The two books under review here are remarkably similar, although (it would appear) entirely independent of each other.Footnote 2 Further, I must say right at the start that, in many respects (perhaps the most important respect), I am entirely in agreement with many of their conclusions, and their manner of reaching those conclusions. So maybe this ‘critical’ review is not critical enough!
- Both authors are English; and both books come across (to me) as very ‘English’: the argument is cautious, fully acknowledging caveats, never overstating the case, etc.
- The prime issue addressed in both is the relationship between Thomas and the Synoptics.
- Both argue for a broadly similar conclusion – that Thomas is dependent (whether directly or indirectly: see below) on at least Matthew and Luke.
- Both reach this conclusion via an almost identical methodology: seeking to identify possible elements of Synoptic redactional activity in Thomas as an indication of Thomas’ dependence.
- Some (but not all) of the evidence cited is common, e.g. Th 5 / / Luke 8:17; Th 14 / / Matt 15:11; Th 31 / / Luke 4:24. (However, other examples do not overlap.)
- Both are reticent about relying too much on theories about Q (perhaps surprisingly, Goodacre less than Gathercole).
- Both see Thomas as potentially mystifying, not necessarily aiming at clarity, but demanding hermeneutical effort on the part of the reader.
- Both emphasise the importance of taking seriously the evidence of the Greek fragments, and indeed both at different points use the Greek evidence closely.
- Both offer powerful critiques of the older form-critical ‘norms’, or ‘laws’, about the development of the tradition, both effectively denying that such rigid ‘laws’ have any validity at all.
- Both address in detail the more general arguments of others defending Thomas’ independence, e.g. on order, the apparent lack of ‘much’ agreement with the Synoptics, the possibility of assimilation in the textual transmission, etc.
- Both also offer good arguments referring to the possibility of being misled by the phenomenon of the Synoptic Gospels about what level of agreement might be expected in a case of dependence. (The Synoptics may be highly unusual in the high level of agreement (in wording and order) they display.)
- Both emphasise the danger of polarising the debate prematurely, of simply claiming that one's own theory has ‘won the day’ and ignoring others who disagree. Certainly both authors here take on, and engage with (in my view very fairly) the views of others who disagree.
The books differ slightly in that Gathercole ranges more widely than Goodacre: a large ‘Part 1’ of Gathercole addresses the question of the original language of Q, an issue not raised by Goodacre; and in a final part, Gathercole addresses the issue of possible dependence of Thomas on Paul, and on Hebrews. (I found the section on Paul rather more convincing than that on Hebrews.) In turn, Goodacre devotes a relatively short chapter (ch. 9) to the date of Thomas, an issue not explicitly discussed by Gathercole.
Gathercole also raises explicitly the important issue of whether agreements between Thomas and Matthew/Luke might be explained by a line of dependency Th → Matt/Luke, rather than Matt/Luke → Th. The possibility is generally ignored by Goodacre who simply assumes that, if redactional elements of Matthew/Luke reappear in Thomas, this must show that Thomas is secondary.
One other difference concerns the nature of the ‘dependence’ for which both argue. Goodacre seems to envisage a much more direct kind of dependence: ‘Thomas’ ‘obtained a copy’ of Matthew and Luke and used it (fairly) directly, although probably sometimes from memory.Footnote 3 Gathercole appeals rather more to the phenomenon of (what has come to be known as) ‘secondary orality’ to explain the agreements between the Synoptics and Thomas (pp. 156–8, 224, though with some caution expressed). However, despite the occasional example of extensive verbatim agreement over around 12 words in Greek (e.g. between Th 26 = P. Oxy. 1.1–4 and Luke 6:42 par.), the level of agreement remains relatively low (or at least less extensive) in other sayings where one can test it. (At times Goodacre relies on retroversions from the Coptic into Greek, but this clearly raises methodological issues.) One wonders if Goodacre's overall theories are rather set by his model of Synoptic relationships (despite his insistence in general terms – quite rightly in my view – that one should not be misled by the Synoptic phenomenon). For instance, he uses the example of Th 26 as his parade example to argue that there is probably a measure of dependence between Th and the Synoptics. A model of some kind of ‘secondary orality’ may fit the evidence as we have it better.
As already noted, I am in agreement with large parts of what is said in both books.Footnote 4 (As both authors here are kind enough to note in footnotes, I have argued similarly myself elsewhere.) However, I note here a few questions (not necessarily disagreements) relating to what is said in the two books.
Greek Fragments
Both authors emphasise the importance of using the evidence of the Greek fragments. In one way, this is clearly right and proper: the Greek fragments predate the Coptic MS from Nag Hammadi by more than 100 years, and the text of the fragments is (probably) in the ‘original’ language of Thomas. Hence issues of possible dependence on the Synoptics, which can often depend on small details of the wording in Greek, are much better served by using (where possible) the evidence from the Greek fragments of Thomas, rather than the (derivative, secondary and late) Coptic text. In principle this is fine in general terms. However, I just wonder if there is a danger of making things too clear-cut, when inevitably, in this sort of enterprise, the situation is probably much more ‘messy’.
There is perhaps a danger of assuming too easily that the Greek fragments, by virtue of their age and language, are not only earlier and ‘better’ than the Coptic text; they are also ‘very good’ in absolute terms, so that one assumes one is almost back to the autograph of Thomas itself. My own (limited) experience of working with texts only weakly attested in MSS would suggest caution. I have worked with the Gospel of Mary (similar to Thomas in being available in one Coptic MS, with two earlier Greek fragments), and also 2 Clement (available in two Greek MSS, and one Syriac, but for the last third of the text, one of the Greek MS is not extant). In both instances, it is tempting to take the Greek evidence as primary; but in both instances caution is necessary: the Greek MSS in question are not always reliable, and the ‘later’, translated version of the text (e.g. the Coptic MS of Mary, or the Syriac MS of 2 Clement) may at some points provide a better witness to the ‘original’ text.
So too with Thomas, the P. Oxy. Fragments may be at times not very ‘good’ witnesses: they contain some clear mistakes; they are not autographs, etc. Also they have lacunae (especially P. Oxy. 655); further, suggestions about filling the lacunae may be based on parallel texts, whether in the Coptic MS of Thomas itself or on Synoptic parallels. Hence the use of this evidence in the discussion about the nature of possible dependence on the Synoptics may be (a) no better than using the Coptic text, and (b) could contain an element of potential circularity.
So too, I wonder if anyone has ever given possible credence to a theory of textual assimilation at the level of the Greek fragments. It is a standard part of the defence of Thomas’ possible independence to appeal to the possibility of textual assimilation to the New Testament text by scribes or transmitters of the Coptic text (usually, it is suggested, in Coptic). But might not the same also apply to cases (quite rare) of very close verbal agreement in the Greek fragments? For instance, the phenomenon of the close verbal agreement between Thomas 26 = P. Oxy. 1.1–4 and Luke 6:42 is the exception rather than the rule. Yet this is used by Goodacre as his parade example to show that there is a prima facie case for accepting the theory of some sort of dependence – between probably the original Thomas and the Synoptics.
In relation to the two books separately, I raise just a few points in relation to each.
Goodacre
‘Missing Middles’
On a number of occasions, Goodacre appeals to the way in which Thomas seems to take up Synoptic traditions, but in doing so makes something of a ‘mess’ of it. He refers to how Thomas regularly seems to omit elements of the material, creating a number of ‘inconcinnities’ which, as such, reveal the secondary nature of the Thomas version. And he devotes a whole chapter (ch. 7) to what he calls ‘Missing Middles’, examples where Thomas regularly misses out the central part of a tradition, at times making a slight mess of things as elements in the closing part then make little sense.Footnote 5
In one way, the argument is appealing; but in another it does raise the question of just how ‘sensible’ or ‘coherent’ one can/should expect an author to be. Goodacre's ‘Thomas’ comes across as just slightly incompetent, making a mess of things, leaving loose ends hanging and unexplained. Further, is there a ‘form critical law’ being quietly presumed here, whereby traditions always progress from the fully rational and ‘complete’ to the slightly disarranged and/or incomplete? Is there perhaps an argument that might go in a rather different direction and say that we should presume, as the default position, that the extant text (of Thomas or whatever) does make sense and that, if we can't see it initially, it is up to us to try a bit harder?Footnote 6
Thomas’ Use of the Synoptics
A second issue is raised by Goodacre's final chapter (ch. 10: ‘Secrecy, Authority and Legitimation: How and Why Thomas Used the Synoptics’) where he deals with the issue of why Thomas might have included all this Synoptic material. Goodacre argues that one possible answer might be that this was a way of providing authentication for Thomas’ picture of Jesus: the Synoptic material reassures the reader that this really is Jesus; and this then opens the way for the more specifically ‘Thomasine’ material which comes in the non-Synoptic material. The Synoptic-type sayings are then the ‘necessary baggage’ (p. 180) to make Jesus sound familiar. Hence too this might explain why there are (virtually) no parallels to Johannine materials: John's Gospel had not yet attained widespread recognition and hence a ‘Johannine-sounding’ Jesus would not have commanded such universal recognition and respect.
In some ways this is not impossible; and indeed I argued similarly in relation to a (much smaller) section of the Gospel of Mary.Footnote 7 But on the other hand, the overall theory seems to me to run into some difficulty explaining the size of the material involved. It is a slight caricature (but perhaps not entirely unjustified) to say that, for Goodacre, the Synoptic material in Thomas loses significance: Thomas seems interested in it only to establish rapport with his readers and convince them to accept more readily the ‘real’ message he wants to get across – in the non-Synoptic material.
This is not totally dissimilar to the model of the older-style redaction criticism, which focused on the redactional elements in the Synoptics as the primary evidence for an evangelist's ‘theology’ (cf. e.g. Conzelmann on Luke): traditional elements (where an evangelist took over a source unchanged) tended to be ignored. Such a model is widely criticised today as being too one-sided: all the material which an evangelist includes is potentially important for determining a writer's point of view and a decision to include a tradition unchanged is potentially just as important as a decision to make a positive change to a tradition.
So then here, simply saying that the Synoptic material establishes rapport with the readers seems rather ‘low-grade’. Does not the decision to include this material suggest that the author found it positively valuable for what s/he wanted to say? (To be fair, Goodacre does at times seek to show this, for example, in the way in which Thomas sometimes glosses Synoptic traditions.Footnote 8) Further, would a policy of trying to establish rapport, and to make his Jesus sound ‘authentic’, really explain why quite so much of the material is used in this way? It is one thing perhaps to frame a major presentation of the teaching of ‘Jesus’ by an ‘authenticating’ context (as may be the case in the Apocryphon of John with the initial reference to the Matthean resurrection scene), or adding a few sayings at the end of a long block of the Saviour's teaching in the Gospel of Mary; but still the bulk of the material remains the author's own. Is there a comparable parallel to an author using such a high proportion of the material s/he does include for such a purpose? Given the relative amount of material involved, should we not first presume that this material does cohere, and contribute to, the author's overall message(s) and/or ‘theology’, and not (just) regard it as something of a vestige of earlier tradition that is almost to be discounted?
Gathercole
The Language of Thomas
Gathercole devotes the whole of a long first part of his book to an elaborate defence of the theory that the original language of Thomas was Greek, engaging with others who have tried to show a Syriac, or an Aramaic, original for Thomas.Footnote 9 As with the dependence issue, I am fully in agreement with the conclusions, and if it fair to appeal to ‘majority views’ on this, so would most others.Footnote 10 Gathercole does an excellent job in pointing out the difficulties, and the inconclusive nature, of many of the arguments which have been proposed.
I wonder though whether it is too easy at times to point out the difficulties and/or inconclusive nature of arguments as a way of ‘proving’ one's own case. I take just one example.
Gathercole refers (p. 44) to the phrase ‘taste death’ in Th 1, which some have argued might reflect a Semitic Vorlage. His argument is to say that, as the phrase occurs in the Greek texts of the canonical gospels (Mark 9:1 and pars; John 8:52, as well as Heb 2:9), there is no need to posit a non-Greek Vorlage for Thomas: such is ‘unnecessary’ (p. 44). But this simply pushes the argument one stage backwards (or sideways): how is one to explain the wording in the canonical gospels? (And is Heb 2:9 reflecting gospel usage?) Could not this diction be used to argue for a Semitic Vorlage for the canonical evidence as well? If this is excluded, is the argument here, appealing to the precedent of the Synoptics, already presupposing a relationship between Greek texts?
A great deal of Gathercole's argument here reduces to a ‘not necessary’ kind of conclusion. This is of course probably fair. But is the demand that the opposite side has to show that their theory is ‘necessary’ perhaps raising the bar which one is demanding of others unfairly high? No kind of ‘proof’ with mathematical finality can be reached in discussions of these issues: we can only work with probabilities, global explanations of the evidence which seem to generate the least number of difficulties.
I do not wish to question Gathercole's overall conclusion here (I am happy to accept that Thomas was originally a Greek composition), but I wonder if the rhetoric is at times claiming slightly more for the argument than is perhaps justified.
Parallels between Thomas and Synoptic redaction
Gathercole (like Goodacre) seeks evidence for Thomas’ dependence on the Synoptics via places where Synoptic redactional elements may reappear in Thomas. In general terms, I fully agree with the methodology. But what precisely can count as valid evidence here? At one or two points I had some reservations.
At one point, Gathercole claims that Th 13 is parallel to, and agrees with, Matthew's redactional addition to Mark in the blessing of Peter following Peter's confession. Indeed Gathercole seems to see this as quite a key piece of evidence, showing not only that Thomas presupposes Matthew's redactional work (and hence might know some traditions in their post-Matthean form), but also that Thomas here betrays knowledge of the existence of Matthew's written gospel as such (full discussion on pp. 169–78).
The text in Thomas concerns the famous interchange where Jesus orders the disciples to tell him who he is like. Two replies, by Simon Peter (‘you are like a righteous angel’) and Matthew (‘you are like a wise philosopher’) are followed by the reply of Thomas himself, saying that he is ‘wholly incapable of saying who you are like’. Gathercole argues that the structure of the scene is closely parallel to that of Matthew/Mark's account of Peter's confession at Caesarea Philippi, and that the commendation of Thomas is closely parallel to the famous blessing on Peter in Matthew 16:17–19, Matthew's redactional addition to Mark. Further, Gathercole assumes that (a) both the first two replies are ‘clearly wrong’ (p. 169), and (b) the choice of ‘Matthew’ as one of the speakers of these inadequate confessions relates to him as the authority behind (or author of) Matthew's Gospel. Hence Thomas here shows opposition specifically to (and hence direct knowledge of) the written gospel of Matthew.
Both these claims are at least contestable. It is not so clear that Peter's and Matthew's responses here are ‘wrong’ (or even necessarily inadequate).Footnote 11 But it is equally not clear that the name of Matthew here acts as a surrogate for the written gospel text bearing his name.
Further, despite the claim that there are ‘two redactional features evident in Matthew [which] appear (mutatis mutandis) in Thomas’ (p. 177: both draw attention to the sources of Peter's/Thomas’ knowledge, and both note this as the occasion for the ‘consecration of a figure who is to be the foundation of Jesus’ community’), the fact remains that any ‘agreement’ here only exists at some level of abstraction. There is no verbal agreement at all. The name ‘Thomas’ is not ‘Peter’; the blessing on Peter in Matthew 16 is couched in totally different words and terms from any ‘blessing’ (no cognate of the word is used) pronounced on Thomas (‘you have become intoxicated from the bubbling stream I have measured out’).
This then raises a methodological question of when one can say that there is ‘agreement’ between the text of Thomas and elements of MattR or LkR. Most would probably insist on some kind of verbal agreement (with of course allowances having to be made in cases where Thomas is extant only in Coptic). But without any such verbal agreement, is a broader ‘similarity’ enough to establish the case? Gathercole's reference to two redactional features of Matthew ‘appear[ing]’ in Thomas is admittedly qualified by a ‘mutatis mutandis’; but some might feel that too much has been ‘changed’ here to make the parallel into an ‘agreement’.Footnote 12
A similar problem may arise in relation to an alleged agreement between Th 44 and Matthew 12:32, as another possible example of an element of MattR reappearing in Thomas. Gathercole sees this as such an example, comparing the extra phrase in Thomas that whoever blasphemes against the Holy Spirit will not be forgiven ‘either on earth or in heaven’ and the extra phrase which Matthew adds to Mark here, saying that anyone who speaks against the Holy Spirit will not be forgiven ‘either in this age or in the one to come’. Gathercole makes a case for saying that Thomas’ formulation is intelligible as a rewriting by Thomas; but the fact remains that any ‘agreement’ here is at the relatively general level of a ‘structural similarity’ (p. 181) which does not however reach the level of verbal agreement.
In general, Gathercole's array of evidence is considerably more extensive than Goodacre's; but I am not entirely convinced that all his examples are quite as persuasive, especially when they turn out to be ‘agreements’ only at a level of some abstraction.
Conclusion
Both authors are to be warmly thanked for their books. The two books together provide a powerful case for the theory of Thomas’ dependence on the Synoptics, even if at times there may be still questions to raise and discuss.
