Date: Tue, 23 Jun 1998 20:50:15 -0400 From: E. Bruce Brooks To: crosstalk@info.harpercollins.com Subject: Re: The Thomas/Q Hypothesis [ The following text is in the "ISO-8859-1" character set. ] [ Your display is set for the "US-ASCII" character set. Some ] [ characters may be displayed incorrectly. ] Topic: The Thomas/Q Hypothesis In Response To: Survey Results and Subsequent Comment From: Bruce Thanks to Mark Goodacre for his several clarifications of the experimental design of the Q/Thomas project. I would like to suggest some inferences for possible general discussion. Truth In Advertising Statement: I haven't the pleasure to believe in Q (which is presumed to exist for purposes of the Q/Thomas Project), and am not presently convinced about the role of Thomas, though I am perfectly willing to see it tested, and will here take the Q/Thomas results as one such test. My synoptic order is: Mk > Mt > Lk. Since, as compared to the 2-Source Hypothesis (2SH) with Mark and Q as the two first-tier texts, this model has only 1 first-tier text, I suppose it might conveniently be called the 1GH or 1-Gospel Hypothesis (on Synoptic-L, I believe following Stephen Carlson, this is instead called the FGM or Farrer/Goulder model, which term however is not symmetrical with the names for most of the other hypotheses, and ignores the previous work of Ropes, etc, not to mention the later clarificatory contribution of M Goodacre). Original Intent: As most recently restated by Mark following discussion by Mike Grondin, the project consisted of examining 57 Thomasine sayings with Q parallels [or, to Q skeptics, Thomasine sayings with parallels in both Mt and Lk) to see which of the following options were true (some numbered options for less satisfactory results omitted here): 2. Those sayings in Thomas that correspond to Q are more like the Matthean form than the Lukan form, 3. Those sayings in Thomas that correspond to Q are more like the Lukan form than the Matthean form with some participants expecting one result, and some the other. As summarized by Bob Schacht on 20 June (with one addition error* corrected), the results for the 57 sayings were: Option 1 (Spurious example) 20 Option 2 (Thomas more like Mt): 7 Option 3 (Thomas more like Lk): 16 Option 4 (Thomas like both): 10* Option 5 (Uncertain result): 4 Total: 57 My personal observations follow. NOTE 1: Despite presumably being preselected for the purpose, 20 or 35% of the 57 passages proved irrelevant for the stated purpose. This is a rather high number to fail expectations, and would seem to suggest that the presumptions behind the experiment might need another look. Having noted this, I leave aside the 20 "spurious" examples and discuss below only the 37 which gave a result, including a balanced or uncertain result (Options 4 and 5). NOTE 2: Some decisions for Option 2 or 3 (preponderant resemblance in one direction) might easily have been adjudicated as Option 4 (resembles both). To take an example from David Cedor's admirably clear report (17 June), rearranged to put Mt ahead of Lk: Mt: Congratulations to the poor in spirit! Heaven's domain belongs to them. Lk: Congratulations, you poor! God's domain belongs to you. Th: Congratulations to the poor, for to you belongs Heaven's domain. David's comment was: "Cedor's vote: 3. Th is more like Lk than like Mt. 'Heaven's domain' is like Mt, but the lack of spiritualization makes this more like Lk's version." But surely this amounts to saying that Th resembles both Mt and Lk in different respects, and the vote for Option 3 is a weighting of one common feature [lack of "in spirit'] against another ["Heaven" vs "God"]. Another investigator might have chosen the Heaven/God commonality as more important, or for that matter the statement vs direct address form, and thus voted Option 2. The objective common ground is that resemblances exist in both directions. It seems that this is the sort of situation for which Option 4 was provided. Similarly, Shane Matsumoto (18 June), ap #41, notes that the decision is more a matter of weighting, than detecting, similarities. I think these situations might be rather numerous among the decided cases, which bulk larger, but not overwhelmingly larger (23 of 37 = 62%), than the explicitly undecided ones. Without attempting to recalibrate people's results, I think that one way of interpreting the already high number of Option 4 reports is: "Thomas is frequently eclectic with respect to Matthew and Luke." This, as far as it goes, would be consistent with the possibility that Thomas is later than, not a source for, Matthew/Luke. Given the known behavior of Matthew and Luke themselves with respect to the source Mark (at least for those who accept this sequence), where the last of the three (Luke) will sometimes prefer the immediate precedessor Matthew and sometimes go back to the more remote predecessor Mark, an oscillation between sources must be acknowledged to be as much an option for the latest text as is a balanced mixing of the two. In view of the comments of critics of, say, Lukan tertiority - please use this word, and notify the OED scanning committee - on the authorial strategies proposed for Luke, perhaps it is even a *more* plausible option. I don't think there is anything in the numbers for Options 2, 3, and 4 that would preclude an inference that Thomas has drawn on Matthew and Luke. That possibility might then deserve separate study. NOTE 3: There are a number of cases where Mt/Lk are virtually identical, and only Thos diverges. Mark Goodacre (18 June, ap #34 and 36) has noted instances, and others are easy to discover. These, if gathered into a subcorpus, might make a good second project. What is the implied directionality between Mt/Lk and Thos, for these cases where the red herring of Mt/Lk disagreement is eliminated, and where simple copying ("redaction") is not available as a frame assumption? Can't put my finger on it this second, but I seem to recall it being observed that a John the Baptist saying in the synoptics is a Jesus saying in Thomas. I would suggest that in evolving traditions generally, the very strong tendency is for sayings (or even text) attributions to gravitate to the better known figure. The direction John > Jesus would be in line with that paradigm. It is much harder to figure why a Jesus saying would be reattributed to John at a later date, is it not? I would take this instance as favoring a Mt/Lk > Thomas directionality. But I would be interested to see the whole set isolated and examined. APPENDIX. I now embark on a suggestion about Thomas for which I think the present project results offer a test, and indeed a confirmation. Sorry it is at once so long and so inadequate to the difficulty of the subject. Archaeologically, we have Greek Thomas (GThos), known in fragments from Oxyrhynchus datable paleographically to c200, and Coptic Thomas (CThos), known complete from Naj Hammadi, datable paleographically to c400. The specifics of these finds seem to me to permit some inferences about Thomas that I do not recall seeing spelled out in what I have seen of the secondary literature (to which we of the Coptically challenged set are in effect limited, being unable to make use of Mike Grondin's website). Here are a few proposals. Inference 1: The Gospel of Thomas was popular enough as of somewhat before c200 that three scribally distinct copies of it (Oxy #1, 654, 655, no two written by the same person) were made in close enough proximity to turn up in the same rubbish heap. Whatever its origins, the text was thus "live" as of c200. Inference 2: The text remained popular enough between c200 and c400 to have gone into a Coptic translation. It is thus not an antique, but a current cultural object, as of the c400 copy of that Coptic translation. Thomas thus had a long future as of c200. Inference 3: The fragments of GThos do not validate the existence of *all* of CThos as of the earlier date, c200. They contain only (in terms of the CThos order) sayings 1-7 (preceded by a heading, so we know they were the first 7 sayings), 24, 26-33, and 36-39, plus 77 attached to 30. The extent of CThos is 114 sayings. That three partial copies should all be confined to the first third (34.2%) of the text is at least somewhat unlikely. A better hypothesis is that the text grew in extent (remained not only popular, as in Inference 2, but textually *active* or in process of formation) between c200 and c400. Inference 4: This is confirmed by the joining of what becomes CThos 77 to what becomes CThos 30. An isolated Oxy fragment containing only 77 would tend to disprove Inference 3, but a continuous text in which 77 is *the second part of 30* is a different matter. Looking at the matter frontwards rather than backwards, GThos 30a and 30b are split up and widely separated in CThos (as 30 and 77), but this is a matter of expansion and reordering in CThos. The provable extent of GThos remains at 39 sayings. That others followed is quite possible, given the fragmentary nature of all three Oxy papyri, but their extent is not directly knowable. See however a trial INVESTIGATION below. Inference 5: In support of the above, note a further instance of rearrangement in CThos. The sequence of sayings in Oxy 655 (I am deeply indebted to Andrew Bernhard's website for permitting some certainty here) is 36-37-24-38-39. Unlike the 30/77 juncture, this detail is not regularly noted in descriptions of GThos, including the commentary (as distinct from the prefatory matter) in the Miller Complete Gospels volume. It is not difficult to see a rationale for the sequence 37-24 (both can be said to deal with the "revelation of the inner light"), so that this sequence is quite plausibly a result of authorial intent rather than scribal error. A topic for investigation is to determine the logic of the relocating of 24 after 23 in CThos. Inference 6: Directionality. It is doubtful that dumped scraps in c200 served as a source for a scribe in c400, but the notes to the GThos translation in Miller seem to make it clear that a *text* like that behind the GThos fragments is also scribally prior to CThos. See the commentary in Miller ap GThos 30:1. Inference 7 (Qualifying the preceding): On the other hand, the text behind GThos is itself at one remove from the main stem, since it has at least one elaboration not followed by CThos (see commentary ap GThos 5:3, and note the clinching argument that the term in question is otherwise not in the vocabulary of the text). Inference 8: Elaboration and Abbreviation. Both processes are visible in the relation between Mark and Matthew (whatever direction one thinks *that* runs). Therefore neither process can be excluded, or favored, a priori for this literature in this period. A case of expansion of GThos in CThos is at 29, where (thanks again to Andrew Bernhard for permitting assurance here) there is simply no room in the unreadable part of Oxy 1 to take all the text which the GThos commentary cites from CThos. GThos has to have been shorter. PROPOSAL: These details suggest to me that a study of the evolution of Thomas between its Greek and Coptic forms might help establish a trajectory for Thomas that would be a useful preliminary for any backward projections of the earlier nature and history of what lies chronologically behind Greek Thomas as we have it. INVESTIGATION: An aspect of that study might be to attempt to determine the probable physical extent of Greek Thomas. One approach might be to look for signs of literary cadence or thematic culmination at, or shortly beyond, saying 39. Separately, and more subtly, one might ask whether the style or affinity of the text changes notably at any point in that range. It is of course difficult to quantify these terms. One highly speculative attempt, not as such worth the time of distinguished colleagues, would be to ask, of sayings in CThos with analogues in Matthew and Luke, which of the two they were closer to. That there should be any systematic difference requires assumptions which are themselves precarious, and hardly worth stating. However, we happen by a stroke of fate to have the results at hand. I note the following, of judged affinities (as summarized by Bob Schacht, and ignoring any subsequent reconsiderations or switching of votes): 4 Matthew 5 Luke 6b Luke 14b Matthew? 16 Luke 20 Luke 21 Matthew 24 Luke 33 Matthew 34 Matthew 39 Luke 41 Luke 44 Matthew 47a Luke 48 Matthew 54 Luke 55 Luke 61 Luke 64 Luke 69b Luke 89 Luke 91 Luke In other words, there seem to be two zones in the text: (1) an early one, extending at least as far as saying 48, where the affinities are quite evenly balanced between Matthew and Luke, and (2) a late one, beginning possibly as early as saying 54, where the affinities are exclusively Lukan. SUGGESTION. Not to seem to be delegating all the work to others, I will hazard the first guess that the end of GThos may have been at sayings 49-50, both of which speak of a return to the Kingdom (49) or the Light (50). The immediately following 51-52 also speak of an End, but a more Messianic and general one than that which it seems to me is envisioned in 49-50. I can see 51-52 as a later anthologist picking up the thread thematically at this point, with 51-52, but in the process rather spoiling a rather magical ending. Notice, in support, that the "rest" in [CThos] 50 picks up the theme of "rest" at the literary beginning of GThos, in saying 2 [as amended]. The so-called saying 1 is really, or so it seems to my eye, a promise about the following sayings ("whoever discovers the interpretation of these sayings") than a saying itself. In that case the conjectured GThos might have contained 49 rather than 50 sayings. Assuming GThos correctly amended, it is worth noting that the CThos counterpart saying does not end in (or even contain) "rest," but rather ends in "rule." The end of CThos, sayings 113-114, echoes this same motif of "rule" and "dominion." It may be, then, that GThos had its framing theme, emphasized at the beginning and end of that text, and that CThos broke down that pattern, rearranged and extended the material, and imposed a quite different framing pattern on the new whole. INTERPRETATION: What might this mean in terms of the history of the text? Accepting the verdict of many that Thomas may usefully be categorized as a Gnostic text, and the verdict of many others that Marcion (mid 2c) is also a Gnostic activist, so that Marcion's priorities may be relevant in some sense for Thomas, I note that the one Gospel Marcion admitted to his canon was Luke. My conjectural Thomas has a first segment of Matthew/Luke affinities, and a later segment of pure-Lukan affinities. The vogue of Luke with Marcion and his many followers (half of Christendom according to some popular accounts) might then have set a fashion for Luke in other strands of Gnosticism as well, and in due course influenced the priorities of this conjectured segment. This might have taken place at some time after c160 in Rome and points north, that is, in Marcion's homeland and the area of his great confrontation, and perhaps post c200 in points south, such as Egypt, where both copies of Thomas were discovered. The scenario is not time-perfect, but perhaps it is salvageable. Just a suggestion. I would be interested in any reactions or further explorations. APOLOGIA. And having said that, I must in courtesy add that I am compelled to be away from E-mail for about ten days, and will thus be unable to acknowledge any criticisms or suggestions that might be made. My next message to Crosstalk will be "unsubscribe," to prevent the Daily Fifty from overflowing the capacity of my University computer account. I will however check in via FindMail on my return, and can respond if needed at that time. Renewed thanks to Mark Goodacre for making this archival continuity possible. Bruce E Bruce Brooks University of Massachusetts