Hi Bob-

Just for the record, I see from the changed subject line on your post that

you thought you were replying to my remarks, but you in fact were replying

to Kurt, instead.

Rick

||-----Original Message-----

||From:

gthomas@yahoogroups.com [mailto:

gthomas@yahoogroups.com] On

||Behalf Of Bob Schacht

||Sent: Saturday, May 01, 2010 2:08 PM

||To:

gthomas@yahoogroups.com
||Subject: Rick Re: [GTh] Re: The Scientific method

||

||

||

||At 11:09 PM 4/30/2010, kurt31416 wrote:

||

||

||

||

|| Hi Bob, thanks for the thoughtful reply,

||

|| My point is a general one that if we can identify the parallels, it

gives us

||the ability to put a mathematical handle on it, to be able to test

theories' ability

||to predict experiments. One hypotheisis would be that the parallels in

Mark are

||random.

||

||

||What do you mean by random? Do you mean that the composer of a saying in

||GTh decides at random whether or not to include a Markan parallel? Or that

||existing "sayings" were selected for inclusion in the list of sayings at

random,

||with respect to whether or not there is a Markan parallel? Please clarify

what

||you mean by "random" in this context.

||

||

||

|| That one you can get a hard math answer to, by looking at that big

Mark

||gap between 66 and 99. Less than 1% the way the Jesus Seminar broke apart

||the sayings, about 1.2% by saying, would have a gap that large or larger

if you

||rolled dice. You can take that to Las Vegas. A universal system to test

the

||randomness of the distribution of things by looking at the largest gap. I

posted

||it here before, I'll go get it. It's never been published elsewhere that

I'm aware

||of.

||

||

||Well, then chances are(!) that your method is poppycock. If no one has

done it

||before, there's probably a reason. Or perhaps your literature search has

not

||been effective.

||

||The better parallel is not rolling dice, but tossing a coin, in which,

say, "Heads"

||means a Markan Parallel, and "Tails" means no Markan parallel.

||"Gaps" of the kind you point to are not uncommon in coin toss sequences.

But

||you can't just look at the gap; you have to look at the entire sequence.

For

||example, a "gap" of 33, such as you point to, would be very rare in a run

of 33,

||but in a run of, say, a million, the odds are much greater of finding such

a

||"gap" somewhere in the sequence. The mathematics of this have been the

||object of many studies. Avoid reinventing the wheel, when this particular

wheel

||has been well studied.

||

||There are also standard statistical tests that could apply to this

situation,

||regarding Ordinal Testing (see, for example,

||

http://www.quantitativeskills.com/sisa/statistics/ordhlp.htm
||<

http://www.quantitativeskills.com/sisa/statistics/ordhlp.htm> ) This

requires

||that the sayings in GTh be divided into two groups: those with Markan

Parallels,

||and those without. It is an ordinal test because the sayings in GTh are

||presented in a certain order that is fixed. These are usually described as

runs

||tests because the sayings in GTh can be described as a "run" in which each

||saying can be represented by a letter indicating whether or not each

saying has

||a Markan parallel. So, for example, if M designates a Markan parallel, and

X

||means no Markan parallel, GTh could be represented by a string such as

||XXMMXMXXXXMMXMXM....

||

||But you still need to play close attention to exactly what your hypothesis

is.

||

||

||

|| You'd think someone had already done the math to calculate the

||probability of a gap of a certain size for this kind of case, but

apparantly they

||never have.

||

||

||Baloney. See above.

||

||

||

|| It's harder than it looks. By saying, I took it as far as proving

it was

||smaller than one out of 40. And experimentally, by having a computer roll

the

||dice hundreds of thousands of times, which consistantly produce a result

of

||about 1 in 80. (Less than 1 in 100 by partial sayings) I'll go bump it

with brief

||commentary.

||

||

||You're barking up the wrong tree.

||

||

||

||

|| There's plenty more strange about Mark. Ratio of Mark to Q before 66

is 7

||times higher than after #66. (Five Gospels parallels, Q defined as Matthew

plus

||Luke, but not Mark, per saying.)

||

|| One could defend the position that before #66 is the Mark part of

Thomas,

||and after #66 is the Q part.

||

||

||Now you're getting into more interesting territory. You're dabbling in

hypotheses

||about the source for *blocks* of text

||

||

||

||

|| In addition to the math, it also gives the human brain, a picture,

that we

||are very good at sorting out intuitively, so we know where to go do the

math.

||One could calculate the odds there the ratio would be at least seven times

||higher, etc, but good to know where to look. For instance, here's the Mark

and

||Q sayings in Thomas, before and after #66, according to the Jesus Seminar.

||Intuitively, does that look random?

||

||

http://www.kingdomofthefather.com/Mark-QInThomas.html
||<

http://www.kingdomofthefather.com/Mark-QInThomas.html>

||

|| The one hypothesis I want to test with it, the one that's

interesting to me

||lately, is that there was a Christianized version of Thomas, a lot like

the notion of

||Q, except Mark saw it too. But that's another whole ball of wax.

||

||

||Yes. For starters, you'll have to define what a "Christianized" saying

looks like.

||

||Always, always be clear about exactly what your hypothesis is.

||

||Bob Schacht, Ph.D.

||Northern Arizona University

||

||