Amaleq13

That would explain the headaches, nausea and double vision.

TedM

How do you know it is ridiculous? Bush could have said something derogatory about people name Gerard. Anyway, replace the name Gerard with the phrase "all males" if you prefer. My point is that probability doesn't try to determine relevancy to specific individuals because it CAN'T by its very nature. It determines NORMS which are apply to groups made up of individuals, not whether the norms are relevant to any specific individuals in the group. You can't know whether a specific factor is relevant for a specific individual when dealing with norms that address an entire group. The only time you will know that a specific factor is relevant to a specific individual's decision is if that individual tells if it is or not. Statistics deal with uncertainties, not certainties.


How do you know? It certain COULD HAVE been a factor. Bush could have said something derogatory about the name Gerard. Or people named Gerard may be more or less conservative due to the kinds of parents that name their son Gerard. We know that Bush didn't make some comment, but we don't always know what influences a person's decision. That's why statistics are invoked. They show what a given population does. They don't explain why they do it. They don't tell what they will do in response to other factors either. But they do give us a benchmark to use to compare individuals with norms in order to predict WHAT an individual is likely to do. Not WHY.

It is fairly intuitive to me that there is no reason to find another poll with real numbers to illustrate the point I'm making: The relevancy of the questions/factors to the specific individual is not always known and is therefore irrelevant as long as that individual is included in the population studied. If a study showed that 99.9999 people in the first century named Paul used 'apo' to mean 'indirect' then we could conclude that St. Paul likely did also without having to know WHY he did it or HOW he used 'apo' in other circumstances.

We may never be able to ask enough questions to come up with probabilities that are reasonable to you. This is my point. We can never get into the individual's mind enough to be certain unless the individual actually comes out and tells us exactly why he does what he does. Had Paul used 'apo' 100 times he could have used it differently in the case of 11:23 than all 100 of the other times on some basis unknown to us. Should we therefore scrap using probabilities altogether?

I"m not sure it does tell us much at all because it seems so unreliable to use as a predictive factor since we can't really get into the specific individual's mind. This is your point, and I agree. However, the very fact that we can't really get into their mind argues for using any and all known statistically significant factors (this shows relevancy) used by a population that includes Paul to get as close as possible to accurately guessing that those factors are also relevant to Paul, which of course they might not be at all. If one isn't willing to do this, then one will never be able to make a decision about what someone else had in mind since one can't know and there is always an alternative possibility one is unaware of. IF 'apo' had a more common use that meant indirect, it would be neglegent to not consider that fact as well as the question of how common it was when looking at Paul.


You don't and can't know that there weren't relevant for any specific individual.


Sure. People who used the word 'apo' 1000 years before or after Paul are more general and less specific to Paul. Paul is still part of the population (I recall that word from the stat courses years ago ) of those who use 'apo', but the use during Paul's time is definitely a more specific group in which he is a member.


By excluding the application to Paul it is the same as excluding him from the group. The use of probabilities by definition is NON-SPECIFIC. It applies to groups in which people are members. Again, if 99.9999% of the uses of 'apo' during Paul's day means X, then it surely wouldn't be 'too non-specific to be useful for accurately predicting the behavior of a specific individual" included in that group, which in this case is Paul. Unless you can refute this statement I think you will not succeed in convincing me otherwise.


I guess I was not clear with my words. I meant "transformation of the incarnation into a new life" or something along those lines. As such the passage seems more supportive to me of retrojection, the original point.

ted

Amaleq13

Because you made it up for no reason except to support your conclusion. Lacking a basis in reality makes it ridiculous. You are, of course, free to show that there is a basis in reality for it. Get back to me on that.

If he had, you would have an arguable basis for including it.

That wouldn't really be an equivalently specific factor.

Probability estimates for an individual are made more reliable by basing them on groups with a greater number of factors relevant to the individual. Take a look at this website on Bayes' Theorem and read through the first example given. The analogous guess is whether J.Doe died in 2000. The first statistic used is the general death rate and this is analogous to the general common/uncommon "statistic". When they change the focus to a more specific statistic (ie death rate for senior citizens) what happens to the number? There is a tremendous leap from .008 to .08! Why? Because a more specific group was being considered that was more directly relevant to the individual in question! Increased specificity of the factors = increased reliability of the prediction about any given individual described by those factors! An analogous change for our discussion would likewise focus on a smaller group that was more specifically similar to Paul.

The more similar the considered factors for the group are to the individual, the more reliable will become the predictions about the individual. This is just common sense. The more like the group you are, the more like the group you can be expected to behave. The "group", with regard to 'apo', has no defining characteristics except the use of the word in question!

There is nothing intuitive about this claim. It is just wrong. The greater the correspondence between a given individual and the population measured, the greater the reliability of attributing the group results to the individual. We are dealing with a "statistic" that describes a completely undifferentiated population except by usage. It is as general a description as is possible and, as a result of that generality, offers the lowest possible level of reliability in predicting the specific behavior of any individual.

I'm not sure why you would think that since I've already given a couple examples of specific factors that I would consider as providing a much more reliable guess as to Paul's intent.

If they don't involve actual numbers and don't involve enough specificity to make reliable predictions about a given individual, yes. They serve no purpose except to create the illusion of support for preferred conclusions.

You acknowledge that it doesn't tell us much at all but insist that we should rely on it anyway because 100% certainty is impossible? That's ridiculous. We should rely on it anyway because we don't have more reliable numbers available? That's ridiculous, too. Just because it is the best you can offer doesn't make it enough.

I completely agree that we cannot determine, with any degree of reliability, what Paul intended from this word alone. I've been saying that from the beginning.

But not specific enough, as far as I'm concerned, to justify the conclusion.

I'm not excluding the application. I'm denying the reliability of the application.

Probabilities describe groups but you can certainly be more specific in defining your group and the more specific the group, the higher the reliability of predicting the behavior of individuals in that group. You are using the most general of terms (common/uncommon) without any numbers at all and expecting me to accept your conclusion as reliable.

And the more specific the group, the more correlation there is between the group's behavior and the behavior of the individual. I realize I'm repeating myself but you seem to be missing this rather crucial point.

That would make the uncommon use group pretty darn specific, so I would think your reliability would be pretty good. I doubt anyone would describe that minority usage as "uncommon", though. "Virtually unknown" would be more appropriate.

TedM

I had another post, but am replacing it with a response to you. Maybe we are talking past each other. I hope to clarify here.

Since we are dealing with math, I see no requirment other than mathmatical consistency. The numbers can be made to work, and there is no question in my mind that the math is duplicated all the time in real life, though certainly not in this made up example. No I can't find such an example (not sure where I'd even look) but it should not make any difference for illustrating the mathmatical principles.

I totally agree. However, we don't know what is relevant to Paul, do we? You've given some factors that might be and likely are, but in the end you may miss the most important factor. General use just might be the most relevant factor to Paul's own use. It is for many if not most words people use. You could guess that someone in Paul's age range of educational background or place of origin is more likely to use 'apo' the same way as Paul does but it is a guess. What if Paul was writing to appeal to a younger audience? What if Paul had a particular education we are unaware of? What if Paul's place of origin that mattered most was a 3 day stay down the street at his uncle's house--an uncle from Egypt?!? We can't know what raises the likelihood of any given individual's personal decision to be reflected in a group percentage, without asking the individual.

Sure, as we get older we are more likely to die. Duh! This is not a good comparison because it a known factor is used to determine the more specific group (ie we are more likely to die as we age). This is bassackwards. This works great when you know what is relevant to the individual. The difference here is that this is a theory borne out by known results. We don't know the results for Paul. We don't know what Paul means by 'apo'. And we don't know what is specific to him, so we can't define the group he should be in. We can try and guess, but we can't know and there may always be differences. As my example illustrates, some elements of specificity that seem to be more related to the individual in question (#5 Gerard) are in fact less related. Your example above works great if we know what is relevant. We don't. That is my point.

Yup, and as far as we know that is MORE relevant than other characteristics you have suggested which are no more than educated guesses.

NO, it offers a lower (not lowest) level than is possible if we knew of more relevant (not specific) information about Paul. We don't. It's the best we can do for certain. Guesses about relevancy could make the pcts less reliable.


And they seem reasonable. But, they are guesses which could actually steer us further from reliability. We know that Paul is a member of the population of general usage. We can possibly determine actual pcts for that group. We don't know that Paul's membership in the groups you have suggested is more relevant to him personally, and as far as I know we cannot determine the actual pcts for those groups anyway.


This doesn't address the reliability of the general usage (ie how reliable it is). It continues to assume the relevancy isn't strong enough. Again, if general usage was 99.9999% accurate, I'm not sure you'd be saying it is not relevant. I see you address that below.....


We need to find out more clearly how the word is generally used, I think.


Well, it would make the uncommon use group small, but I don't see how you would say that is relevant to Paul since it seems to me that you think factors such as educational background are the more relevant factors. You seem to be taking both sides here: Reject the general usage unless it becomes so common that we can redefine the uncommon group as somehow being 'specific'. Or do you mean the common group is somehow now 'specific'? Specific to what I don't know. To Paul? How so?


ted

andrewcriddle

This partly depends on whether 1 Thessalonians 2:14-16 is Pauline or an interpolation.

More generally, although one could argue that claims of the responsibility of the Jewish leaders for the death of Jesus are a product of later hostilities between Jews and Christians, if Jesus was actually killed by the Roman government without any type of involvement by his own people then it is difficult to see what in early Christian teaching would have caused serious hostility by the Jewish authorities.

The main possibility for conflict unrelated to responsibility for Jesus' death, would seem to be that Stephen's speech in Acts 7 is based on early tradition, and that some pre-Pauline Christians argued that with the coming of the Messiah the temple cultus had become dispensable.

Andrew Criddle

NOGO

Hi GakuseiDon,

Thanks for the feedback. I was keen on getting your perspective on this.

Actually my point was more to say that I expect Paul to claim Jesus as one of the sources of revelation. Of course it does follow from this that Paul would also be expected to use Jesus' teachings but I make the distinction anyway and I am sure that you can appreciate the nuance.


As a non believer I think that all the messages are either directly from scriptures or inspired from scriptures. I have a good idea of where the Lord's Supper comes from.

Are there any other?
I can't answer offhand but you must also realize that some are not as obvious as others. For example the bit about the third day.



I am a bit astonished at this argument.
Paul was a believer, right?
He believed that Jesus was God, right?
How can Jesus have said or done something contrary to scriptures?
I conclude that Paul could not possibly have been afraid to talk about Jesus' life and teachings.

In fact one of my points is exactly that. How could Paul fail to show that Jesus was the one prophecised in scriptures.

To be continued ...

NOGO

I like your "partly because there may have been no other information around".
That is precisely the point of this thread. There was no other information around. Jesus' life and teachings found in the Gospel was also reconstructed from scriptures. If you do not believe that all of it came from scriptures then you must believe that most of it did.

So what did Paul know about the human Jesus if human Jesus there was.

a) son of David
b) crucified

Maybe you can help me here. I can't think of anything else. I can't even say "crucified for our sins" since that probably came from scriptures as well.

The problem with this is as follows:
how did they know that this man was indeed the Christ which they saw in scriptures?

From the Gospel perspective how did things work out?
Initially there was only a handfull of people who knew Jesus as a man and then witnessed his resurrection.
Did these people try and convince others by quoting scriptures?
OR
Did they try and relate Jesus' life and teachings to scriptures in order to show that he was the Christ?
OR
Did they go around saying that this man resurrected and therefore he must have been the Christ ... which would have gotten them nowhere.

Do you see my problem?

The lack of details about Jesus does not point to a man who "accomplished scriptures" and started a new religion.

Amaleq13

I agree but it must be consistency with reality not consistency with your preferred conclusion.

"[D]uplicated all the time" but you can't find an example? It isn't just that your numbers are fabricated, it is your poor choice of factors as well.

We can't describe Paul? By "relevant to Paul" I'm saying the more specific, common attributes the group shares with Paul, the more reliable the guess. The hope would be that we would find a common attribute that obtained something approximating your theoretical 99.9%. I would think education would be a great place to start.

That isn't how it works. You look at the data according to these variables. What does the percentage look like when we break it down according to educational background? Where is Paul on that graph? Same with place or origin.

Sure. You don't know what the numbers will suggest until you look.

We can't "know" but we can certainly improve the reliability of our guess.

What is relevant is what correlates with a high percentage of one choice over the other. You won't know until you look at the numbers associated with a given factor.

We wouldn't know until we looked at the numbers for 'apo' usage.

Yep and I can't imagine a less informative fact.

Yup.

Right. Reject the general usage numbers as insufficiently specific to make reliable predictions unless they are so skewed as to become predictively specific. We are looking for factors that break down into more indicative categories than "common"/"uncommon".

It would be specific to one particular outcome over the other by a significant degree.

TedM

Since this is your position, I'll try to make my points without appeal to a made up example, below.

We can describe Paul, but we don't know what groups are more specific(implied 'relevant') to Paul because we don't know what 'the group' that more frequently uses 'apo' a certain way is. If we KNEW that those with an educational background similar to Paul's used 'apo' one way 95% of the time, then that would be both specific and relevant to Paul. And, a good measure. But, we don't know that is the case at all.

You suggested earlier that it would be relevant to see how Paul uses 'apo' elsewhere. It sounds like a reasonable approach. However, what did you find? An example subject to interpretation. Paul's own use elsewhere didn't help.

We simply don't have the samples you would like, and that MIGHT provide more specificity and relevancy. Until we have them though we can't even know if there are ANY more relevant, specific groups. It very well may be that the general use of apo is the most relevant factor that is specific to Paul.


I agree. But we don't have the data to know if these variables are relevant or not.


Exactly. Maybe we are saying the same thing.

I thought your original rejection of 'general use' was based on the idea that it was a group without characteristics that defined Paul more clearly. I thought it had nothing to do with the degree of common usage for that group. It seemed to me that if I said 65% of the general use was one particular way, and there is a 65% chance that Paul's use was the same way, you would have disagreed with me based on what you were saying. Was that NOT the case? Here is what you wrote that led to this expanded discussion:

I think you were incorrect about the part I bolded, which is your example to invalidate the value of a general statistical description of the whole population.

I thought that because you rejected the value of the general group by implication you once meant 'specific' in the sense that it was a subset of the general group of people, and now you mean it in the sense of being more predictive than the word 'common' would seem to indicate to you, ie a high pctg of likelihood. It sounds like your position has changed, but maybe I've misunderstood them.

In any case, if I read your current position correctly, finding out just how common the general use was IS something that would be of value, as it might reveal enough statistical significance to render the need for a subset of the general group unnecessary.

I don't know if this entire discussion has been helpful to you or not. I'm not sure it has been helpful to me.

ted

Amaleq13

We don't know until we look at the numbers. That's the whole point of trying to get more specific, Ted. Even if we had numbers to define "common" and "uncommon", they wouldn't be very reliable predictors of Paul's specific behavior unless they were so skewed (eg your 99.9% figure) as to make this broad division definitive. As I said from the very beginning, that extreme figure doesn't appear to be likely but even a 70/30 split isn't all that specific. It really depends on how large the total population is and what the actual numbers are.

It is what you would need to establish your case. If someone is familiar enough with the usage to call one "common" and one "uncommon", they could also consider the more specific factors of the authors.

Nope. What is your point? That we shouldn't have considered it? You don't know what you've got until you look at the numbers, Ted, and the more specific the numbers the more reliable your "look".

We don't any numbers at all. I'm well aware that we don't have any numbers. I'm saying they are what is necessary to support your claim because, right now, you don't have anything but a general claim that one use is more "common" than another. That really isn't much to go on, is it? It really doesn't make sense to say "Hey, that's all we've got so let's pretend that this is the most we could possibly know and just go with it!", does it? I mean the only reason anyone would engage in such an approach is to support a conclusion they prefer and avoid any possibility of evidence to the contrary.

That would be very unfortunate since it isn't very specific and, therefore, isn't a very reliable predictor.

"Yes" to the first, "not exactly" to the second. The extreme statistic you were using as an example would make these broad categories pretty specific since so few would be expected to choose the minority usage.

A low difference between factors suggests this breakdown is not a very reliable predictor of individual behavior. High difference = good predictor.

I rejected the reliability of a general group statistic that approached 50/50. My use of "specific" has been consistent but is applied to different contexts. I'm suggesting we would need to focus on factors that are specifically relevant to Paul and look for percentage differences that are specifically predictive (ie have a high degree of reliability). Any factor that is specific to Paul and obtains a high percentage of one particular use enhances the reliability of our prediction.

I've held with the former from the onset of this discussion and provided clarification on the latter when it became clear it was required.

I think it is important to understand how to interpret statistics because they can be so easily manipulated to serve a preferred conclusion.