended the second reference to be I Thessalonians, chapter 2. There is no mention there of either a trial or any involvement by the Sanhedrin.

I'm still not going to read the entire book of Acts trying to find whatever reference you have in mind there. Please me me a specific citation.

[Paprika]

Now that we have clearly established what your E is,

The sources don't matter much.

They do matter. Do we accept Josephus on the trial specifically? Do we accept Josephus' and the Talmud's general description or portrayal of the Sanhedrin? I wonder if you have run Bayes' Theorem on them? Or is application of Bayes' Theorem only for sources which you dispute?

For any particular application of Bayes' Theorem, any fact not in dispute counts as background knowledge, and nothing else does. If you and I don't agree on whether something is a fact, then we need to resolve that disagreement before we can go any further with Bayes.

As in my post above, it would be most enlightening if you can quote and cite any authoritative source to show that your application of Bayes' Theorem is valid.

[Doug Shaver]

The references in in Acts 5:29-32 and 7:52-53 point to the active complicity of the Sanhedrin in the death of Jesus.

Thank you. I will work them into my analysis.

[Doug Shaver]

Are you now allowing for the fact that historical probability is not merely confined to that produced by Bayesian analysis?

I wouldn't say that. Historical probabilities have to be arrived at by use of evidence. Bayes tells us how the evidence can be applied logically. I fail to see any reason to think any non-logical method could justify our belief in any historical proposition.

I don't endorse the application of Bayes' Theorem to historical questions at all.

Have you any reason for that attitude, aside from the apparent fact that Bayesian analyses tend to controvert your religious dogmas?

It would be most enlightening if you could produce any authoritative literature to show that Bayes' Theorem is a normal way historical scholars evaluate evidence.

I strongly doubt that you would regard as authoritative any literature that I could cite. Maybe -- emphasis on maybe -- I will try, though, if you will tell me the criteria you use for deciding whether literature is authoritative.

As to my objection: you rule out the gospels, which are our main sources for who Jesus was, his character, his aims, his actions.

You objection would be well founded if I were ruling them out as evidence. I am not. I am ruling them out as a source of background knowledge. Any fact, to count as background knowledge, must be undisputed. Nothing in the gospels is undisputed except their claim that Jesus had a preaching ministry and was crucified by Pontius Pilate. Actually, quite a few people dispute even that, but I'm ignoring them for purposes of this discussion.

[Paprika]

I wouldn't say that. Historical probabilities have to be arrived at by use of evidence. Bayes tells us how the evidence can be applied logically. I fail to see any reason to think any non-logical method could justify our belief in any historical proposition.

So you're saying that historical work to judge whether a certain event did or didn't happen can't be done except by using Bayes' Theorem?

Have you any reason for that attitude, aside from the apparent fact that Bayesian analyses tend to controvert your religious dogmas?

It's hardly a fact; though if you want to establish it I welcome you to try. My attitude is because (as far as I know) most historians, including historians studying Jesus, simply don't use Bayes' Theorem to do analysis.

I strongly doubt that you would regard as authoritative any literature that I could cite. Maybe -- emphasis on maybe -- I will try, though, if you will tell me the criteria you use for deciding whether literature is authoritative.

I am looking for a book written by at least one credentialed historian on historiography. And now that you've asked me what I consider "authoritative", I would be interested in your criteria.

You objection would be well founded if I were ruling them out as evidence. I am not. I am ruling them out as a source of background knowledge. Any fact, to count as background knowledge, must be undisputed. Nothing in the gospels is undisputed except their claim that Jesus had a preaching ministry and was crucified by Pontius Pilate. Actually, quite a few people dispute even that, but I'm ignoring them for purposes of this discussion.

Earlier you ruled out the gospels sans trial narratives out as not part of E:

then what happens is that the rest of the gospels, (ie the gospels sans trial narratives) are neither part of B or E.

That is correct.

[Doug Shaver]

The sources don't matter much.

They do matter.

Not just because you say so. Background knowledge by definition is everything we know or think we know before we examine the evidence. The means by which we acquire that knowledge is beside the point. If you happen to question any source, then for you, what that source says is not part of your background knowledge, and if I'm debating with you, I'm wasting my time if I count it as part of my background knowledge.

Do we accept Josephus on the trial specifically?

He doesn't mention any trial of Jesus.

Do we accept Josephus' and the Talmud's general description or portrayal of the Sanhedrin?

We accept as background knowledge anything about the Sanhedrin that is not in dispute.

I wonder if you have run Bayes' Theorem on them? Or is application of Bayes' Theorem only for sources which you dispute?

It's not about checking sources. It's about checking what we believe, though that would include whatever we believe about our sources. Bayes' Theorem is a way to check the logical consistency of our reasoning from any evidence to any conclusions that are supposed to be based on that evidence. It is universally applicable, but if we actually tried to apply to everything we believe, we'd never have to time to get anything else done.

[robrecht]

So, Doug, is 'background knowledge" in this Bayesian context subjectively defined, either individually or by consensus of a specific group?

"We wish to use background information, not to express degrees of subjective belief, but to avoid being misled by our subjective beliefs, biases and desires!"
http://errorstatistics.com/2013/07/2...ive-beliefs-2/

Thus background knowledge is identified as a source of potential error?

[Paprika]

Not just because you say so. Background knowledge by definition is everything we know or think we know before we examine the evidence. The means by which we acquire that knowledge is beside the point. If you happen to question any source, then for you, what that source says is not part of your background knowledge, and if I'm debating with you, I'm wasting my time if I count it as part of my background knowledge.

If you say that the sources don't matter but the facts themselves, then it falls to you in any case to lay what what you believe are the background facts.

He doesn't mention any trial of Jesus.

My mistake; I meant their involvement in his death

We accept as background knowledge anything about the Sanhedrin that is not in dispute.

What do you mean by not in dispute? Not in dispute by any historical scholars?

It's not about checking sources. It's about checking what we believe, though that would include whatever we believe about our sources. Bayes' Theorem is a way to check the logical consistency of our reasoning from any evidence to any conclusions that are supposed to be based on that evidence. It is universally applicable, but if we actually tried to apply to everything we believe, we'd never have to time to get anything else done.

So if you don't apply it to all your beliefs, then why do you apply it to some beliefs and not others?

[Doug Shaver]

Earlier you ruled out the gospels sans trial narratives out as not part of E:

then what happens is that the rest of the gospels, (ie the gospels sans trial narratives) are neither part of B or E.

That is correct.

Considering the context, I thought my intended meaning was obvious. Those parts of the gospels that have nothing to say about the trial are not part of the evidence for the trial, which is the only E that is relevant to this discussion.

So you're saying that historical work to judge whether a certain event did or didn't happen can't be done except by using Bayes' Theorem?

No, not saying it can't be. Saying it shouldn't be.

I would also say, though, that it is routinely applied by people who don't know they're applying it. Whenever a cogent argument is made from some body of evidence to a justified conclusion, it is a Bayesian argument or equivalent thereto. It's like a carpenter who knows that a 3-4-5 triangle has a square corner. He is using the Pythagorean theorem even if he has never spent one day in a geometry class and wouldn't know Pythagoras from Pygmalion.

Have you any reason for that attitude, aside from the apparent fact that Bayesian analyses tend to controvert your religious dogmas?

It's hardly a fact; though if you want to establish it I welcome you to try. My attitude is because (as far as I know) most historians, including historians studying Jesus, simply don't use Bayes' Theorem to do analysis.

OK, so your reason for rejecting it is an argument from expert consensus. You don't know why they don't use Bayes, you just know they don't, and that satisfies you.

Maybe -- emphasis on maybe -- I will try, though, if you will tell me the criteria you use for deciding whether literature is authoritative.

I am looking for a book written by at least one credentialed historian on historiography.

So, you equate credentials with authority? Then here is my authority: Carrier, Richard C. . Amherst, NY: Prometheus Books, 2012.

And now that you've asked me what I consider "authoritative", I would be interested in your criteria.

I will acknowledge that someone is an authority if they are regarded as such by other members of their scholarly community. However, I'm getting too old to be relying solely on authorities to tell me what I should believe. I'm not content to know that lots of authorities say such-and-such. I want to know what evidence they're using and how they argue from that evidence to their conclusions.

[Paprika]

Considering the context, I thought my intended meaning was obvious. Those parts of the gospels that have nothing to say about the trial are not part of the evidence for the trial, which is the only E that is relevant to this discussion.

So what makes me wrong when I said that "you rule out the gospels, which are our main sources for who Jesus was, his character, his aims, his actions", ie. (as I intended) the gospels sans trial narratives, out from the discussion altogether?

No, not saying it can't be. Saying it shouldn't be.

I would also say, though, that it is routinely applied by people who don't know they're applying it. Whenever a cogent argument is made from some body of evidence to a justified conclusion, it is a Bayesian argument or equivalent thereto. It's like a carpenter who knows that a 3-4-5 triangle has a square corner. He is using the Pythagorean theorem even if he has never spent one day in a geometry class and wouldn't know Pythagoras from Pygmalion.

Do you mind demonstrating this in clear detail?

OK, so your reason for rejecting it is an argument from expert consensus. You don't know why they don't use Bayes, you just know they don't, and that satisfies you.

You misunderstand it. I haven't rejected it; I merely haven't endorsed it.

So, you equate credentials with authority? Then here is my authority: Carrier, Richard C. . Amherst, NY: Prometheus Books, 2012.

I asked not just for citation, but quotations.? To reference my original request: "It would be most enlightening if you could produce any authoritative literature to show that Bayes' Theorem is a normal way historical scholars evaluate evidence." Could you quote Carrier to show this? Also, how does Carrier use Bayes' Theorem? Is your usage in line with his?

I will acknowledge that someone is an authority if they are regarded as such by other members of their scholarly community. However, I'm getting too old to be relying solely on authorities to tell me what I should believe. I'm not content to know that lots of authorities say such-and-such. I want to know what evidence they're using and how they argue from that evidence to their conclusions.

My concern here is the methodology. When someone like yourself proposes and uses methodology that very few historical scholars, to the best of my knowledge, use, it is very reasonable for me to view that methodology with wariness.

[Doug Shaver]

So, Doug, is 'background knowledge" in this Bayesian context subjectively defined, either individually or by consensus of a specific group?

I'll give you my take on it, but I strongly urge anyone who wants to really get it to do lots of googling. I am confident that my understanding conforms to what most of the real experts would say, but I wouldn't dare ask anyone to just take my word for it.

You have a proposition P, and you have some fact or set of facts E that you think is or might be evidence for P. Strictly speaking, your background knowledge is all of the knowledge you have other than E -- every last thing you know or think you know. Of course, usually most of it will be irrelevant to any consideration of P's likelihood, but this cannot be decided arbitrarily. No fact can be ignored just because it might influence your analysis in some way you'd like to avoid.

It would say it is defined for and by everyone participating in a given discussion about a particular proposition. If you and I are discussing the authenticity of the Testimonium Flavianum, neither of us can be held to account for any facts of which neither of us is aware, no matter how many other people might already have learned them. Of course, if those facts should be taken into account, then even if we both accept the outcome of a Bayesian analysis of the Testimonium, we could both be wrong. Bayes can't tell us the actual truth about anything. All it can tell us is whether, considering the information we actually have to work with, we're being logically consistent in our application of that knowledge.

Thus background knowledge is identified as a source of potential error?

The error would lie in disregarding any portion of it that ought to influence the probabilities we assign to the equation's variables.

If I want to know whether Josephus wrote some part of the Testimonium Flavianum, then chances are it will be safe for me to ignore anything I know about the pre-Columbian Aztec civilization. But if I also happen to know something about Philo's writings, then I'd better pay some attention to it. That noted . . . suppose someone challenges my analysis by saying, "Actually, you should revise your estimate of P(E|~H), because we know that the pre-Columbian Aztecs did such-and-such." Then I might have a problem, depending on how good an argument my adversary has.

[Doug Shaver]

If you say that the sources don't matter but the facts themselves, then it falls to you in any case to lay what what you believe are the background facts.

You're asking me to tell you everything I have ever learned throughout my entire life. Nobody can do that.

I meant their involvement in his death.

If a source does not say there was a trial, it is not evidence for a trial.

What do you mean by not in dispute? Not in dispute by any historical scholars?

I mean not disputed by anyone participating in the discussion.

Bayes' Theorem is a way to check the logical consistency of our reasoning from any evidence to any conclusions that are supposed to be based on that evidence. It is universally applicable, but if we actually tried to apply to everything we believe, we'd never have to time to get anything else done.

So if you don't apply it to all your beliefs, then why do you apply it to some beliefs and not others?

I believe you already have my answer. I do not refuse to apply it to anything, but I cannot apply it to everything. Whether I apply it to a particular proposition depends on my interest in the proposition and the time available to me.

[Doug Shaver]

So what makes me wrong when I said that "you rule out the gospels, which are our main sources for who Jesus was, his character, his aims, his actions", ie. (as I intended) the gospels sans trial narratives, out from the discussion altogether?

What makes you wrong is your failure to remember that the discussion is not over yet. Or at least, I'm not done with it.

I would also say, though, that it is routinely applied by people who don't know they're applying it. Whenever a cogent argument is made from some body of evidence to a justified conclusion, it is a Bayesian argument or equivalent thereto.

Do you mind demonstrating this in clear detail?

From scratch? I'd have to write a whole book. If you think you know a counterexample, I could do something with that.

I haven't rejected it; I merely haven't endorsed it.

Very well. I stand corrected.

I asked not just for citation, but quotations.? To reference my original request: "It would be most enlightening if you could produce any authoritative literature to show that Bayes' Theorem is a normal way historical scholars evaluate evidence."

All right.

Source: Carrier, Proving History, p. 45

Since BT is formally valid and its premises (the probabilities we enter into it) constitute all that we can relevantly say about the likelihood of any historical claim being true, it should follow that all valid historical reasoning is described by Bayes's Theorem . . . . If I'm correct, and it is true that BT models what all historians actually do when they think and reason correctly about evidence and explanations, historians would do well to know more about it.

© Copyright Original Source

When someone like yourself proposes and uses methodology that very few historical scholars, to the best of my knowledge, use, it is very reasonable for me to view that methodology with wariness.

The scholars who do history right are using Bayes. Most of them might not know it, but they're using it, just like that carpenter is using Pythagoras.

[Paprika]

You're asking me to tell you everything I have ever learned throughout my entire life. Nobody can do that.

Then at least lay out the relevant background facts pertaining to the Sanhedrin.

If a source does not say there was a trial, it is not evidence for a trial.

The gospel texts do not say it was a "trial" either.

[Paprika]

I'll give you my take on it, but I strongly urge anyone who wants to really get it to do lots of googling. I am confident that my understanding conforms to what most of the real experts would say, but I wouldn't dare ask anyone to just take my word for it.

You have a proposition P, and you have some fact or set of facts E that you think is or might be evidence for P. Strictly speaking, your background knowledge is all of the knowledge you have other than E -- every last thing you know or think you know. Of course, usually most of it will be irrelevant to any consideration of P's likelihood, but this cannot be decided arbitrarily. No fact can be ignored just because it might influence your analysis in some way you'd like to avoid.

You're asking me to tell you everything I have ever learned throughout my entire life. Nobody can do that.

The question raised as this juncture is whether anyone can practically use "every last thing you know or think you know" in a Bayesian analysis. Do you?