Mark/Thomas Parallels (Thomas Order)
- To: GThomasCc: GPGOn: Mark/Thomas Parallels (Thomas Order)From: BruceI guess I should do something to redeem my previous crimes and shortcomings, and maybe this will help. Besides the Correlation feature on his Home Page, Stevan Davies also has a list (which is probably where I got confused) of Mark/Thomas parallels. That list is in Mark order, and anyone wanting to look at it from the Mark side of the table can find it atI took the trouble to rearrange it in Thomas order, and I will give those numbers here, in case anybody else was curious.[With some adjustments and different labelings, this is based on the Koester list, Ancient Christian Gospels p108. It has 27 sayings vs Koester's 28, but is extensively the same. I think we have found the source].Thomas/Mark ParallelsGTh Mk4b 10:316b 4:229 4:3-913 8:27-3014c 7:15-2020 4:30-3221d 4:26-2921e 4:9 (also occurs elsewhere in GThos)22a 10:14-1525 12:3131 6:433b 4:135 3:2744 3:28-2947c 2:21-2248 12:22-2355 8:3462a 4:1165 12:1-866 12:10-1171 14:5879 13:1799 3:31-34100 12:13-17104 2:18-20111a 13:31113 13:21It was obvious from the original (Markan-order) list that GThos favors certain parts of Mk. The interest of seeing the data in Thomas rather than Mark order, or so it occurs to me, is that we can tell if there is clustering in Thomas: if one part of Thomas relates more extensively than another to Mark.[Always assuming that this list of parallels is adequate for the purpose, and I don't take that up here. Koester points out that there is some Mark/Q overlap, which for me is a no-no; Q, if defensibly defined, is what Mt and Lk both contain, but didn't get from Mk because it is not present in Mk. For those who wish to make the adjustment, Koester identifies "as many as seven" of these parallels - he actually marks eight - as having Q overlap, namely GThos 5, 6 [one item in Davies], 20, 31, 33b, 41, 44, 62a, 55b, 71].Or we could do this if there were an adequate statistical test of clustering available. What do our statisticians say upon this matter? How far does the above profile depart from what could be reasonably thought of as a random distribution of Mk contacts across the 114 Thomas passages? In terms of runs, what is the longest run of GThos passages without a Markan contact that we would expect, from a random scatter of the material? How many consecutive GThos passages could have Markan parallels, without departing from what is attributable to random distribution?Parable of the Sower all over again, but this time (hopefully) with statistical aids to interpretation.BruceE Bruce Brooks
Warring States ProjectUniversity of Massachusetts at Amherst
> What do our statisticians say upon this matter? How far does theSince you joined only recently, Bruce, you missed the big discussion we
> above profile depart from what could be reasonably thought of as a
> random distribution of Mk contacts across the 114 Thomas passages?
> In terms of runs, what is the longest run of GThos passages without a
> Markan contact that we would expect, from a random scatter of the
had in May on these very issues. Perhaps the most productive course of
action would be to review those messages before going on with this?
(See also Rick Hubbard's postings that same month on "parallels".)
Thanks for your suggestion to check the backlist. Among the rest, I did note
your 13 May 10 message (#9378) declaring the topic of computational analysis
closed. Verdict accepted and apology herewith tendered; I will continue my
own investigation elsewhere.
For those equally new to the GThomas list, but interested in reviewing
previous discussion on the open GThomas archive, there were three principal
threads. "Probability of No Mark Parallels For 29 Sayings in Thomas" began
with #8594 (1 Feb 09, Rick Van Vliet), ran for quite a spell, faded out, and
then resumed on 2 May 2010. "The Markan Gap in Thomas" began with #9356 (11
May 2010, Mike Grondin). "Parallelomania" (#9352, Rick Hubbard) was perhaps
usefully balanced by a later caution about "Parallelophobia."
Among many other things of interest, I liked Rick Van Vliet's demonstration
that the Wald-Wolfowitz Runs Test fails ingloriously to detect manifestly
nonrandom distribution of A among B; I recall that Willi Feller made an
equally ludicous mistake about the pattern of V-bomb hits in London (which
he thought random, though to her credit, Florence David, who was on that
ground and analyzing the data at that time, got it right: bivariate
Gaussian). My hope to find the "Jesus Seminar" list that people were talking
about (though the actual specifications seem to have shifted somewhat during
the conversation) was at last seemingly gratified by Mike Grondin's posting
(the code seems to refer to The Five Gospels), which varies significantly
from the Koester list previously mentioned. That long a gap in Mark-related
Thomas passages (on the chart, it is GThos 68-98 inclusive) is definitely
merkwürdig, but there is no such gap in the Koester distribution, which
tests out (I am elsewhere informed) as indeed insigificantly different from
a random result. Koester recognizes Mark similarities at Thomas 71 and 79,
which (if correct) plugs the gap statistically.
Of course the assumption of Thomas literary unity underlying the various
statistical tests employed in this discussion may not be correct. If there
are internal compositional (not just formal) boundaries, the whole
statistical map would have to be redrawn.
Next thing to do along this line, I should think (I make this remark for the
GPG small audience, such matters being precluded for GThomas), is to
examine, and adjudicate, the differences of opinion as to what is a "Markan"
contact. Tim Lewis directed attention some time back to Tuckett's article
(Thomas and the Synoptics, NT 30 #2 (April 1988) 132-157), which spends some
time demonstrating that, inter alia, the Mustard Seed mention in Thomas 20
is clearly closer to Mark than to the later Second Tier Gospel variants. On
the other hand, Tuckett finds Thomas 5 and 16 to be closer to Luke.
Directionality perhaps still to be determined.
And so on.
E Bruce Brooks
Warring States Project
University of Massachusetts