BookTalk.org

Tim Hendrix wrote: Interesting. I can only quote OHJ: "if I assign ~h to be the theory defined by Premises 1 through 5".

Hmm. Let's continue the quote:

On the historicity of Jesus, p. 55 wrote: if I assign ¬h to be the theory defined by Premises 1 through 5, I can safely assume that h entails historicity (given my minimal definition of historicity as a hypothesis in Chapter 2) and that these exhaust all relevant possibilities, and therefore I have a proper binary test, h and ¬h, just two hypotheses to compare against each other, such that if one is false, the other is true.
Certainly, when framed like this, technically ¬h (non-historicity) must also include all Jesus myth theories not defined by Premises 1 through 5 (that is, all theories of the evidence for Jesus that entail historicity is false and at least one of Premises 1 through 5 is false), but since their prior probability (even collectively) is surely less than a tenth of one percent (as I just reasoned) and their posterior probability not sufficiently high to make enough of a difference (especially in relation to minimal historicity), these theories share such a small portion of the probability-space occupied by ¬h that they can simply be ignored.

I can only assume that your copy of the book has suffered some terrible accident damaging page 55, because surely it must be above you to provide a misleading quote on purpose.

Tim Hendrix wrote: If you don't understand why you have to use the same numbers as me (which are from OHJ) in order to overturn or criticize an argument based on those numbers I can't really help you .

It's not my fault that you use the wrong numbers. No one - not Carrier, not me, not you - considers those numbers to be correct and unbiased. So why do you insist on using them?

Tim Hendrix wrote: I agree that a 5% bias won't overturn Carriers upper bound on mythicism. But that is simply saying that starting out from a position where the conclusion is very clear, limited bias will be magnified so as to weaken it but not fully overturn it: Bias is still magnified in both cases.*

If it is 'magnified', it must be magnified in relation to something else. What would this be?

Really, what you have done is just to take a case where the available evidence is almost exactly balanced in favour of two hypotheses. Only the prior here gives a slight edge to non-historicity. Now, you complain that the calculation is not 'stable'. But what can we expect in such a case? We do not need to use any math to realize that if the evidence is completely 50/50, the slightest amount of bias, or even differing perspectives, will cause different referees to slightly favour different positions. This is not a flaw in the calculation. Really - if it was 'stable', it would be wrong. When you model an unstable system - and this is such a system - your model should also be 'unstable' in the relevant range.

Now, can I just recommend that you read a little from the book. You don't need to open it this time, just the cover.
"On the Historicity of Jesus
Why We Might Have Reason for Doubt"

Notice the subtitle. It is not 'Jesus almost certainly never existed'. Even though that is Carrier's conclusion given his own best judgement of the facts. He is, indeed, humble enough to recognize that other reasonable people may come to different conclusions, and may have valid counter-arguments to his arguments. He is judging the evidence as best as he can - but also inviting others to offer differing opinions.

Tim Hendrix wrote: Please give me the general formula which gives the numerical value of two probabilities from their ratio. Let us make it concrete. Suppose I tell you that:

P(A)/P(B) = 1.07

Please tell me what value P(A) and P(B) originally had. I am waiting with baited breath .

There are of course an infinite number of possible values so I choose the ones where the larger value is 100%, in this case P(A). Therefore, P(B) is 1/1.07 or about 93.5%. See, it was not that hard. For the general case, if the odds are below 1 it will be P(B) that is 1 and P(A) will be equal to the odds value. Now you will no doubt complain that I made a hidden assumption to use the largest possible values and that it is possible that the original values were something different, but of course I never claimed to have done anything else. For the purpose of adding bias I want those probabilities that are scaled to a relative maximum, so as to make the bias term more uniform. If you had just asked nicely instead of immediately assuming I made an error, you would not have had as much egg on your face momentarily.

Tim Hendrix wrote: I believe (and still do) historians agree with Ehrman Jesus existed and it will be easy to find historians who in their writings clearly indicate so much. I don't think I can find literature of historians who sets out to defend the existence of Jesus since his non-existence is considered too fringe.

They don't have to defend historicity as such, they just need to have published something about the historical Jesus. Now, I think it is pretty obvious that if we made a poll of who was the most important human being in history, Jesus would come out very near the top, most probably as number 1.

But I guess your next move is to say that of course no historian would ever want to research such an obscure subject as the most important person in history..

JohanRonnblom wrote:

Tim Hendrix wrote: Interesting. I can only quote OHJ: "if I assign ~h to be the theory defined by Premises 1 through 5".

Hmm. Let's continue the quote:

On the historicity of Jesus, p. 55 wrote: if I assign ¬h to be the theory defined by Premises 1 through 5, I can safely assume that h entails historicity (given my minimal definition of historicity as a hypothesis in Chapter 2) and that these exhaust all relevant possibilities, and therefore I have a proper binary test, h and ¬h, just two hypotheses to compare against each other, such that if one is false, the other is true.
Certainly, when framed like this, technically ¬h (non-historicity) must also include all Jesus myth theories not defined by Premises 1 through 5 (that is, all theories of the evidence for Jesus that entail historicity is false and at least one of Premises 1 through 5 is false), but since their prior probability (even collectively) is surely less than a tenth of one percent (as I just reasoned) and their posterior probability not sufficiently high to make enough of a difference (especially in relation to minimal historicity), these theories share such a small portion of the probability-space occupied by ¬h that they can simply be ignored.

I can only assume that your copy of the book has suffered some terrible accident damaging page 55, because surely it must be above you to provide a misleading quote on purpose.

Well, all I can say is that the quote does not alter the fact that Carrier "assign ~h to be the theory defined by Premises 1 through 5". What else is ~h assigned to? Carrier even acknowledges this is not technically correct in the very quote you cite, but provides an argument as to why he believes it does not matter from a pragmatic perspective. That does not make it technically correct though.

JohanRonnblom wrote:

Tim Hendrix wrote: If you don't understand why you have to use the same numbers as me (which are from OHJ) in order to overturn or criticize an argument based on those numbers I can't really help you .

It's not my fault that you use the wrong numbers. No one - not Carrier, not me, not you - considers those numbers to be correct and unbiased. So why do you insist on using them?

The numbers are from OHJ; I use them because it is easier for a reader to relate to about 1/3 than about 1/12000.

JohanRonnblom wrote:

Tim Hendrix wrote: I agree that a 5% bias won't overturn Carriers upper bound on mythicism. But that is simply saying that starting out from a position where the conclusion is very clear, limited bias will be magnified so as to weaken it but not fully overturn it: Bias is still magnified in both cases.*

If it is 'magnified', it must be magnified in relation to something else. What would this be?

...the unbiased value, i.e. the value in OHJ. This is not difficult, I included the relevant graph a few pages ago.

JohanRonnblom wrote: Now, you complain that the calculation is not 'stable'. But what can we expect in such a case? We do not need to use any math to realize that if the evidence is completely 50/50, the slightest amount of bias, or even differing perspectives, will cause different referees to slightly favour different positions. This is not a flaw in the calculation. Really - if it was 'stable', it would be wrong. When you model an unstable system - and this is such a system - your model should also be 'unstable' in the relevant range.

Wow, that sure sounds like you are acknowledging that the calculation is affected by bias. I guess in 10 more pages or so you might acknowledge the basic fact from error analysis that that bias will tend to be increased by the multiplication of many terms...

JohanRonnblom wrote:

Tim Hendrix wrote: Please give me the general formula which gives the numerical value of two probabilities from their ratio. Let us make it concrete. Suppose I tell you that:

P(A)/P(B) = 1.07

Please tell me what value P(A) and P(B) originally had. I am waiting with baited breath .

There are of course an infinite number of possible values so I choose the ones where the larger value is 100%, in this case P(A). Therefore, P(B) is 1/1.07 or about 93.5%.

Hm. It seems like there was no general formula to get the probabilities from a ratio after all so I guess I get to keep my dayjob? .
So to bring that into context of my previous point: When you assign one probability to be 100%, you are both doing something which is unreasonable considering the situation (it is for instance not reasonable to assign a probability of 100% to the Gospels because surely, it is not a given that the four gospels with their specific stories would have a 100% chance being produced in their present form), but it is also affecting your analysis of bias because your arbitrary choice of 100% then induces a cutoff. But I guess you will now explain why that is all wrong using the word "complain" a lot .

Tim Hendrix wrote: Well, all I can say is that the quote does not alter the fact that Carrier "assign ~h to be the theory defined by Premises 1 through 5". What else is ~h assigned to? Carrier even acknowledges this is not technically correct in the very quote you cite, but provides an argument as to why he believes it does not matter from a pragmatic perspective. That does not make it technically correct though.

No, he explains how to make it technically correct and makes it technically correct. I'm sorry that you refuse to understand this. Maybe if you considered what it would take to make it technically correct, you would realize that this is exactly what Carrier does. But probably not, I've given up. You're not susceptible to reason.

Tim Hendrix wrote:

JohanRonnblom wrote: If it is 'magnified', it must be magnified in relation to something else. What would this be?

...the unbiased value, i.e. the value in OHJ. This is not difficult, I included the relevant graph a few pages ago.

We don't have the unbiased value. So this is like complaining because Usain Bolt didn't run 100 m in 5 seconds. Do it better yourself first, then we can talk.

Moreover, this is completely technically incorrect. The unbiased value - by definition - has no bias. Therefore, magnifying the bias of the unbiased value will produce something that also has no bias. This is a fairly fundamental property of multiplication.

Tim Hendrix wrote: Wow, that sure sounds like you are acknowledging that the calculation is affected by bias.

The world is affected by bias. It's not a problem of the calculation. If you create a calculation that is not affected by real-world bias, your calculation will be wrong.

Tim Hendrix wrote: Hm. It seems like there was no general formula to get the probabilities from a ratio after all so I guess I get to keep my dayjob? .

If it was up to me, no way. You clearly have no idea what you're talking about, dressing up your lack of relevant knowledge in fancy terms you don't really grasp in the hope that no one else will either.

I'm done with you.

But, if anyone knows of an historian who has published anything about Jesus in a somewhat scholarly context, I'm genuinely interested in knowing, so I would appreciate a post in this thread. While the apparent lack of historians publishing about Jesus evidently does not prove that Jesus was not historical, it should in my opinion at least provide a strong reason for doubt. It is something of a definition of being historical that you're someone historians write about - at least if you're important. And I doubt many historicists would claim that Jesus is unimportant.

JohanRonnblom wrote:

Tim Hendrix wrote: Well, all I can say is that the quote does not alter the fact that Carrier "assign ~h to be the theory defined by Premises 1 through 5". What else is ~h assigned to? Carrier even acknowledges this is not technically correct in the very quote you cite, but provides an argument as to why he believes it does not matter from a pragmatic perspective. That does not make it technically correct though.

No, he explains how to make it technically correct and makes it technically correct.

I don't think this can persuade me from what I think the passage plainly says. I think it is best to simply agree to disagree.

JohanRonnblom wrote:

Tim Hendrix wrote:

JohanRonnblom wrote: If it is 'magnified', it must be magnified in relation to something else. What would this be?

...the unbiased value, i.e. the value in OHJ. This is not difficult, I included the relevant graph a few pages ago.

We don't have the unbiased value. So this is like complaining because Usain Bolt didn't run 100 m in 5 seconds. Do it better yourself first, then we can talk.

Well, all I can do is to go back to the basics of error analysis: When you analyze the stability of an equation, you examine what the equation does when the inputs are perturbed: Is the perturbation increased in magnitude or decreased (and by which order). This analysis is independent of whether you believe you have the correct input values in some platonic sense as the conclusion is simply: Over a reasonable range of input values this is what the equation actually does.

That you have another interpretation of what error analysis is is very interesting and all, but I can't see why I can't include an example which shows what the equation does when the inputs are biased. Once more, I feel this very basic material...

JohanRonnblom wrote:

Tim Hendrix wrote: Wow, that sure sounds like you are acknowledging that the calculation is affected by bias.

The world is affected by bias. It's not a problem of the calculation. If you create a calculation that is not affected by real-world bias, your calculation will be wrong.

Well, I agree. It almost seems like the next natural step is to examine how the equation is affected by bias, i.e. if the bias is increased or decreased. How might one do this... I guess the simplest way would be to perturb the input values and see how that affected the result of the computation in a graph...

JohanRonnblom wrote:

Tim Hendrix wrote: Hm. It seems like there was no general formula to get the probabilities from a ratio after all so I guess I get to keep my dayjob? .

If it was up to me, no way. You clearly have no idea what you're talking about, dressing up your lack of relevant knowledge in fancy terms you don't really grasp in the hope that no one else will either.

Then all I can do is hope you are not my PI .
I find it interesting that you feel this discussion has to include so many insults and ascribing an emotional response to the person you are arguing with, but I guess it simplifies things.