Clement of Alexandria and the (Canonical) Gospel of Mark (a pedantic post)

In Oxford this term we have a colloquium or reading group working through and discussing Simon Gathercole's brilliant new book, The Gospel and the Gospels: Christian Proclamation and Early Jesus Books (Grand Rapids: Eerdmans, 2022). Now this is a five hundred page book, so when in this post I point out an error on one of those pages, please (Simon!) take this in the spirit in which it is intended (straightforward academic one-upmanship) (pursuing the truth) and with a sense of proportion. (Also Simon Gathercole is not the only victim of this error; Chuck Hill made a similar error; cf. here in connection with my comment re p. 72 - this post now resolves the problem noted there.)

Anyway, here is page 4:

 

The specific problem that is obvious here is that Clement of Alexandria (like other patristic writers) barely ever quotes Mark's Gospel. The general problem is that all the figures for the canonical gospels look inflated and thus suggests the overwhelming preference of Clement for the canonical gospels. But the actual data is more complicated. Gathercole is quite clear that he gets these figures from Mutschler's monograph. He is also basically not adjudicating this data, but in his actual argument is showing that scholars differ in their assessment of the attestation of the various gospels so we need a theological content discriminator to make an argument for the distinctiveness of the canonical gospels from the noncanonical gospels. So maybe Gathercole can be slightly let off the hook; but the data as cited here is (as I will show) plainly wrong. Here is the relevant page from B. Mutschler's monograph,  Irenäus als johanneischer Theologe: Studien zur Schriftauslegung bei Irenäus von Lyon (STAJ; Tübingen: Mohr Siebeck, 2004) (it is not necessarily a problem to go to a monograph on Irenaeus' use of gospel traditions to get data about Clement of Alexandria - since Mutschler has a chapter comparing the two writers; but in this case we are being mislead because of decisions made by Mutschler) (thanks to the Bodleian Library Scan & Deliver Team): 

We can see from this that Gathercole has accurately transcribed the numbers from Mutschler's chart (although Mutschler does not compare the noncanonical gospels so Gathercole has got those numbers - which are correct - from somewhere else). So where does Mutschler get his numbers and what are they actually counting? 'Anzahl der biblischen Bezugnahmen' means the number of biblical references; but really what is going on here? Well, now it is a little unclear. There is no footnote to this chart on p. 101. On the previous page, in relation to a more general chart we learn from the footnotes in Mutschler that Mutschler claims to get his data (the raw numbers) from A. van den Hoek, ‘Techniques of Quotation in Clement of Alexandria: A View of Ancient Literary Working Methods’ Vig. Chr. 50 (1996), 223-243, and that A. van den Hoek got her data from the indices to the Stählin GCS edition of Clement of Alexandria (vol. 4, pp. 1-59) (Mutschler note 5 on p. 100: 'Die Zahlen sind A. VAN DEN HOEK, Techniques, 240f Anm. 55 entnommen, die sich dabei auf die GCS-Indices von O. STÄHLIN beruft (s. GCS Clemens Alexandrinus IV, l- 59 STÄHLIN).'). The complication here is that van den Hoek doesn't give any figures at this level of granularity (i.e. for all the different books of the Bible), so the implication would seem to be that in fact Mutschler has taken van den Hoek's idea and done his own count based on the indices in the Stählin edition of Clement of Alexandria.

Now, an interesting subpoint here is that both van den Hoek and Mutschler are aware of problems with the data they are using. van den Hoek says that 'the index of Stählin, however, is less accurate [than the TLG] since it is a vast vessel of very diverse materials that were collected over the centuries. Experience suggests that it contains too many parallels, not all of which are valid.' (van den Hoek, ‘Techniques of Quotation', 230). Since she is mostly interested in Clement citing by author's name she thinks that the combined search (TLG and the Index to Stählin) for e.g. Παῦλος will be sufficient for her needs (and that the errors are probably the same for all of his searches, so that the comparison will still be worthwhile). Mutschler has an interesting comment about potential problems where he suggests that the Stählin Index might cause the results for the synoptic gospels to be overrepresented by almost three percent (note 7, p. 100-101). 

So it is time to check out the Index to Stählin. Here is a portion of page 14

It is true, I think, that there are 278 entries for Mark, but that doesn't really matter. If we look at the first example we will uncover the fundamental problem here. The index is supposed to be listing places where literary dependence is assumable or at least probable. (But it could include indirect or secondary citations - for these you need to check the apparatus to the edition.) Entries have italics for direct citations (from Stählin GCS vol. 4, pp. IX-X). 

The first entry here is: Mark 1.3: I, 9, 18f; 64, 10; 224, 16.

So vol. 1, p. 9 lines 18-19 should be a direct citation. Here is the page:

 

What we have there is a brief citation from Is 40.3 in a form that doesn’t appear to directly match any of the NT uses (or the LXX for that matter): εὐθείας ποιεῖτε τὰς ὁδοὺς κυρίου. The apparatus refers to Is 40.3 (Matt 3.3; Mark 1.3; Luke 3.4; John 1.23) – the four places where this passage is cited in the NT. Clearly there is no direct evidence for any relationship between Clement and Mark’s Gospel at this point.

Vol. 1, p. 64 line 10 should be a passage with literary dependence that is not a direct citation. Here is the page: 

We can see that this is a general allusion to this passage. The apparatus has ‘vgl.’ i.e. cf. Matt 3.3; Mark 1.3; Luke 3.4. Again no direct evidence for Clement and Mark's Gospel. 

Vol. 1, p. 224 line 16 should be a passage with literary dependence that is not a direct citation. Here is the page: 

This passage is also a general allusion to the making straight terminology (in the context of John the Baptist) and the apparatus has ‘vgl.’ i.e. cf. Matt 3.3; Mark 1.3; Luke 3.4; John 1.23 (Is 40.3). Again no direct evidence for Clement and Mark's Gospel.

I didn't bother going any further. These indices are not a little bit problematic (van den Hoek) or three percent out (Mutschler), they are totally useless and misleading on the use of Mark in Clement of Alexandria! There is no useful data to be drawn from the total number of references to Mark in the apparatus (i.e. 182 times, as noted above). It is all ‘contamination’ from the synoptic parallels. Please, please, please, may no one ever refer to this again. 

 Actually we do have useful books on this subject.

 M. Mees, Die Zitate aus dem Neuen Testament bei Clemens von Alexandrien (Quaderni di ‘Vetera Christianorum’ 2; Bari: Istituto di Letteratura Cristiana Antica, Universita di Bari, 1970). I bought a copy to check this out. He finds 18 passages where Clement might exhibit knowledge of Mark. 

C.P. Cosaert, The Text of the Gospels in Clement of Alexandria (NTGF 9; Atlanta: SBL, 2008). Cosaert is quite critical of Mees lack of methodological control. He finds only three passages in which Clement quotes Mark (one of which is the extensive and rather wild text of Mark 10.17-31). 

Lessons learned: 

1. Numbers are not actually raw data, there is usually a chain of scholarship and assumptions behind them (Gathercole → Mutschler → van den Hoek → Stählin).
2. The chain of scholarship and assumptions needs to be checked against actual data at some point. 
3. Statements that look or feel wrong may well be wrong, but you still need to check the actual data.
4. The attestation of Mark's Gospel in the early period is less full than the other canonical gospels (and in actual fact is more comparable in terms of citations and manuscripts with some of the non-canonical gospels).
5. It probably is better to cite monographs on the actual author or question when they are available (e.g. both Mees and Cosaert are far more helpful and accurate than Mutschler).
6. None of this makes much of an impact on the overall argument of Simon Gathercole's book.