Thoughts on Antiquity

This is a comparison of the Burton L. Mack’s and John Dominic Crossan’s work on Q’s compositional history and the historical Jesus. See previous post for clarification.

Of Crossan’s 503 original complexes attributed to Jesus in early Christian writings, Mack declared that less than 10% of them are candidates for authenticity (Lost Gospel, 191). Mack does not note that Crossan subsequently revised his count of 503 to 522. Mack attributes 44 of complexes to Q1. This leaves no more than six other complexes as possibly authentic (44+6 [is less than] 50.3). Given that he finds the possibility of authentic complexes in other pre-gospel sources, he does not even believe that one can trust the whole of Q1 as a source for data. (Such documents might include his pre-Markan chreia source, earlier form of the Gospel of Thomas, and “parable collections.”)

By my count, there are 23 complexes in Mack’s Q0 (16 per Kloppenborg) and 13 candidates for authenticity in the Markan chreia source, approaching the 10% he suggested. This list is brief, consisting of the “aphoristic core” of the Q1 clusters and perhaps other, originally unrelated, sayings that were appropriated into them as well.

This form of the document is reconstructed by removing Q1’s redactional features, namely the framework of its argumentative structures, leaving us with the ethos that the clusters were written to defend. However, Mack’s use of Q0 is odd and somewhat troubling. In his reconstruction of Q, he places the beatitude for the persecuted (6:22–23), as a whole, in Q2. Many, including Kloppenborg, have objected to this placement, instead preferring to keep Q 6:22–23b in Q1 and attributing only 6:23c to Q2, for its obvious Deuteronomistic theme (see above). Confusingly, Mack believes that this part of this verse was present in Q0, but why, exactly, is unclear.

Crossan, by contrast, believes that about 25% of his 522 complexes are authentic (see link above). In Historical Jesus, Crossan places 47 items in Q1, 33 of which he believes to be authentic (including his five “1Q?” complexes). If we take them as a whole, 70% of Q1 items are taken to be authentic to the historical Jesus. Regarding those sayings whose layer is ambiguous, seven of the eleven complexes (64%) are authentic. For Crossan, 51 sayings are securely in Q2, twenty are authentic, about 39%. While one might reasonably look at the data and assume that behind the correlation between stratigraphical location and authenticity there is also an issue of causation, there is more to what is going on than meets the eye. Of the 31 complexes that he deems inauthentic in Q2, 23 are singly-attested, probably indicating that a later layer is significant when coupled with another factor, in this case lack of external testimony. Thus, if we look only at the multiply-attested complexes in Q, we see that Q1 is 72% authentic, and 65% of Q2, relatively close. This is consistent with Crossan’s claim of preference for Q1 over Q2 material due to the former’s relative frequency of plural attestation.

If looking at those sayings that are found only in Q (i.e., singly-attested complexes), very different numbers come up: 64% of Q1 is authentic, but only 12% of Q2. Crossan’s method is to assume that all complexes attested in the first 60 years CE (including all of Q) that have plural attestation are authentic unless argued otherwise. Given that Q material was composed quite early in Christian history in his mind, it is quite reasonable for him to prefer these documents to ones composed later, even if he is accepting parts of them rather uncritically. Indeed, his preference for the earliest layer of Q is similarly unsurprising, as one would very reasonably suspect that more, and a greater proportion of, authentic memories would be in the tradition at that time.

But given the high rate of inauthenticity among singly-attested sayings in Q2, one might be surprised to see that Crossan treats singly-attested special-Lukan material much more favorably, despite the fact that it is unrepresented in the Jesus tradition before 90–120 CE to his mind, and totally absent outside this gospel. 22% of these special-Lukan complexes there are authentic, despite its content having no independent attestation. Another point of contrast may be drawn to singly-attested special Matthean material, 9% of which he believes to be authentic, despite having been composed shortly before Luke. Given how quickly Q’s otherwise-unattested complexes became inauthentic, it is surprising he gives so much uniquely-Lukan material the benefit of the doubt.

I will soon do a much shorter post with Kloppenborg’s changes to his stratification of Q over the past 25 years. This edge-of-your-seat excitement will finally reach its conclusion.