2023-02-27

A Brilliant New Book on Gospel Origins

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

by Neil Godfrey

If you are looking for a serious, easy-to-read and up-to-date study of the question of how the gospels came to be written, what sources their authors used, what their authors were trying to achieve, and for the most part is delivered in conversational style, then you will have found it in Rhetoric and the Synoptic Problem by Professor Mike Duncan.

While acknowledging and questioning other views in New Testament scholarship, Duncan clearly presents a logical case for the various gospels all being polemical re-writes of the Gospel of Mark. He introduces insights that strengthen Mark Goodacre’s revamped case that the author of Luke used both Mark and Matthew and that, consequently, there is no need to postulate, as most scholars have done, a long-lost source (Q). He even demonstrates the physical process of how Luke copied Matthew and Mark without Q on the widespread understanding that authors of the time wrote with scrolls on their knees and in so doing shows that the most common argument against Goodacre’s (Farrer’s and Goulder’s) view — that Luke was unlikely to have broken up Matthew’s Sermon on the Mount if he knew it — all but vanishes.

Duncan is a scholar of rhetoric and communications but his study is very different from the pioneering “gospels through rhetoric” analysis of classicist George Kennedy that I posted about some years back. The simple justification for a rhetorical approach lies in the fact that the gospels were written to persuade and rhetoric is the study of how persuasion works.

I am not a biblical scholar, a seminarian, or even a Christian. To write to any of these audiences would be therefore disingenuous. I am an academic rhetorician who works in a university English department. I often write on early Christianity and rhetoric, and I am an agnostic who holds no text sacred. As such, I make no pretense to offer this book as a contribution to the longstanding field of biblical studies, especially as practiced by its many evangelical academics, or, on the other end, militant atheists. I have no dog in that fight; I do not care if tomorrow someone solves the [Synoptic Problem] by way of a method other than the Farrer Hypothesis that I tend to prefer, although it will be a minor annoyance in that I will have to find another example for my ideas on unsolvable problems and rhetoric. As such, this book is offered in the same spirit as Frank Kermode’s The Genesis of Secrecy, an analysis of the Gospel of Mark from the perspective of literary studies, save I’m using rhetoric as a focus, and I’m looking at all the canonical gospels at once. (p. 2)

One point of method that I particularly liked was Duncan’s demonstration that certain characters and events in the gospels function to make specific polemical points. If Occam’s razor be our guide, that means the events or characters originated in the authors’ imaginations rather than from oral tradition about a presumed historical event — though Duncan does accept the historicity of Jesus and John the Baptist. Here is an example. The author of Mark is apparently responding to the “tradition of resurrection appearances” that we read in Paul’s first letter to the Corinthians by introducing the “empty tomb”.

The author of Mark had plenty of sources available for inspiration for an empty tomb narrative, including Psalms 22, 23, and 24, the widespread Orphic theology as well as the end of the Iliad, the story of Elpenor in the Odyssey, and Plutarch’s and Livy’s accounts of the death of Romulus. But the source of the story is not as important as the kind of rhetorical claim that it allows the author of Mark to make. If the author of Mark invented the empty tomb story, for what purpose was it done?

— It could be to show a “removal”—a Hellenistic showing of an empty grave as evidence that the gods have conferred hero status on its missing occupant.

— It could also just be a simple dramatization of what the author thinks probably happened that day, working with current Jewish custom for visiting the recently deceased.

— The empty tomb could also be a narrative “promise of a personal resurrection to later Christian martyrs”—an important point for post-70 CE Christians in the wake of Jerusalem’s devastation, suggesting a similar physical resurrection for them: you, too, will die, but you will rise again.

But these are just suppositions. I can make a more defensible observation that does not require any of them. Holding the earlier points for historical plausibility in stasis for a moment, the author of Mark’s empty tomb narrative allows Jesus’s resurrection to be implied rather than witnessed. In other words, Mark can have Jesus rise without granting either Peter or the apostles any authority that they might have gained by having witnessed it. Theodore Weeden took a similar position that the authority of the twelve disciples, derived from their witnessing a post-resurrection Jesus, is removed by this maneuver, though he did not note that the empty tomb serves a dual rhetorical function by removing the necessity of eyewitnesses. In any case, the appearances in 1 Cor 15 suggest visual appearances that were witnessed, but Mark’s version lacks appearances; this could also be a subtle way to reconcile 1 Cor 15:35-49’s spiritual resurrection with the 1 Cor 15:3-8 list of physical appearances.

The narrative skill by which Mark accomplishes this maneuver, coupled with Paul’s obliviousness to any empty tomb story, refutes the notion of a longstanding tradition of the discovery of a vacated tomb. . . . 

(pp. 48f – my formatting and highlighting; italics are original)

You might recall from the Epistle to the Galatians that Paul saw himself in some kind of rivalry with Peter and resisted tendencies to exalt Peter’s status above his. We have many hints of a leadership struggle in those earliest documents. The Gospel of Mark, many scholars believe, favours Paul over the other apostles, especially Peter. The author of that gospel speaks through his literary figure of the young man in the tomb an assurance that Jesus has been resurrected and, implicitly, that he will be seen again.

The author of Mark’s argument does not need a post-resurrection appearance by Jesus to make its ultimate point: Jesus prophesied truly and not even his disciples, many of whom started a religion after his death, really understood the true implications. For Jesus to appear like a parlor trick and say, “Told you so!” would deflate the author’s call for much hardier discipleship that the original followers of Jesus mustered. (p. 52)

So where does that leave the later gospels that do contain descriptions of resurrection appearances to leading apostles?

With this understanding of the rhetorical role of Mark’s gospel as a denunciation of apostolic authority in hand, the variances of the post-resurrection appearances in the other two synoptic gospels can be better explained. They are not simply variances in tradition as many exegetes posit, but rejoinders in a hostile rhetorical conversation with peculiar rules dictated by rapidly developing theology and power struggles. (p. 53)

Duncan, as you can see in the above example, addresses explanations about this or that biblical text that many of us may have encountered and obliges us to think more clearly and thoroughly about their ultimate worth.

The book explores the various accounts of the women at the tomb of Jesus, the comparable but different versions of a few miracles, characters and sayings to demonstrate similar points of polemical rivalry among the gospel authors but concentrates on a selection of key areas. I’ll mention the others shortly.

Technical terms are introduced gently and simply for the lay reader. The scholarly literature often refers to “redaction”. Duncan clarifies the different kinds of that process (adding, deleting, tweaking, reordering, retaining) with digestible explanations along with his preference for the simpler (and, he explains, more neutral) term, “editing”. He also offers an easy guide to the different ways we tend to make decisions about various problems, including “the synoptic problem” — deduction, induction and abduction.

The one thing I did not at all like — being a scholarly kind of reader myself — was his relegation of citations and tangential discussions to endnotes instead of being on the main text page. But others, of course, much prefer that style.

Not that the book is “for the lay reader”. It is most decidedly targeted at a scholarly readership as well. The history of the scholarship and current debates are addressed in enough detail to assure both sets of readers of currency and adequate thoroughness.

I referred to the detailed discussion of the resurrection appearances above. The other key topics addressed are the John the Baptist material across the four gospels (this includes observations on “the invention of the gospel genre”), the contrasting treatments of the Twelve Apostles, and the Sermon on the Mount. An appendix exploring the question of the dates of the gospels is included at the end. Duncan does not confine his Baptist discussion to the moment of Jesus’s baptism but shows how the differences in the opening chapters of all the gospels are best understood when one considers the way the Baptist is handled in all references throughout the gospels.

The Sermon on the Mount

Why would any author want to demolish the grandeur of Matthew’s three chapters of the Sermon on the Mount? Before tackling the question directly, Duncan guides the reader through a word by word examination of Luke’s prologue. He also studies the method and style of the author of Matthew and the reasons Matthew created this classic instance of an ancient “speech-in-character” section for Jesus. With that background, the reader is prepared for a serious analysis of how the author of Luke worked and what his complaints with both Mark and Matthew were.

I have been somewhat chary about the claim that ancient authors did not use tables and therefore copying from diverse scrolls was not as easy as moderns might imagine, but it is one that I have to accept as a serious point. (Duncan gives the reasons that are advanced in its favour.) If the author of Luke wanted to break up the lengthy sayings block of the Sermon on the Mount then how easy would it have been in reality, given that the author had his scroll for writing on one knee and the scroll for reference on the other? What was involved in consulting different scrolls as one wrote? To help answer this question, Duncan sets out ten pages of columns demonstrating how the author of Luke moved between Mark and Matthew in composing his own work. (He adds 1 Kings as an additional source for the composition of the first two chapters of Luke.)

This table helps one visualize the point where the author’s “fatigue” set in (following Mark Goodacre’s explanation for certain inconsistencies of narrative and reverting to the story as per the main source). We are given a dozen specific observations arising from this table. The one with particular reference to the Sermon on the Mount:

After Lk 6:19, the author begins to deal with the issue of the Sermon of the Mount in Matthew, switching back to the scroll of Matthew. Again, when dealing with Matthew content, the author continues an alternating pattern where he copies closely from Matthew and then immediately inserts passages of his own. The passages in Matthew are more scattered this time —moving back and forth through the scroll, but not switching to Mark. And yet, the author still returns to the Matthew narrative in Mt 5 and 7. (p. 233)

For comparison, he sets out a similar table to demonstrate how the author might have worked with the hypothetical Q document instead of Matthew. Two of his seven conclusions on this exercise are worth noting:

• The Q Hypothesis chart is easier to read and understand. There are no “ambiguous” source entries because there is no choice between Matthew and Mark for similar material; Mark is the only option other than the initial use of 1 Kings in the LXX. The author of Luke’s fidelity to Mark remains the same as in Table 6.1.

• The flip side of the ease of understanding is that the lack of Matthew as a source option creates 9 content “discrepancies” in scroll usage, where Luke somehow manages to reproduce content in Matthew (and not in Mark) without using Matthew. Three are near the beginning of Luke, and five near to the end: Lk 1:26-38; Lk 2:1-20, Lk 3:23-38, and Lk 22:23-30, Lk 22:47-53, Lk 22:54-71, Lk 23:33-43, Lk 24:1-9. The remaining one, the Lord’s Prayer in Lk 11, might be explained by a differing oral tradition, but the rest, collectively, are a serious problem for the Q Hypothesis. The Q-centric criticism tend to call these discrepancies “Major Agreements” when they constitute more than a word, and “Minor Agreements” when they are only a word, but there are an alarming lot of them. If we drilled down further, we would find even more of these “Minor Agreements,” scattered throughout Luke at the level of small words and phrases but I find these 9 the most significant for the question of scroll-switching in terms of content. (p. 241)

So why does Duncan think the author of Luke broke apart Matthew’s Sermon on the Mount? He identifies three reasons:

• The Sermon, as a long philosophical speech, disrupts the rapid pace of the gospel narrative established first by Mark. In order to preserve the supposed words of Jesus in the sermon, but save the overall quick narrative pace, the author of Luke constructed a shorter version—the “Sermon on the Plain” and distributed the rest throughout the gospel. This would then be a classic example of an editing task.

• The author of Luke knew or concluded before they started composing that the Sermon was a Matthew-edited pastiche of Jesus’s sayings, and as a result, felt completely free to disassemble it to meet their own rhetorical ends as it was not authentic. The artificial ordering of the material— remember the emphasis on “accuracy” and “order” in the prologue of Luke—then had no special status to the author, despite any later aesthetic judgment of superior arrangement.

• The author of Luke disagreed with the holistic message of the Sermon as being inaccurate to his view of Jesus’s core message, and while not daring to delete the material entirely, removed the “skeleton” of the arrangement to focus on their individual wisdom sayings in turn. (pp. 242f)

Mike Duncan’s style is, as I have mentioned, largely conversational and readily keeps the reader engaged — as is surely appropriate from a professor of communication! He makes it plain that he does not present his argument as “The Answer” to the Synoptic Problem. Rather, it is an explanation that arguably leaves us with fewer problems than the other options, and those options are themselves proposals that are likewise necessarily limited by the simple fact that we are working with very patchy sources of data.


Duncan, Mike. Rhetoric and the Synoptic Problem. Lanham: Fortress Academic, 2022.

— with thanks to the publisher for granting me access to an electronic review copy


The following two tabs change content below.

Neil Godfrey

Neil is the author of this post. To read more about Neil, see our About page.


If you enjoyed this post, please consider donating to Vridar. Thanks!


5 thoughts on “A Brilliant New Book on Gospel Origins”

  1. It was the empty tomb in Oedipus at Colonus which first set me down the path of realising Jesus never existed at all – that’s it’s all stitched together from old motifs and what were at the time cliched stories, possibly as a rationalisation of John the Baptist’s failure as a messiah underneath (and possibly not).

    The reference to a lack of tables is a new one to me. I would have thought a slave would have been available to hold the reference scroll(s) or even to read out from them rather than the writer balancing them on his knee.

    1. Tables were certainly around, but some doubt they were used for the sort of writing we are talking about here. Anyway, here is Mike Duncan’s discussion point:

      However, what if you wanted to look at two rolls at once, say, to copy from one to the other? We know this happened all the time, but not exactly how. Writing desks or use of other similar surfaces do not seem to have been commonly used in antiquity.13 The reasons are unclear, given tables and chairs certainly existed (though they were far lower than modem equivalents) but given such physical constraints, one “source” roll might be draped over one knee, and the other knee used for composition, while the scribe sat on a stool or in a cross-legged or kneeling position, or perhaps only one scroll was held in the lap with another source scroll placed elsewhere.

      Robert Derrenbacker has held that working with more than one source at a time with papyrus was technically impossible for the evangelists, given writing desks were not in use,14 and common sense might seem to support this argument. The knees arrangement seems odd, of course, to a contemporary writer, who would chafe and improvise. Just sitting on flat ground and placing simple weights such as flat-bottomed rocks at the corners of a rolled-out papyrus would allow a scribe to not only see far more of both scrolls but would allow viewing more than two scrolls at once. But this does not seem to have been common practice either. The surviving “scriptorium” at Qumran in Palestine suggests a third possibility; scribes sat at low seats in a way that elevated their knees; a slightly higher platform in front of them may have held the text they were copying from, facilitating more than one scribe copying the same scroll. Again, this looks rather strange to modern eyes, but it is functional, and closely matches similar illustrated arrangements over the next 1,000 years.

      Different physical arrangements suggest, possibly, different habits of composition. . . .
      (96f)

      The Derrenbacker reference is to: Ancient Compositional Practices and the Synoptic Problem

Leave a Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Discover more from Vridar

Subscribe now to keep reading and get access to the full archive.

Continue reading