The Jewish-Christian Gospels, pt. 2: The Nazaraean Gospel in Alexandria

After a year of life-changes (graduate degree, new job, new city), it's finally time to start posting again. I'll continue my series on the Jewish-Christian gospels and the documents known as the "Gospel of the Hebrews". So, picking up where I left off:

4. The Alexandrian school (Clement, Origen, and Didymus) also seems to have known the Nazaraean gospel, but indirectly.

The Jewish-Christian Gospels, pt. 1: Ebionites and Nazaraeans

I've been promising this for a long time, so I better finally deliver :)

To explain the identification of Q with GEbi, and of both GEbi and GNaz (and perhaps GPet) with "GHeb", we'll need to sort out the evidence regarding these gospels and their relationships with not only GMt but also GLk-Ac. This will take several posts. I largely draw on Ben Smith's extremely helpful pages on the Jewish-Christian Gospels (helpfully indexed here), but augment this occasionally with material from elsewhere. I do not really rely on any one study of these gospels (such as Klijn or Luomanen) because I think that so far, no one has sorted out these gospels quite right. I believe that the HSH will help us see the true relationships.

I generally use Ben Smith's translations, but rely on Frank Williams' translation of Epiphanius. I copy the Greek from Ben Smith as well, who does not use diacritical marks. When translating the phrase "according to the Hebrews", I place it in quotation marks: otherwise when referring to it I use acc. Heb.

(ETA: fixed the translations from Jerome--I had accidentally deleted an intermediary block.)

So, let's begin with our first observation:

1. The gospel of the Ebionites, and that of the Nazareans, were different texts, as nearly all scholars now recognize.

A Home for GEbi, and a Solution for Q

ETA: In 2010, I began identifying Q with the Gospel of Peter, and both with the Gospel of the Hebrews. Later that year, I distinguished between GPet and GHeb, but continued to identify the former with Q. 

Over the years, my thinking on this evolved considerably, until late in 2013, I realized I had it wrong: Q is to be identified with GEbi and (one version of) GHeb. GPet was posterior to Q/GEbi.

This post, and several earlier posts, have been edited to reflect this realization. I have tried to note where I had made edits.

This is the post in which I first identified Q with GEbi. Some material has been removed; one bullet point has been added, as noted below.

Preliminaries, and a new diagram

I have little time this week, unfortunately, to devote the attention I need to the question of the Gospel of Thomas and the Synoptics (though I have already drafted some comments), but I did have time to put together a new HSH diagram.  Here it is:

The HSH in the Context of Johannine Studies

I thought I'd finally put this post up--I wrote it last fall as part of my ongoing exploration of the role that the Signs Gospel plays in the Hyper-Synoptic Hypothesis, but got busy with other things and it's been sitting for a while.  But I'd like to explain how the Hyper-Synoptic Hypothesis relates to the study of the Gospel of John, so I've revised this post and here it is.  


The HSH, at least with regards to the role of the Signs Gospel, does fit into a previously-established tradition of Johannine scholarship, though this tradition is perhaps not a mainstream one.  The Signs hypothesis in general (that is, that there was a source for the Gospel of John that contained a sequence of symbolic miracles called "signs") is not universally accepted, and indeed one might argue that it has never been mainstream opinion.  I wanted to explain that I am aware of this, but that there is nevertheless an ongoing thread in Johannine studies that does assume the existence of a Signs document (of some sort or other), even one that relates to the Gospels of Mark and John in the way that I claim it does.


I'll present this as a more-or-less chronological study of the concept of a Signs source in the history of New Testament criticism.

A Brief History of Scholarship on the Relation of GJohn to the Synoptics

Adding Egerton and Syriac GTh

(This post gets a little technically dense; sorry about the brevity. ETA: typos corrected!)

I was mulling over the Egerton Gospel, as I do from time to time, trying to think of where it might fit into the HSH.  I had become particularly  interested in its mixture of Johannine and Synoptic elements (which were always triple-tradition elements).  I'd never been able to fit it into the diagram before, but what if it were a kind of proto-John?  The "Johannine-only" elements would then have been taken from the Signs Gospel; the synoptic triple-tradition elements from the original Gospel of Mark, i.e. Secret Mark.

This would have been a slightly complex idea, however, because I would then need a reason why the canonical Johannine redactor either edited out or heavily rewrote the synoptic elements.  This wasn't an impossible idea, but I didn't see anything in those elements (the questioning of Jesus by his opponents, and the healing of a leper) that seemed to require special treatment by a Johannine author.  So this remains an intriguing idea, but I don't think it's the best one.

I then turned to the Q side of the diagram to reconsider placing it there.

The Lost Ending of GMk: Conclusion

In my last post, I presented a completed ending to original GMk.

So what happened to it?

The Lost Ending of GMk, pt. 4

In the last post, we completed a chiasm for the original ending of GMk--two chiasms, actually, including the pericope begun by Mk 16:1-8.  Although we completed each bracket using material shared among the sea-miracles in GMt (the Water-Walking and the Fish and the Coin), GLk (the Calling of the Disciples/the Catch of Fish), and GJn (the Appearance in Galilee/the Catch of Fish), the end result seemed a bit brief, as though there were still more material to add.  In this post, we'll find that remaining material in the same manner, by searching for shared material among GMt, GLk, and GJn.  We'll find that it fits nicely into chiastic brackets, thus completing the original ending of GMk.

Again, here is what we have so far:

The Lost Ending of GMk, pt. 3

Before I proceed, I wanted to mention this article that James McGrath published in February of this year, essentially on the same subject: the probable existence of a missing ending to GMk based on the internal, textual evidence of GMk up to Mk 16:8, and the witness of GPet to the content of that missing ending. 

At any rate, in the last post, we got a kind of rough outline of what a missing ending of GMk might have originally looked like, based on the principle that the first verses of Mk 1 would form a kind of super-chiasm bracket with it--that is, that we should expect to find the specific themes of Mk 1 replicated at the end of the Markan gospel.

And I argue we can go even further with this, and figure out the specific details found in this final pericope.  In a previous post on the signs in GMk, I provided two lists from Raymond Brown’s Anchor Bible commentary on GJn: one of parallels between Lk 5:1-11 and Jn 21, and another of parallels between Mt 14:28-32 and Jn 21. 

The list of Lukan parallels is as follows (with triple parallels in the Matthean Water-Walking underlined)

The Lost Ending of GMk, pt. 2

Now consider the idea that the entire Markan gospel was originally one big super-chiasm.  This idea has been best put forward by M. Philip Scott, in Biblical Theology Bulletin ("Chiastic Structure: A Key to the Interpretation of Mark's Gospel", BTB 15.1, pp. 17-26), laid out in rather intriguing detail.  Pairs of opposed ideas do seem to be spread throughout the gospel, in nested conceptual brackets, just like they are within individual pericopes.  For example, Scott characterizes the beginning and end of the gospel (ibid, p. 18) as "An angel witnesses to his coming" with "An angel witnesses to his going", though I would amend the second half slightly to “An angel witnesses to his returning”.

So once we recognize the overall chiastic structure of Mark's gospel in general, the remaining pieces of the missing ending fall into place, along with the original super-structure.  Notice that what we have at the beginning of Mark’s gospel is not just John’s announcement of the coming of a messiah, but also some substantial baptism imagery.  Not only do we have the baptisms of John's followers, but we also notably have the baptism of Jesus.  Furthermore, we have John's promise that Jesus will baptize in the Holy Spirit--and, following Mark's adoptionist theology, we also have Jesus' own baptism in the spirit.

The Lost Ending of GMk, pt. 1

I did promise a post on the lost ending of GMk, didn't I?  Well, alright, here it is.  This will be a series of two or three posts, I think.  Then we'll get back to describing the final sign in the Signs Gospel.

The key to understanding the lost ending of GMk lies in understanding the chiasmic nature of the original GMk.  As I have often mentioned, Michael Turton (following principally John Dart, but others as well of course) has helped show how each authentically Markan pericope has achiastic structure with a twinned or paired center.  I don’t think Turton gets every one quite right (and in fact I’m probably going to do my own version of GMk, chiasm by chiasm), but he at least makes a good go at it, and some are very insightful.   

Turton tries his hand at constructing a chiasm for Mk 16:1-8, showing that it seems to be missing an A' bracket, ending enigmatically--and infamously--with "they were afraid, for" (EFOBOUNTO GAR).  There is a strenuous debate in the scholarly literature about whether this is a sentence fragment or not, focusing in particular on the question of whether GAR can serve as the end of a sentence, as well as at the end of an entire work.  To summarize, yes, sentences can occasionally end with GAR in ancient Greek (including the Septuagint), but only one other work, the Third Treatise of Plotinus, ends in GAR.

There is also the problem of why the Markan author would predict an appearance of the risen Jesus in Galilee, but fail to describe it, or why he would leave the reader hanging on the questions of whether the women would tell the disciples about what they saw, whether the young man's prediction would come true, whether Jesus were truly risen at all, etc.

Here is Turton's attempt at a chiasm:

The Seventh Sign, Pt 5: The Catch of Fish

Like the Water into Wine, the Catch of Fish also leaves traces behind in GMk.  It will turn out that they're harder to find, because the sign has actually gone missing, but nevertheless they're still there.

The first location is found in Mk 1, towards the beginning of Mark's gospel.

A Catch of Believers

When the first disciples are called in GMk, we find Simon and Andrew casting a net in the sea.  We also meet James and John in a boat with their father Zebedee, mending their nets. Jesus tells them that they will become "fishers for men" and makes them an offer: "follow me".  They abandon their nets and follow him. 

These elements all have parallels in the Catch of Fish as we have it in GJn.  There, Simon, James, and John--the last two identified as the "sons of Zebedee"--are all fishing.  Jesus tells them to cast their net into the sea.  After he reconciles with Peter, he tells him "follow me". 

Bethsaida, Secret Mark, and the Not-So-Great Omission, pt. 4

In the last post, I made a case for my claim that an original version of GMk was used by the Q-author to produce a Q gospel that lacked the Bethsaida section.  This Q gospel was then combined with a canonical version of GMk to produce GMt, which used the Bethsaida material in canonical GMk; and GLk/Mc, which left out the Bethsaida material in canonical GMk (since it was absent in the Q gospel).

To further demonstrate the plausibility of this scenario, we'll have to show that the Q-author would have had reason to omit each of the original pericopes in the Bethsaida section in his Q gospel.  Once again, these original pericopes were determined, by Michael Turton, to have been:

a) The Water-Walking
b) The Feeding of the 4000
c) The Blind Man of Bethsaida

We deal with each pericope in turn:

Bethsaida, Secret Mark, and the Not-So-Great Omission, pt. 3

In Part 2, we showed how Michael Turton, in his Index to Historical Commentary on the Gospel of Mark, reduced the so-called "Bethsaida section" of GMk (roughly Mk 6:45-8:26) to only three original pericopes:

• the Water-Walking
• the Feeding of the 4000
• the Blind Man of Bethsaida

These three pericopes were the only ones written by the original Markan author, i.e. the author responsible for most of the Gospel of Mark as we know it--its overall narrative, its structures, and its characteristic features.  Turton simply calls this author "Mark" (though this author was not the editor of the canonical GMk.)

So there was an original version of GMk that featured these pericopes alone, in this order, in this section of the gospel. 

We know that Matthew nevertheless used canonical GMk in composing his gospel, because he has all the material in the Bethsaida section that the canonical author has.  But did Luke/Marcion also use canonical GMk? 

Bethsaida, Secret Mark, and the Not-So-Great Omission, pt. 2

We now turn to Turton's analysis of the Bethsaida section of GMk.  

Turton individually examines the pericopes in this section of GMk using chiastic analysis.  He eliminates any pericopes that do not exhibit what he feels to be typically "Markan" characteristics; chiastic brackets that "talk to each other" (i.e. exhibit thematic or vocabular parallels), central bracket pairs that feature chreia, sayings by Jesus that neatly summarize teachings or deftly refute opponents, etc.  As summarized in his "Excursus on Mark without Bethsaida", this leaves him with four pericopes that are arguably Markan in this section:

Bethsaida, Secret Mark, and the Not-So-Great Omission, pt. 1

We now begin an examination of the "Bethsaida section" in GMk, so-called because at its beginning in Mk 6:45, Jesus and his disciples set out in a boat for Bethsaida, and at its end in Mk 8:27, there is a healing in Bethsaida of a blind man.

Michael Turton has done excellent work in showing that this section is riddled with non-Markan editing and interpolations, and that there was an earlier, original, "Markan" version, containing much less material than we now find in GMk.  In other words, Turton thinks there was an earlier version of GMk, that lacked much of the Bethsaida section.

What I'll show is that Cerinthus would have had every reason to eliminate the few pericopes in this authentic, earlier version of GMk.  In other words, Cerinthus performed a "Not-So-Great Omission" on original GMk. 

The Unforgiable Sin Logion, pt. 5

Matthew also replaces Mark's unusual AMARTHMATA with AMARTIA, the same word as in Didache.  (AMARTHMATA is actually Pauline, showing the influence of Paul, especially Rom. and 1 Cor., on Mark.)  But in Mark's other uses of "sin" (AMART-) Matthew pretty much follows him verbatim.  So why would he have changed Mark's word here?  Matthew likely had another influence in this case, and Didache provides the best candiate for this influence, whether directly or indirectly.

Remember also that Matthew follows the same order as the Didache, PASA AMARTIA...AFEQHSETAI, whereas Mark uses PANTA...AFEQHSETAI...AMARTHMATA  How could Matthew have coincidentally followed the same order as Didache in these three words, if it had not influenced him one way or another (again, directly or indirectly)?


To finish our analysis, let's turn briefly to Luke's version.  Notice that he keeps the Unforgivable Sin logion away from the Beelzebub incident altogether--the Beelzebub incident is Lk 11:14-23, but the Unforgivable Sin logion is Lk 12:10.  Why would he do this?

The Unforgiable Sin Logion, pt. 4

Let's get back to the Matthean version.  Tuckett also thinks (in agreement with most scholars) that there was a Q version of this saying.  Most scholars think that it began "He who speaks against the son will be forgiven..." in parallel with GTh.  But notice...

...that both Matthew and Luke begin their versions (Mt 12:31//Lk 12:10) with either PASA (GMt) or PAS (GLk).  The parallel is even more direct in GMt, for he writes PASA AMARTIA AFEQHSETAI where the Didache writes PASA GAR AMARTIA AFEQHSETAI.  Tuckett suggests that Didache derives from GMt here.

But what if...the Didache preceded GMt?  Indeed, why would Matthew use PASA, and Luke PAS, instead of Markan PANTA?  It would seem on the face of it...that both Luke and Matthew are drawing from a single source, one that was somehow related to the Didache, and that contained the Didachean statement PASA (GAR) AMARTIA AFEQHSETAI.

The Unforgivable Sin Logion, pt. 3

GTh logion 44 ("He who speaks against the father will be forgiven...") seems to have parallels in both GMk (Mk 3:28-29) and in the Q-gospel (Q 12:10).  From these sources, it seems to have found its way into GMt and GLk--Matthew combines the two, keeping the saying in its Markan position in the Beelzebul incident, whereas Luke blends them together and relocates it somewhat.  In the synoptics, it is called, in general, the "Unforgivable Sin logion".  We'll be investigating it to see how it could have ended up in both the Markan and the Q traditions.  Our conclusions will be, I think, somewhat surprising.

The Unforgivable Sin Logion, Pt. 2

In our last post, we ended by wondering if there is direct textual evidence that the author of Mk 3:28-29 relied on Did 11:7 as his source for the saying about the unforgivable sin, rather than GTh 44.  This post will present that evidence.