Rewriting and the Gospels

Rewritten Scripture Recently

Scholarship on rewritten scripture as a concept begins in earnest with Geza Vermes’s seminal work Scripture and Tradition in Judaism (1973). Vermes defined this phenomenon loosely in his initial work, later confirming that he thought of the concept as both a genre and process (2014: 8). He subsumed a variety of Jewish works under this category, including material composed well beyond the Second Temple period. The function of the category for Vermes was fundamentally interpretive in nature: ‘in order to anticipate questions, and to solve problems in advance, the midrashist inserts haggadic development into the biblical narrative – an exegetical process which is probably as ancient as scriptural interpretation itself’ (1973: 95).

Critical interest in rewritten Bible continued with the publication of Charlesworth’s pseudepigrapha volumes (1983; cf. Petersen 2014: 19-27). For example, Daniel J. Harrington continued Vermes’s emphasis on narrative and interpretation, but expanded the category, including works related to biblical narratives, but not based upon them in a substantive way (1986). He also narrowed rewritten Bible to ‘products of Palestinian Judaism’ (1986: 239), acknowledging that works like 1 Enoch, 2 Baruch, 4 Ezra and others share similarities with this category. Harrington concludes that rewriting is best viewed ‘as a kind of activity or process’ in contradistinction to a genre (1986: 243). George Nickelsburg (1984), too, described rewritten Bible as an exegetical strategy and included a number of compositions that expand upon particular episodes or that have tangential relationships to their putative sources (e.g., Epistle of Jeremiah, additions to Greek Esther). His understanding of the concept was more expansive, corresponding to a broader phenomenon of exegetical engagement, and moving beyond purely narrative material. The underlying exegetical similarities in these works, regardless of the genre of their base traditions, outweigh the differences in their generic features.

Pushing back against these process definitions, Philip Alexander theorized rewritten Bible on purely generic terms (1988). After examining Jubilees, Genesis Apocryphon, LAB and Josephus’s Antiquities, Alexander concludes that these works comprise a distinct literary genre based on a series of characteristics that define the rewritten work to its putative source(s). His strict generic approach emphasized the authority dynamics between works.

Following the complete publication of the Dead Sea Scrolls, the usefulness of the category of rewritten Bible was called into question, since it was used to describe an increasing network of heterogeneous Jewish literature. In response, a cacophony of terminological solutions was offered, including ‘parabiblical’, ‘parascriptural’ (Falk 2007), ‘pseudo-X’ (e.g., 4Q225 Pseudo-Jubilees), ‘paraphrase of X’ (e.g., 4Q382 Paraphrase of Kings), ‘exposition on X’ (e.g., 4Q464 Exposition on the Patriarchs) or ‘commentary on X’. George Brooke subsumes many such works under the rubric of rewritten Bible, defining them as works ‘which follow closely their scriptural base text and which clearly display an editorial intention that is other or supplementary to that of the text being altered’ (2000: 778). For him, rewritten Bible ‘is a general umbrella term describing the particular kind of intertextual activity’ (2000: 780; 2014: 119-35). The fluctuation between genre, process and authorization continued.

It is this growth of the concept following the publication of the Cave 4 material against which Moshe Bernstein reacts, swinging the pendulum back towards genre. ‘When “rewritten Bible” becomes a process rather than a genre’, he says, ‘much of the value of Vermes’s tight descriptive classification has been lost’ (2013: 1.46). Although I am not sure how ‘tight’ Vermes’s initial description was, Bernstein carefully distinguishes between ‘rewriting the Bible’ as an activity and ‘rewritten Bible’, preferring a generic definition. Interestingly, despite his appeal to Vermes, he omits works that Vermes considered central, like the medieval Sefer ha-Yashar, which, according to Vermes, ‘fully illustrates what is meant by the term “rewritten Bible”’ (1973: 95).

Recently, the question of genre has received a more nuanced analysis by Molly Zahn (2012), who points to three features that characterize rewritten scripture. First, scribes who produced these works perceived their compositions as ‘doing something different from scribes who produce[d] expanded copies of biblical books, even though both sometimes use[d] similar techniques of revision and rewriting’ (2012: 282). Second, rewritten scripture is an interpretive endeavour. Not all interpretive texts are rewritten, but all rewritten works are interpretive. Third, the authors of rewritten scripture intended their works to be conceived of as the same ‘kind of text as the text that formed the basis of the rewriting’ (2012: 284). Zahn’s approach helps to define rewritten scripture as a phenomenon distinct from, but integrally related to, other works that retain interpretive elements, while at the same time acknowledging the diachronic flexibility of genres. Her approach acknowledges that the exegetical processes undergirding the production of rewritten works are basic to a plethora of early Jewish literature. The genre/process debate continues unabated.

In addition to this contentious debate, authority and the relationship between reworked and reworking text has remained a consistent point of discussion. Michael Segal, among others, has argued that rewritten works are designed to usurp the authority of their base traditions (2005: 11). Reworking is not only an interpretive activity, but an activity that imbues the re-presented interpretation with authority; rewritten works do not replace their antecedents, since such a move would be ultimately self-defeating. One rewrites out of a place of anxiety, and authority is the overriding concern for crafting a work indebted to antecedent literature (cf. Mroczek 2016: 56-68).

Not all agree that authority works in this mono-directional way. Brooke construes relationships between rewritten works and their base traditions as analogous to Genettian hypertexts and hypotexts (2010). Hypertexts do not merely usurp authority, but bestow it upon their hypotexts – reworking a text is an acknowledgment or assertion of a base tradition’s authority. The conferral of authority is a reciprocal process. In other words, ‘hypertexts enable the authentic renewal of their hypotexts’ (2010: 52) as part of a broader process observable in all manner of early Jewish literature.

In all of this, it is clear that rewriting in all its complexity was a fundamental aspect of early Jewish literary composition. This reality has led to a series of wider questions that remain at the forefront of critical discourse, particularly as it relates to the authority relationships between texts. It should come as no surprise that the NT is indebted to similar processes of rewriting, some examples of which may satisfy even the rigorous generic standards advocated by Alexander. The Gospel of Matthew’s relationship to Mark offers a brief example that assists in rethinking the underlying motivations behind rewritten works in a way that transcends concerns for authority conferral.

Matthew’s Use of Mark

Matthew’s relationship to Mark has only rarely and recently been contrived in terms of rewrittenness (e.g., Doole 2013; O’Leary 2006; Luz 2005: 18-36). Nonetheless, Matthew expands, omits and reworks Markan material in ways commensurate to rewritten early Jewish works, although his editorial interventions are on the more conservative side of this spectrum.

The obvious commonalities shared by Matthew and Mark signal their interrelatedness. Matthew generally follows the serial order of Markan narrative, transposing some events and adding a significant quantity of non-Markan material. The similarities of their plots are further supported by their shared climax (passion narrative) and resolution (resurrection). The many small-scale editorial changes that Matthew makes to Markan material are easily observable by comparing their texts at any number of pericopae.

What distinguishes Matthew from Mark, along with Matthew’s editorializing, is its many larger-scale additions, differences that are often interpretive in nature, especially at the beginning of the work. The Matthaean genealogy (1.2-17), birth narrative (1.18-25) and Herod-Magi-Egypt episode (2.1-23) are obvious additions to the Markan narrative (cf. Lk. 1.5-80; 2.21-38, 41-52). Parallels to Mark’s narrative appear in Mt. 3.1-6 with the manifestation of John the Baptist and the quotation of Isa. 40.3 that corresponds to Mk 1.2. Matthew’s pre-baptism material leaves the arrangement of the beginning of the Markan narrative otherwise undisrupted, although not its wording or narrative voice. The Matthaean Isaiah quotation (Mt. 3.3) omits the first part of the Markan quotation (Mk 1.2) from Mal. 3.1 and places the quotation in the mouth of John the Baptist instead of the disembodied voice of the narrator. John is a more active character in the Matthaean pericope. His first person speech in Mt. 3.2 defines his message directly in contrast to the third person narrative report in Mk 1.4. Matthew 1–3 illustrates both the expansive and reworking components of his literary sensibilities.

Matthew’s expansive approach to the beginning of the narrative functions exegetically, offering a solution to the terseness of Mark’s introduction and correcting a potential implication that Jesus’ messianic status goes back only to the baptism (Yarbro Collins 2007: 150). Beyond the mention of Jesus in Mk 1.1, there is little to contextualize his appearance at the Jordan. He becomes the main character following his divine acclamation as ‘son’ (1.11), but the lack of context for Jesus’ appearance assumes that the audience has a deeper reservoir of extra-textual knowledge regarding Jesus. Matthew’s expansions contextualize Jesus’ pre-baptism life by connecting his lineage to the story of Israel, addressing an ambiguity from Mark with a concrete story. Matthew 2 also backgrounds Jesus’ ministry in the context of the politics of Judaea and foreshadows Jesus’ royal identity in the Magi’s question (‘where is the child who has been born king of the Jews?’; 2.2) that culminates in the sign hung on the cross (‘This is Jesus, the King of the Jews^’; 27.37). The Matthaean expansions fill gaps in Mark, namely its abruptness, lack of context, and potential insinuation that Jesus’ sonship only takes shape at his baptism (cf. Peppard 2011: 93-131). The expansions connect the outset of Jesus’ ministry more concretely to the story of Israel and foreshadow key themes and events that dominate the discourse of Jesus’ depiction, including the nature of his sonship and his kingship (cf. Mt. 2.16-18).

Matthew is also expansive in the resurrection account. The additions address, again, potential problems with the Markan narrative, which are also engaged through redactional changes in shared material. The Pharisee’s pre-emptive attempt to head off possible claims of resurrection (Mt. 27.62-66) addresses rumours surrounding Christian claims to Jesus’ resurrection. The additional material in Matthew apologetically strengthens Christian assertions – the Pharisees set the bar for false claims much higher, and in so doing ironically sowed the seeds of Christian exceptions to the claim that Jesus’ disciples had stolen his body.

Matthew also provides an explanation for the rolling away of the stone, claiming that an earthquake precipitated by an angelophany was responsible (Mt. 28.2). Mark does not explain the event. These expansions anticipate objections to the Markan version, explaining in fuller detail the mechanics of the resurrection.

Jesus’ resurrection appearances in Matthew are also expanded in relation to Mk 16.8, which concludes with the fearful inaction of the women who meet the young man in white. After the women in the Matthaean version run to tell the disciples the news, Jesus himself meets them on their way. They worship him, and he encourages them to tell his brothers to go on to Galilee (Mt. 28.9-10; cf. Jn 20.15-17). The narrative then returns to the Pharisees’ attempt to secure the tomb (Mt. 28.11-15). The guards report the events to the chief priests. The Pharisees bribe the guards, telling them to spread the rumour that his body had been stolen. The additions about the guards at the tomb and their shadiness (Mt. 27.62-66; 28.11-15) address a lacuna in Mark’s account that leaves open the possibility of resurrection fraud. Might Mark’s lack of an explanation for the rolling away of the stone imply that it was the work of his disciples? Matthew categorically rejects this potential reading. The Markan Jesus never makes an appearance in the short ending, we are not told how the stone was rolled away, and the appearance of the young man in the tomb is ambiguous. Matthew increases the explicitness of these events.

Furthermore, the response of the women in Mark is contrary to the young man’s request that they go and tell the disciples, despite the ways that the various endings of Mark repair this rupture. Matthew’s reworking and addition of subsequent narrative confronts implied concerns relating to the nature of Jesus’ resurrection that might make fodder for opponents of early Christian claims. Matthew’s expansions are interpretive insofar as they explain in more coherent detail the resurrection account and have an apologetic bent against resurrection deniers.

The final addition in Matthew’s resurrection account, if we construe Mark’s long ending (16.14-18) as a later composition, is the great commission (28.16-20). Jesus’ declaration that ‘All authority in heaven and on earth has been given to me’ (28.18) corresponds to the Magi’s question to Herod in 2.2 about the location of the ‘king of the Jews’. Jesus’ kingship, anticipated in Mt. 2, is realized. The commission also offers a concrete action for readers to take in response to kingship, commanded by Jesus himself. The short ending of Mark leaves it to the reader to interpret the proper response to the narrative, while Matthew offers a precise course of action.

Matthew, Mark and Rewritten Scripture

This brief analysis of Matthew’s supplementation of the Markan narrative suggests that Matthew can be characterized, to a degree, as a rewritten version of Mark. On the one hand, the relationship between them can easily be described using the compositional language often used to describe rewritten scripture (cf. Zahn 2011: 17-18). It is obvious that Matthew contains additional material not present in Mark, some of which is taken from other traditions. Matthew’s rewriting is archival, grafting in a compendium of pre-synoptic traditions onto Mark’s basic framework. Many of these additions also neutralize potentially problematic interpretations of Mark.

Likewise, Matthew omits material. For example, Mt. 12.4 omits from the Markan version of the grain plucking disputation the notice that David entered the house of God and ate the bread of the presence ‘when Abiathar was high priest’ (2.26). Another example occurs in the ointment incident in Simon the Leper’s house at Bethany (Mt. 26.6-13//Mk 14.3-9). The Matthaean version omits the details that the ointment was ‘pure nard, very costly’, that the jar was broken, and that the woman was reproached (Mk 14.3-5). Small-scale omission is an omnipresent feature of Matthew’s reworking.

Matthew also alters Markan material in multiple ways, sometimes dividing a Markan saying between two new contexts (e.g., Mk 9.43-48 in Mt. 5.30 and 18.8-9) or replacing a verb indicating direct speech with a synonym (e.g., Mt. 21.24//Mk 11.29). The compositional facets that are inherent to rewritten scriptural works are also central to Matthew’s composition.

However, an alternative way to think about Matthew’s relationship to Mark problematizes the concept of rewritten scripture. Beyond the point that Matthew is not presented as related to Mark – it is not self-consciously rewritten – Matthew also incorporates and interprets a large body of non-Markan material. The status of these non-Markan traditions vis-à-vis Mark is difficult to discern, but the fact that Matthew extends his Markan core suggests that to describe Matthew as a rewritten form of Mark is only partially true. In this sense, Matthew functions as a literary archivist, shaping an array of Jesus traditions around a Markan core. Matthew, like Ben Sira, who describes his own literary activity using the metaphor of an overflowing river of tradition (Sir. 24.25-34; 39.1-8), ‘does not understand his work as a “book” in the sense of an original and final written composition, but as the malleable and necessarily incomplete continuation of a long tradition of revealed wisdom’ (Mroczek 2016: 93). The difference is that, instead of wisdom material, Matthew is concerned with Jesus tradition. The point of Matthew’s activity is not to replace, rework, rewrite or even interpret Mark – although these activities are central to his mode of composition – but to offer a narrative fleshing-out of a thicker body of Jesus tradition.

Beyond the compositional features of rewritten scripture, Matthew also corresponds suggestively to previously theorized generic categories (e.g., Alexander 1988; Zahn 2012). But, while Matthew’s adoption of the form of Mark is an important part of his compositional strategy, this choice of genre is not something that Matthew was constrained to choose because the evangelist thought that he was relating to Mark via the formal genre of ‘rewritten scripture’. Matthew’s engagement with Mark indicates that rewriting was an active part of the process, but not to an extent that it constitutes its own generic category.

However, while their self-designations highlight their presentational differences – Mark is an εὐαγγέλιον (1.1), Matthew is a βίβλιον (1.1) – Matthew and Mark are the same kind of work. The similarity of governing voice and the general contours of the plot show even more consistency than classical rewritten works. There is no persistent perspectival play like the Genesis Apocryphon, where parts of the Abraham story are narrated by Abraham in the first person (e.g., Gen. 13.1-18; Gen. Apoc. 20.33–21.22), and there is no overarching framing device that articulates a formal break from the base tradition like Josephus’s Jewish Antiquities, in which he claims to write a ‘history’ (ἱστορίας; 1.1) or ‘ancient record’ (ἀρχαιολογίαν) translated ‘from the Hebrew records’ (1.5). The myriad of formal differences between works classified as ‘rewritten scripture’ suggest that a generic category is more challenging to define than the process of rewriting.

In essence, the composition of Matthew is similar to Segal’s model of the composition of Jubilees (2014; cf. Mroczek 2016: 139-55). Segal envisions a compiler-redactor who incorporated extant sources, juxtaposing them with legal and chronological material and leading to logical coherence issues between these sections. The ‘composer’ also writes an introductory framing narrative for the entire work, incorporating a range of Enochic traditions. Matthew too reframes the Markan narrative, adding material at the beginning and end of his adopted framework. He also incorporates a body of pre-synoptic material, but generally does not redistribute this material across Mark, inserting it instead in blocks of mostly first person didactic speech and narrative framing (e.g., Mt. 5.2–7.27; 10.5-42; 11.7-30; parts of the parable discourse in 13.1-52; 18.10-35). Like Segal’s ‘composer’, Matthew edits portions of his received traditions in quantitatively small-scale ways. Also, like Jubilees, the process of reworking and composing did not end. Segal argues that two of the Pseudo-Jubilees fragments (4Q225-226) represent continued rewritings of Jubilees. Further literary reflection on the narrative of Jesus’ life continued unabated, most immediately in the canonical gospels of Luke and John. Matthew is but a point in a more dynamic tradition of transmission and composition, even though he does not explicitly articulate his location with this stream. Like Jubilees, which consciously places itself within a long history of textualized revelation and which obviously does not engage nakedly with Genesis, Matthew’s conceptual library is much larger than Mark (cf. Mroczek 2016: 143).

The Gospels and Thinking about Rewritten Scripture

Can this analysis help us to overcome the overemphasis on the instrumentality of authority conferral and the intractability of genre/process conversation? Maybe. First, Matthew’s use of Mark shows that the definition of rewritten scripture as a genre is problematic. The exegetical processes and interpretive moves that define these works are ubiquitous across ancient Jewish and Christian literature to different degrees. Moreover, when their relationship to their source traditions are not considered part of their generic identity, the formal features of rewritten works are diverse. It is better to conceive of rewriting as an activity that offers access to streams of tradition, providing opportunities to partake in revelatory discourse. Matthew’s reworking of Mark, for example, is interpretive insofar as it gap-fills, but it is more than this. Matthew fundamentally alters Mark’s presentation of Jesus by combining his edited Markan core with other traditions. Matthew considers non-Markan Jesus traditions as part of the same discourse that Mark presents. Luke states this explicitly in his prologue (1.1-4). For Luke, revelatory information pertaining to Jesus is not restricted to textualizations, but located in the memory of eyewitnesses and their descendants. Matthew’s use of Mark highlights the problematic nature of rewritten scripture as a genre, but emphasizes the productive potential of rewriting as a compositional and potentially revelatory activity.

Matthew’s composition is deeply indebted to a range of sources. His use of Mark solidifies the idea that implicit exegesis is a central characteristic of Jewish compositional praxis, not simply rewritten scripture. But something more is at play than the filling of gaps and games of bibliographic authority. What underwrites Matthew’s reworking of Mark is the idea that Mark did not have a monopoly on revelatory traditions. The totality of Jesus tradition was not localized within Mark, and the importance of the character generated more textual material about him. Mark’s presentation of Jesus was contingent. Matthew too does not claim comprehensiveness in his work. NT authors recognized that the totality of revelation is something that one text cannot contain, nor is revelation defined purely in textual terms.

And this is the point where we begin to move beyond strategies of authorization as the driving force for producing rewritten scripture. Surely there are historical literary relationships between works – Genesis is anterior to Jubilees; Mark is anterior to Matthew – and these create hierarchical relationships between texts. But the question of one or the other’s scriptural status or attempt to gain authority are not necessarily key to discerning the underlying motivation for this type of engagement. Instead, Matthew’s use of Mark emphasizes the fact that even works considered authoritative or scriptural for particular communities were not idealized as comprehensive in their scriptural content.

Mroczek’s recent work on the Psalter is helpful in this regard (2016: 33-85). Her suggestion that the tradition that David prophetically composed 4,050 psalms, inspired by a ‘discerning and enlightened spirit’, underwrote the variety of psalm collections at Qumran is suggestive (11QPs^a 27.4-11; cf. 1 Chron. 28.11-19). The sheer quantity and inspired nature of psalms in the Davidic tradition endorsed the variety of arrangements and the on-going composition of other psalms and liturgical traditions, not unlike the writings that become associated with Enoch (Jub. 4.17-24; 10.17; 1 En. 12.4; 14.1-4; 39.2; 40.8; 72.1; 82.1; 83.1-2), Noah (Jub. 10.10-14), Solomon (1 Kgs 5.12; Josephus, Ant. 8.44), Ezra (4 Ezra 14.44-47) and Moses (Jub. 1.4-5, 26-2.1), especially when these writings are associated with angelic revelation (cf. Mt. 4.11), heavenly tablets (e.g., 1 En. 81.1-2; Jub. 1.29) or divine inspiration. Although collections of psalms long pre-date the Second Temple period, these compositions were not considered to contain the fullness of the Davidic tradition. Scribes continued to write psalms, rearrange material based on various thematic perceptions (e.g., 11QapocrPs in which Ps. 91 concludes four songs are related to exorcism rituals) and rework particular psalms (e.g., Ps. 151; cf. Segal 2002), locating their activity within the Davidic tradition. Some psalms also migrate between textual oases (e.g., 11QPs^a 21 and Sir. 51.13-19).

The growth of the Davidic psalm tradition, organized around a particular inspired figure, is related to the growth of the synoptic tradition. The fact that a significant body of Jesus tradition existed beyond the confines of the textual gospels motivated the expansion of the narratives due to their association with Jesus. Because Mark presupposes that his readers know more about Jesus than he records, licence is given to the Matthaean evangelist to incorporate these other traditions around the Markan framework. The existence of a greater body of Jesus material begs further development. The Gospel of John comments on this idea in its final verse (21.25), granting licence to later writers to continue to develop Jesus traditions: ‘But there are many other things that Jesus did; if every one of them were written down, I suppose that the world itself could not contain the books that would be written’.

Just as David’s inspired 4,050 psalms underwrite ongoing literary activity, so too does the idea that the gospels are only selective representations of Jesus’ activity. Rewriting offers opportunities to partake in inspired discourse, be it Davidic, Torah-centric or Christocentric. The question of authority is tied into participation in these discourses, but the quest to gain or confer authority is not the only driving force behind the act of rewriting.

Regardless of how one reads the evidence from the gospels, it is clear that the gospel tradition adopted a number of compositional proclivities from early Judaism. The gospels should feature more fully in discussions about the dynamics of transmission, exegesis, literary growth and rewriting in early Jewish texts. Synoptic problems are not unique to the gospels but are endemic to early Jewish textual culture.