The Rhetoric of Gospel Re-writing
SBL Annual Meeting, November 1999

Mark A. Matson

At the risk of over-simplification, there are two primary methods for interpreting the gospels. On the one hand, one can approach them from the standpoint of the author. From that perspective, the interpreter analyzes the sources, the structure of the narrative, and the modifications to the narrative which the author/evangelist made on his or her sources. From these analyses of the raw material of the gospels, and the author’s use of the raw material, the interpreter seeks to understand what the author was trying to say -- might we even say we seek to reconstruct some idea of his or her intention? On the other hand, the interpreter can approach the gospels strictly from the standpoint of the reader. From this perspective, using such approaches as reader-response or ideological readings, the focus is on the impression the text leaves upon the reader, or which the reader finds in the text. The text’s meaning is located in the reaction to the text, without any reference to the author.

These two broad hermeneutical approaches, both of which I want to affirm as having significant value to the biblical interpreter, would appear to share very little in assumptions or methodology. But, as they generally are applied to the gospels, they often do share some fundamental assumptions; assumptions which might, in fact, be false or misleading. A major assumption often shared by interpreters from various approaches is the naivete of the readers. That is to say, the gospels are assumed to be written to readers who are previously unaware of the gospel story, at least in any current narrative form. The reader is, then, expected to be receiving new information, a virgin story so to speak, that transmits new information or interacts with the reader’s perceptions with little or no previous information about the story of Jesus to impede or interact with that gospel’s construction. And in this case, from the perspective of the author, the writer is packaging information about Jesus the he or she is sending forth to the readers – the meaning is wrapped up in the contents of the material, the "what" of the story. Or, likewise, the reader is seen as unwrapping the hidden package; and while each reader might approach the task slightly differently, the contents hidden in the package ultimately are the goal of the enterprise.

A corollary to this image of "new information" transmitted to a naive reader is the uni-vocal or directed-ness of the communication model. A gospel writer sends information off to the reader – the gospel is sent off on its journey from one location, and ultimately received in another; the rocket is launched from one place, and falls to earth in another. Granted, these metaphors conjure up rather extreme images of distant communication between a writer and reader. Surely this is different if the writer is writing to a more intimate and known group, perhaps his or her own community? And this is the normal perception – that the evangelists addressed the needs of particular communities, particularly ones they participated in. But even with this model, the communication model is uni-vocal if the reader is assumed to be naive – that is to say that community recipients are no more participating in the construction of the gospel message than a single recipient at a distance if they are primarily receiving new information.

It is my contention that this standard conception of the gospel being addressed to a naive reader within a local community, of which the author is conceived to be a member, may well distort the location of meaning in the gospels. If, as I will suggest, the gospels were widely circulated by design – that is they were intended to be and were in fact broadly published – then this would have a far-reaching effect on our understanding of the readers, and the rhetorical purpose of the gospels. Particularly for the later gospels, Luke and Matthew, it would significantly affect the way we might read them in light of the original readers’ previous literary exposure to the earlier gospels. Indeed, if Luke’s audience, to take an example, already knows Mark and perhaps Matthew or Q, the primary sources for Luke, because they have already been broadly circulated within the church, then we should be conceiving the literary and rhetorical task as one involving dialogue – both with the gospel sources, and with the audience.

In a recent book, The Gospels for All Christians, and particularly in the lead essay in that book, Richard Bauckham has called into question a major underpinning of the leading assumptions about interpreting the gospels. What Bauckham questions is the idea that the gospels were written to and for distinct communities. Rather than seeing Mark as directed to a group in Rome, and Matthew to a church in Syria, and Luke perhaps to a church in Caesarea, [1] Bauckham argues that the gospels were written for a universal audience, and were in fact distributed broadly at once. Without elaborating extensively on all of his paper, which I find intriguing and generally convincing, the following gives some idea of the main lines of argumentation:

A. There is no objective evidence that the gospels were addressed to single narrow communities.

The understanding that gospels were produced within communities and for that community’s use became prevalent in 20^th century scholarship, partly as a result of B. H. Streeter’s formulation of the synoptic relationships. It certainly made it easy to understand why there were four distinct gospels which all tell the story of Jesus in different ways: each arose in and for a different local community. From the vantage point of such an "intra-community writing," the special orientation of each gospel reflects the concerns and social realities of each community. The gospel material can then be read, as Bauckham points out, as allegories of the communities in which they were produced. Mark’s apocalyptic material can be seen as a reflection of a community’s urgency about the future in the difficult years preceding the destruction of Jerusalem. [2] Matthew’s emphasis on the Pharisees can point to a tensive relationship with developing rabbinic Judaism, perhaps not too distant from Jamnia. [3] Luke’s emphasis on rich and poor themes could suggest that concerns of divergent social status were being struggled with in the Lukan community. [4] Perhaps the most extensive development of the idea of the community orientation of a gospel is seen in the interpretation of the Fourth Gospel, especially under the influence of J. Louis Martyn’s influential work. According to Martyn, the evangelist of the Fourth Gospel is reflecting his community’s concerns about becoming distinct from the synagogue community from which it sprang, and from which it is being excluded. [5]

But while such readings are possible, as indeed the proliferation and ease of such interpretations suggest, aided in part by the increasing interest in social-scientific methodologies, they are by no means necessary and certainly not mandated by the text or by early testimonies about the gospels’ production. The narrative interests of the gospels are not readily or obviously reflective of a single narrow community interest; indeed they may well point to an ideal world imagined by the author, built creatively on the basis of previous gospels or other source material.

One argument which could go a long way to supporting the idea that gospels were written in and for specific communities would be the success such a strategy has in understanding the gospels. But Bauckham argues that the perspective in which the gospels were written for specific communities nonetheless fails. This failure can be seen in two ways. The first is that a community model of writing is not necessary to explain the major features in the gospels. In other words, the idea that gospels were written to distinct communities does not ultimately explain much in the text. Even the most conspicuous case, that of Martyn’s reading of John 9, that the introduction of the birkhat ha minim (the cursing of the heretics) into the synagogue liturgy caused a crisis for the Johannine community, fails to add significantly to the historical understanding of John. For indeed if the rabbinic council had mandated the addition of the birkhat ha minim in the liturgy, as suggested, its impact would have been widespread and not limited to a small narrow community. In other words, the specific concerns could be seen as widespread, not parochial in scope.

Secondly, this strategy of reading has produced diverse interpretations. While the idea of a community within which a gospel arose, and to which it is addressed, would seem to give some insight into the purpose of the gospels, the reality of scholarship using this approach suggests otherwise. As Dwight Peterson notes in his study of the use of the Markan community: "... this commitment to the interpretive determinativeness of the Marcan community does not work. ...virtually every scholar who discovers a Marcan community behind the Gospel – that is, the community for which the Gospel was written, and which is supposed to serve as a control for a reading of Mark – discovers a different Marcan community." [6]

Up until now, the main arguments proposed have been negative, poking holes in the local community model of gospel production. But there are a number of indications that the gospels were intended to be broadly distributed from the outset.

1. The genre of gospels suggests a broad audience. It is not directed toward specific concerns, but seems to be deliberately open. Given the predominant oral culture of the time, writing often served as a substitute for the presence of the author. Certainly this is true for epistles, like Paul’s, which substituted for the living voice of the one writing the letter. But it is likely true for the gospels as well. Why write a gospel down when the "living voice" can retell the story at any time? The very act of publishing a story, then, suggests some geographical distance, some separation between the voice and the hearer/reader. And the absence of clearly stated specific concerns suggests it was intended for more than just one community.

2. Bauckham notes that individuals in the early Christian communities, not to mention leaders such as Paul, Timothy, Barnabas and Peter, were very mobile. And this mobility would suggest that the early church was aware of the geographical spread of the early religion. Certainly the remarks in Acts and Paul’s letters point to a number of itinerant members: Priscilla and Aquila, Phoebe, Epaenetus, Stephanas, Fortunatas – and many more, testify to a healthy movement of Christians to distant locations. With such traffic in people and ideas, it would not be surprising that the broad transmission (publication) of documents was not only not accidental, but intended.

3. The early Christian movement saw itself as a world wide movement. The letters of Paul speak to a concern that cuts across regions. He takes up a collection from Macedonia for Jerusalem. He addresses Rome despite not having gone their yet. First Peter addresses a whole series of regions in Asia Minor. These documents do not suggest hermetically closed communities, but rather a flow of information, and of mutual concern, that was certainly very broad. Is it not likely that the gospel writers have conceived of precisely this kind of broad audience?

There is then, at the very least, a strong possibility that the gospels were not written for narrow consumption, but for broad distribution. And once some gospels were broadly distributed, or thought to be broadly distributed, subsequent gospels would have to account for their presence in the very Christian communities that would receive the "new editions."

Now if Bauckham is correct, and I want to simply hold that out today as a strong possibility without offering additional arguments in its favor, then the implications for the later synoptic gospels, and in particular Luke, are significant. [7] For in that case, Matthew and Luke can be reasonably seen to be offering up gospels to audiences that they might assume already know and perhaps use Mark. Let me reiterate this point just a bit more strongly. If we assume, with Bauckham, that Mark, the first written gospel, was intended for a broad audience and was published broadly, then Matthew and Luke would be publishing their own gospels to audiences that, at least a significant portion of which, have heard Mark already and perhaps use it as a regular part of the church’s instruction and liturgy. Judging from the way the gospels became used by the early churches, we might even assume that where Mark was heard, it was read and re-read. In that case, the universal audience and broad publication would deny, at least for Matthew and Luke, the probability of a naive audience. Instead, Matthew and Luke re-tell the story, using in large part the same material – often word for word – as the church already is using and knows.

So what? Well, in the model most often applied to Luke or Matthew, which assumes no prior knowledge of previously written gospels, one avoids dealing with the question of how the author engaged and interacted with the audience by referring to a prior source. [8] But in this model, Matthew and Luke would have to anticipate their audience knowing the gospel, and reacting to both the similarities and the differences. The communication model would then be dialogical, not uni-vocal. Thus the audience’s anticipated reactions are part of the writing process – the author writes the text in dialogue with the audience.

In this case, the use of Mark becomes not simply a source, but an inter-text. Intertextuality works from the assumption that all utterances are social and already involve an ongoing dialogue of thoughts and ideas and perceptions into which the author joins. In other words, when we utter sentences, we join our thoughts with – indeed with think through the previous thoughts – of those who have gone before us. These previous discourses are echoed in major and minor ways in our choice of language, allusions, figures of speech. This is the idea broached by Mikhail Bakhtin in his theory of discourse – that communication is continually working within a web of dialogues with previous texts and conversations.

While this is true generally of all discourse, it is even more valid in certain literary discourse, especially prose. Citation, reference and allusion are all part of the literary matrix, and point to an author’s engagement with both prior texts and with an audience’s perception of those texts. This dialogue with previous texts can take place in a wide variety of "degrees of presence" (to use Todorov’s term) or "volume" (to use Richard Hays term): repetition, appropriation, modification, debate, parody, ellipsis. To the degree the audience shares in the prior dialogue the referenced texts, the author is indeed engaging in a rich and complex intertextual dialogue. Furthermore, drawing on Kenneth Burke’s discussion of the broad nature of rhetoric, even in literature, the use, misuse, or modification of existing texts works as a rhetorical device. [9] The audience is drawn into a stance of identification with the story that they already know, but then are challenged and exhorted to modify their opinion based on the degree of distance from the intertext.

It is precisely this richness of the dialogical act, in which both author and audience engage in appropriating already well-known texts that, I am suggesting, best fits the character of Luke’s gospel.

It is clear that Luke has drawn on Old Testament Scripture in multi-faceted ways in his gospel. In addition to direct citation, Luke evokes other texts (as for instance the Magnificat evokes Hannah’s Song from 1 Samuel) – making allusions which are surely meant to engage the audience in comparison with the Scripture. [10]

And it is not by any means a new idea that New Testament authors were creatively re-writing prior gospels as an act of interpretation, as for instance Michael Goulder’s Midrash and Lection in Matthew, Eric Franklin’s Luke: Interpreter of Paul, Critic of Matthew. But these approaches are blunted if the audience is not conceived of as an active participant in the re-writing process. In Goulder’s case, for instance, if Matthew’s midrash on Mark is written as a dialogue with an audience that also knows Mark, then an entirely new dimension is added to the discussion.

What I propose, then, is a reading of Luke, in light of Bauckham’s theory of gospel audience and distribution, which would mean that Luke was produced in dialogue with, and to be read in dialogue with, his predecessor texts. But when we consider the inclusion of the audience, and particularly the audience’s prior knowledge, in the construction of Luke’s gospel, we enter into the realm of rhetoric.

At this point, it would be well to see whether this understanding of Luke’s re-telling of the gospel story to an audience already familiar with at least one narrative, Mark’s, gives us a new and helpful angle to approach the Third Gospel. What I propose is to test an example, by which we might gain an initial assessment of the usefulness of my observations. The text I wish to consider is that of Pilate’s trial (Luke 23:1-25).

Luke begins the Pilate trial pericope very similarly to Mark, with the assembly of the chief priests and scribes delivering Jesus to Pilate. Unlike Mark, however, Pilate does not unexpectedly and without any prompt ask Jesus if he is the King of the Jews. While Luke’s Pilate does ask Jesus this question, it is only after the Jewish leaders have brought a formal accusation, many of the points of which are demonstrably false according to Luke’s narrative, in which they specifically accuse Jesus of claiming to be the King . Luke’s introduction at this point modifies the story to shift the burden of the hearing from Pilate to the Jewish leaders. Pilate is seen as somewhat passively responding to the false charges of the Jewish religious elite. While Mark has the leaders offer unspecified charges, situating them after Pilate’s curious question in effect hides them, leaving Mark’s Pilate the main actor and interlocutor of Jesus.

Similarly, the subsequent response of Pilate in Luke displays a telling modification. Pilate immediately declares Jesus innocent, which brings an additional charge of insurrection by the Jewish leaders. In Mark, by contrast, Pilate makes no initial effort to free Jesus, and the trial seems to stall with Jesus’ quiet response.

Luke then inserts the account of the Herod trial, which is totally absent in Mark. In it Pilate appears even more open-minded and willing to let the Jews (as Herod seems to be viewed) judge their own. The subsequent appearance before Pilate finds him declaring Jesus innocent again, concurring with Herod’s own judgment of innocence, and this time specifically citing and rejecting the very charges brought by the Jewish leaders. This second appearance is an expansion beyond Mark’s account.

Finally, the trial moves to the conclusion in both Mark and Luke with similar features: the freeing of Barabbas instead of Jesus, and the final judgment of crucifixion against Jesus. But the two narratives move in very different ways. In Mark, Pilate is seen as attempting to play off the people against the Jewish leaders, and introduces Barabbas as pawn in this game. The chief priests, of course, stir up the crowd to request Barabbas’ freedom and Jesus’ death, but Pilate is hardly seen as actively seeking Jesus’ release – the best one can say is that Pilate asks in response to the calls for Jesus’ death "what evil has he done?"

In Luke, by contrast, the Jewish leaders demand Barabbas instead of Jesus in response to Pilate’s desire to free Jesus. They introduce the switch, not Pilate. In response, Pilate seeks once more to free Jesus, but the leaders overpower him with loud cries, so that their voices prevail.

It is clear that Luke and Mark tell very different stories here, yet most of the raw material is the same. Although very different in tone, it is remarkable how much of Mark still remains in the Lukan account. Thus while some critics, because of the careful editing and addition of material, would argue that Luke has relied on a non-Markan pre-Luke source, it seems to me more likely that Luke has simply used Mark to his greatest rhetorical advantage. Luke has added material to the Markan account, and rearranged some of the narrative, and in the process dramatically shifts the blame to the Jewish leaders and away from Pilate. This we all know from studying Luke. But how would an audience have reacted, indeed how would Luke have expected his audience to react, if they already knew Mark?

At various junctures, the modification of the Markan narrative would be either intriguing or startling to an audience which knew the gospel of Mark, and I propose, would make the point of the Lukan narrative even more compelling. An audience that knew of Pilate’s sudden query "are you King of the Jews?" from Mark, would find this startling question muted and contextualized by the prior accusations against Jesus. The audience, in reading of the Herod trial, which expands on Pilate’s finding of Jesus’ innocence, might well find it undermines their confidence in the sparser Markan account. The story of Barabbas’ release in preference to Jesus has become an even stronger illustration of the intensity of the Jewish leaders prosecution of Jesus: when they unsolicited demand that Jesus by taken away, reader who knows Mark’s account is shocked at the motives displayed. And when they intensely cry "crucify, crucify" over the objections of Pilate, the point is driven home even more emphatically.

It would appear that Luke utilized Mark at precisely the dramatic points in the story to shift the blame from Pilate to the Jewish leaders. Was this, perhaps, because Luke knew the dramatic effect they would have on those who already knew the gospel story from Mark, and he thus subverted the existing story to shift emphasis? Is this not, indeed, what Luke means when he says he intended, "after investigating everything carefully, to write and orderly account …" (Lk 1:3)?

In a standard view of Luke’s authorship, we might see Luke as simply having a different take on the trial of Jesus. But if we consider the dialogical, and hence rhetorical, nature of Luke’s project, if we bring the audience’s knowledge into the discussion, then we are left with a different view of the narrative. Luke is deliberately engaging his audience to rethink the responsibility of Jesus’ death, to mitigate or remove the calumny from Pilate, and to shift it completely onto the Jewish leaders. I propose that this re-imaging of the respective roles of Pilate and the Jewish leaders is carefully crafted, and depends for its effectiveness on the prior knowledge of Mark by the Lukan audience.

This paper began with an invitation to consider the importance that prior knowledge of the gospels would play in the meaning of later gospels. The standard conception, which is actually fostered by the model of distinct geographical audiences for the gospels, is that prior knowledge of gospels is usually not there, or if possible, unimportant. Bauckham’s critique of the narrow geographical audience raises serious questions about the viability of this viewpoint.

But equally important is a discussion of what it would mean to re-use a text which is already well-known by the recipients. What would that do to the way we read, the way we construct meaning? Drawing on Bakhtin’s concept of dialogue, from which much of the modern theory of intertextuality is developed, I suggested that such a dialogical approach would force some rethinking on how we should read Luke. The brief sample seems to bear that out: in this pericope from the passion narrative, Luke appears to be engaging his audience in a reconception of important elements of the Jesus story. This engagement encourages dialogue, and opens the text up to further re-readings and re-interpretation. There is an apparent meaning in the simple narrative of Luke – but a fuller reading, a rhetorically nuanced reading may be had, surprisingly, through the agency of earlier forms of biblical criticism – source and redaction criticism.

1.  This follows B. H. Streeter’s formulation in The Four Gospels, (especially pp. 9–15), which has been very influential.

2. Howard Clark Kee, The Community of the New Age, p. 176.

3. Davies and Alison, The Gospel According to Saint Matthew, p. 136-7; cf. further p. 138-147 on the discussion of location of authorship, which seems to relate the internal issues in the gospel entirely to the possible audience.

4. Philip Esler, Community and Gospel in Luke-Acts, p. 25.

5. J. Louis Martyn, History and Theology of the Fourth Gospel. Cf. Raymond Brown, The Community of the Beloved Disciple.

6. D. N. Peterson, "The Origins of Mark: The Marcan Community in Current Debate," Unpublished Ph.D. dissertation, p. 216.

7. Assuming, of course, Markan priority. Just what constitutes the "later" gospels will vary according the theory of gospel relationships. But while the practice of interpretation might differ, the significance of the gospels’ audiences on the interpretation process discusses in this paper will remain equally valid under any theory of synoptic relationships.

8. So, for instance in Joel Green’s commentary, The Gospel of Luke, (p. 15) he notes "Similarly, while we remain reasonably confident that one of Luke’s major sources was the Gospel of Mark, and that other written narrative materials was available to him, we do not imagine that his first readers had access to their own copies of the Second Gospel or that they could (or would have been interested to) compare the texts of Luke and Mark in ways consistent with modern redaction criticism."

9. Kenneth Burke, A Rhetoric of Motives (Berkeley: University of California Press, 1950)

10. See, for instance, the very suggestive work by Robert L. Brawley, Text to Text Pours Forth Speech: Voices of Scripture in Luke-Acts (Bloomington: Indiana University Press, 1995).