Chapter 2. Biblical Rewriting and the Metaphrastic Habit: The Life of Thekla within the History of Ancient Paraphrase

Prologue: Erasmus and the Conflict over his Paraphrases on the New Testament

For a paraphrase is a plain setting foorth of a texte or sentence more at large, with such circumstance of mo [i.e. more] and other wordes as maie make the sentence open, clere, plain, and familiar whiche otherwise should perchaunce seme bare, unfruitefull, hard, straunge, rough, obscure, and derke to be understanded of any that were either unlearned or but menely entreed [i.e. entered=instructed]. And what is this, but a kinde of exposicion, yea and that of the most pithie and effectuall sorte?
This quotation is taken from Nicholas Udall’s introduction to the English version of Erasmus’ voluminous Paraphrases on the New Testament (1548 and 1551–1552), for the first volume of which Udall (“poet, playwright, and sometime headmaster of Eton college”) was the general editor. [1] For Udall, the biblical paraphrase was a helpful guide to the hard places of Scripture; this manner of exposition, as he says, could provide a real sense of the meaning of the Bible—which it apparently lacked on the surface—for those with only a basic level of education. Erasmus’ Paraphrases were not well received in France; the original Latin edition was condemned in 1527, shortly after its first printing. [2] By contrast, the Paraphrases were very well received in England and had a discernable impact on early Anglican exegesis and preaching; they have {67|68} also been found, perhaps more tellingly, in the library records of hundreds of English parish churches. [3] Therefore, Udall’s English edition would appear to have achieved its aims, and, at least in light of its reception, his general assertions about the usefulness of biblical paraphrase were well founded with regard to his contemporary Anglican audience.
As a recent volume edited by Hilmar Pabel and Mark Vessey amply demonstrates, Erasmus’ Paraphrases were intimately connected with the development of biblical criticism and printing in the early sixteenth century (Pabel and Vessey 2002). The reception of the paraphrases in Reformation Europe, they argue, should be understood from the point of view of the complex intertextuality of the paraphrases and their translations, in which Erasmus’ theological commitments and his affected literary rhetoric interweave in a striking fashion. From their initial conception, the Paraphrases were Erasmus’ attempt to offer, within the biblical text itself, some conclusions of Renaissance exegesis—eventually in octavo (i.e. pocketbook) format—so that (even casual) readers could glean the benefits of advanced biblical criticism. But as Pabel and Vessey point out, the dissemination of the Paraphrases was not without interesting twists and turns. As already mentioned, Erasmus from the start met with criticism in France for changing the ipsissima verba of the New Testament and introducing foreign (if enlightened) comments into God’s Word. [4] Whereas Erasmus claims to have intended that the Paraphrases assist believers and congregations in understanding what the Bible really said, as Udall asserted in his translation, several scholars at the time of their publication contended that biblical paraphrase only confuses the reader and, what is more, it adulterates Scripture by introducing ideas that are merely human and thus not divinely inspired.
Erasmus’ method of paraphrase would be striking in itself, even if its reception history were less controversial. For each New Testament book he paraphrased, he tried to be conscious of the persona of the author: of each author’s style and syntax, of course, but also of the character that tradition had assigned to him. The most interesting example of this method comes from his paraphrase of the Gospel of Luke, to which he prefaced a dedication to Henry VIII. [5] Erasmus presents Luke’s Gospel as a drug or medicine that, when {68|69} taken (i.e. read), heals the effects of sin and death in the patient. He links this metaphor directly to the tradition of Luke as a physician-historian and emphasizes that the efficacy of the Gospel is based on its historical veracity, as vouchsafed by this educated companion of Paul. Erasmus thus speaks, in his words, sub evangelistae persona, taking for himself the traditional characteristics of Luke, the style of his narrative, and his medical authority. Moreover, the two prologues to Theophilus in Luke-Acts provided for Erasmus an exegetical “space” in which to create an audience, not just Henry but all his readers. As Vessey points out:
Unlike normal commentary, which always declares its supplementarity with respect to the source-text, even when the commentator is merely explicating one biblical passage by another, paraphrase stands up—in its first-person, Erasmian mode, speaks up—in the name of Scripture itself. [6]
In this way a personified “Luke-voice,” Erasmus’ biblical ego, offers the paraphrast an opportunity to discuss the very nature of Scripture: how it works, what it lacks, and what it means; paraphrase becomes Scripture explaining itself, defending itself, and claiming itself. [7] Thus, despite sixteenth-century Catholics in France who resisted Erasmus’ free play with the text, the theoretical force of Erasmus’ Paraphrases is monumental both in their reception among reformed Anglicans as well as, on a more general level, in what they say about the perceived ontology—the malleability—of the Bible.
This debate was also, of course, about the freedom of readers and, consequently, about the degree to which textual criticism should have a say in how the Bible is read. These latter are primarily Reformation issues, which we have inherited, but I would venture to suggest that the battle over Erasmus’ Paraphrases can be seen, in its essentials, as a battle for cognition as much as for canonicity, since the central question at stake is what, precisely, is necessary for a reader’s understanding when he apprehends received texts like the Gospel of Luke? Theories of scriptural cognition are common to all generations of Bible readers, and in this way the key issues in the production and reception of Erasmus’ Paraphrases can potentially be detected in every paraphrase ever written. {69|70}

Towards a Modern Theory of Paraphrase: Goody, Alter, and McKenzie on the Mutability of Texts

As a way of defending Erasmus somewhat against the charge of literary invention, it is important to make clear—Pabel and Vessey do not attempt this—that biblical paraphrase has an ancient and revered tradition within Judeo-Christian literary history. [8] In choosing to invest so much effort in paraphrasing New Testament books, Erasmus could very well have been inspired by a number of ancient Jewish texts that famously included paraphrased Scripture, such as the Septuagint, Josephus, or the late antique Targums in Aramaic. [9] There is also evidence that he was aware of some early Christian paraphrases of biblical books, which were less numerous but definitely in circulation (in print) by 1520. [10] Therefore, keeping Pabel and Vessey’s examination of Erasmus’ Paraphrases in mind, I would like to consider in this chapter some of the broader issues of biblical paraphrase as it was practiced in antiquity.
I do this in order to contextualize the close reading of the Life of Thekla which I presented in the last chapter. As a late antique paraphrase in Greek, the Life appropriates the paraphrase tradition as it had been practiced for centuries before and was also currently employed in both Jewish and Christian circles. This tradition clearly flourished at least as much in the eastern Mediterranean as it did in the West, and the eastern side of the tradition {70|71} was spurred on in late antiquity by the strong influence of Hebrew exegesis. However, most of the important issues raised by the contributors to Pabel and Vessey 2002 have never been addressed in a late antique setting, even though late antiquity was precisely when the eastern Christian tradition of biblical paraphrase was coming into its own. Therefore, in the following sections of the present chapter I attempt to present a brief and selective history of this tradition, beginning with the evidence from the Hebrew Bible and ending with the fifth century AD, when the majority of our earliest (extant) Greek Christian paraphrases were written. I shall also make a brief comparison between this first flowering and the apex of Byzantine paraphrase in the tenth century.
The questions must be addressed at the start, however, of what a paraphrase is and how it seeks to represent the text that underlies it. One could argue that a unique theory of paraphrase must be generated by each individual author, given that the underlying text, its Vorlage, can be so determinative of the character of the paraphrase, or “hypertext.” [11] One could likewise argue that the socio-cultural contexts of specific paraphrases and rewritings forbid any kind of synchronic examination. To be sure, examples occur in the history of paraphrase that suggest there are interpretive barriers of this kind, and out of regard for such concerns I attempt below to draw attention to the characteristics particular to each paraphrase examined. Despite this need for literary atomism, recent models of textuality and cognition—from the related fields of anthropology, literary criticism, and bibliographical studies—offer opportunities for rewriting to be seen as a project common to human experience and not limited to any one historical or cultural sphere.
In particular, models of “literary” elaboration in oral cultures can, I suggest, provide some help in attempting to analyze the evidence of ancient paraphrase and rewriting. At the head of recent research on literature in oral societies is anthropologist Jack Goody, whose conclusions have become standard fare for anthropologists, as well as for those working on Renaissance book culture, the transition to print, and modern information networks. [12] Goody’s numerous publications focus mainly on tribes in West Africa among whom versions of the Lo Dagaa myth of the Bagre were still being recited. [13] Some {71|72} of these tribes had set down written versions of the myth, thus providing a testing ground for explaining the oral-to-written transition. Goody argues (persuasively, for many) that, while the writing-down of myths seems to limit their elaboration, variations among oral versions of myths are actively encouraged. These variations are seen as parts of an ancient whole: the individual teller of oral myth, even if patently inventing a new tale, often sees himself as recovering the lost knowledge of his ancestors. “The Speakers, even at the moment of creation, think of themselves as recovering the irrecoverable.” [14] By contrast, when myth is put into text (and only then) variations from it are consistently seen as heterodox. For Goody, cultural memory is thus essentially oral: a vast storehouse of social awareness passed down and elaborated upon by each successive generation. Oral variation is a sign of vitality, whereas the printed versions tend towards stagnation:
The myth was in a perpetual state of transformation. So we have an infinity of oral versions of the Bagre, which in practice the actors find difficult to compare. But there are now two printed versions, which unfortunately some have begun to take as the truth, as orthodoxy, because of the prestige of writing and because they had been recited by ancestors now dead. A new measure of truth, a new concept of archive, has emerged. [15]
Goody’s concept of the decadent “archive” is set in explicit opposition to Jacques Derrida’s program of textualizing the spoken word. [16] And Goody insists on the autonomy of the oral in the face of the post-structuralist project to see “inscription” as pervasive, even in illiterate or semi-literate societies. For Goody this is “an irresponsible attitude towards words” and cannot account for the variation found in oral “texts.” [17]
Goody accepts a basic “textuality” to social self-definition and power—indeed, he has proudly pointed out that anthropology, in his estimation, {72|73} anticipated the ideas of Derrida and Michel Foucault by a few years. [18] However, Goody’s “textuality” is still purely oral, in that he prefers to see the textuality of the written or printed word (in oral or semi-literate societies) as contributing less to the refashioning of literature, history, or social consciousness in general. At a basic societal level, Goody argues, writing cannot supplant oral tradition as a force of change. [19]
Goody, of course, does not view writing as a negative force; [20] he is trying first of all to explain the “interface” between the written and the oral in order to better understand the cognitive processes involved in the transition from one to the other. While his conclusion is that writing and “archives”—the momentary (and thus blinkered) capsulations of a constantly fluctuating discourse—inhibit the creativity of oral literature, he has said many shockingly “textual” things about the character of literary variation and revision within the oral sphere. He has isolated, for example, a trend towards antiquarianism, or “scholarship,” among oral composers for whom the language of the people and the language of their tales has begun to diverge. [21] He has also pointed out the tendency for the establishment of an oral canon to stimulate further elaboration of the Lo Dagaa Bagre myth among the tribes he studied.
The picture that emerges from Goody’s writings is of a vibrant, unencumbered “textual” culture, that is constantly revising its own “textual” history, encouraging the extension and “rewriting” of the oral myths in every generation. However, he rarely moves beyond the (mainly) oral evidence of the Bagre that he so painstakingly accumulated. By way of extending and problematizing Goody’s seminal analysis somewhat, could it not be asked if there is any case where writing does in fact encourage the vibrant literary creativity that Goody has isolated in oral societies?
Paraphrase can arguably be seen as one example of this, and the brief history of paraphrase below seeks to provide evidence from the ancient and late antique Near East. Some studies have been made that already point to this conclusion. In his recent book on the reception of the Bible in modern literature, literary critic Robert Alter has shown (in contrast to Goody) how the {73|74} institution of a written literary canon almost invariably encourages further elaboration and creativity, in writing:
The imaginative response to the Bible of writers in a wide variety of languages bears witness to a power of canonicity that is not limited to doctrine or strictly contingent on belief in the inspired character of the texts invoked. [22]
The canon thus serves as a “vehicle” for imaginative literature, which takes its inspiration from Scripture in a wide variety of ways: by mimicking the “earthy” language of a culturally dominant translation (Faulkner); by “wrestling” with biblical self-interpretation (Kafka); by using biblical language to re-present modern paradoxes (Bialik); by “intricately coordinating Scripture with Homer” (Joyce). [23]
The commingling of biblical myth and Homeric epic is characteristic of fifth-century AD Greek paraphrase, as I shall demonstrate below, but it is appropriate to point out here that the attitude of classical writers to Homer and myth in general adds weight to Alter’s insights and further enriches Goody’s oral model. In her recent study of Literate Education in the Hellenistic and Roman Worlds, Teresa Morgan has gathered an impressive amount of papyrological evidence pointing to the manipulation of Homeric texts in a school context. “Texts oscillated between two statuses: that of the particular canonical version of the story, and that of a tool which could be used and altered.” [24] Homeric canon could be, therefore, a stimulant to literary activity on a very literal level: Morgan’s evidence consists of rewritings of individual words and phrases as well as the wholesale recasting of epic into both prose and verse. [25] On a wider view, the Homeric myths formed an imaginative world for ancient writers, a “site” on which they could play with an ancient, received literary history. Both the incidental details of the myths and the narrative holes left unfilled by the poet became opportunities for expansion. Indeed, as Froma Zeitlin has recently argued at length, ancient patterns of “traffic in Homer” were widespread and varied, often taking the form of imaginative (even visual) reconstructions of the myths and even the persona of the poet {74|75} himself. [26] In other words, nothing was out of bounds, and almost any aspect of received tradition could become the object of paraphrase. Yet the fact remains that the reception of Greek myth in the Hellenistic period by poet-scholars like Apollonius and Callimachus set an enduring pattern for Roman and late antique elaboration, and there is ample evidence that major Greek poets and paraphrasts of the fifth-century AD, such as Nonnus, were taking direct inspiration from their Hellenistic predecessors. [27] The example of Marianus of Eleutheropolis, an official at the court of the emperor Anastasius (491–518), offers a view from the crest of this trend: according to the Byzantine Suda encyclopedia, he wrote iambic paraphrases of the hexameter works of all the important Hellenistic poets—Theocritus, Apollonius, Callimachus, Aratus, Nicander, “and many others.” [28]
In Greek, the term for paraphrase one usually finds is μετάφρασις—less often μεταβολή. [29] While μετάφρασις is sometimes mistakenly rendered into English as “translation”—its meaning in Modern Greek—ancient writers typically invested more in the word than “translation” allows. Josephus, for example, parallels μετάφρασις with μεθερμενεύω, “to interpret,” when describing his own project of retelling the Hebrew Scriptures in his Antiquities. [30] Several Greek writers from the Roman period use the word in this way, and the meaning persists into the Byzantine period when μετάφρασις became a major literary project in its own right. [31]
What, then, is the ancient theory of μετάφρασις? How is a retelling to be understood that is not merely translation or re-presentation? On the basis of theoretical models from anthropology and literary criticism, I have suggested that paraphrase as can be seen as a method of imaginative elaboration. The elaboration is dependent on a canonical or received text, from which it takes inspiration and/or narrative material. The fixity of received texts used for rewriting and paraphrase is less important on a doctrinal or ideological level than on a cognitive one. {75|76}
Both Goody and Alter point to the fact that rewriting is necessarily concomitant with any reception of “text,” be it oral or written, especially when that text has taken on a dominant, self-defining role in a culture. Whether it is the Lo Dagaa Bagre myth, stories from the Hebrew Bible, or the vast Homeric and related mythologies of ancient Greece, human cognitive response invariably tends towards elaboration and rewriting, sometimes on a very literal level, as in Hellenistic school exercises. The received text naturally becomes, often without any external pressure, a “site” or a locus of rewriting and “play”: this play, of course, has as much to do with refashioning contemporary identity as it does with reformulating ancient mythology. [32] In paraphrase the two are inseparably linked; but this play, which Goody would describe as being at root a cognitive activity, occurs whenever a received text is altered, no matter how slightly.
In the context of the copying of ancient manuscripts and codices, Kim Haines-Eitzen has described this cognitive activity in the following terms: “Copying an exemplar meant producing a ‘resemblance’ not an identity.” And she goes on to quote Michel Foucault:
Resemblance has a “model,” an original element that orders and hierarchizes the increasingly less faithful copies that can be struck from it. Resemblance presupposes a primary reference that prescribes and classes. [33]
For rewriting individual words, as much as for paraphrase, the issue of “resemblance” is central: any alteration or elaboration of a received text depends on the original for its new “identity” and, at a secondary level, the re-casted text depends on a pre-existent discourse of classification and power. Exactly how the resemblance is constructed defines the inherent meanings of the text, and, from a structuralist point of view, also describes the society that produced it.
The fifth-century AD Life and Miracles of Thekla, the object of the present study, can provide a textual “place” in which to examine these issues. When this text had been previously studied by scholars, the complex issues of paraphrase and textual elaboration have not been addressed. By setting this half of the text in a literary historical framework—and then by doing the same for the second half in Chapter Four—I hope to be able to say more about how the {76|77} text works internally and, more importantly, how it relates to the culture that produced it.
In a similar vein, the bibliographer Donald McKenzie addressed the question in his 1985 Panizzi lectures of how “textual artifacts” should be treated in an age when the printed word threatens to overwhelm the human ability to process. [34] In a self-conscious attempt to redefine the vocation of bibliography for the new millennium—trying on the label “sociology of texts”—he insisted that scholars should pay close attention to the physical properties of the texts they study, since these properties can tell us as much about what the text means as can the intentions or ideologies that appear on its surface. McKenzie writes:
My argument therefore runs full circle from a defense of authorial meaning, on the grounds that it is in some measure recoverable, to a recognition that, for better or worse, readers inevitably make their own meanings. In other words, each reading is particular to its occasion, each can be at least partially recovered from the physical forms of the text, and the differences in readings constitute an informative history. [35]
and further:
If a history of readings is made possible only by a comparative history of books, it is equally true that a history of books will have no point if it fails to account for the meanings they later come to make. [36]
The present study, like McKenzie’s new bibliography, takes seriously the changes, the resemblance, of the Life and Miracles to its Vorlage, the ATh. Thus, I examined in detail in the previous chapter the elaborations made by the author in an attempt to highlight and further explain the literary nature of the Life on its own.
According to McKenzie’s model, however, the Life and Miracles would never have existed on its own and, consequently, must today be read through the history of books—more specifically in our case, the history of rewritings and paraphrases—that came before it. The brief history of rewriting that follows is not an attempt to excavate origins or to show direct influence. Rather, it {77|78} discusses the synchronic unity of paraphrastic activity through a diachronic survey. Additionally, I provide at the end a sense of how the tradition continued in Byzantium, post fifth-century, in order to compare an instance where paraphrase, μετάφρασις, became an epoch-defining literary project. Nevertheless, an argument already made explicit in this study is that rewriting (either oral or written) is a basic cognitive activity. Therefore, contrary to previous studies of Byzantine μετάφρασις that insist on its uniqueness to that culture at that time, my study will argue for its near ubiquity in Greek Christian literature. [37] Furthermore, I hope this study can contribute to the larger picture of the process of textual inheritance that has emerged from neighboring disciplines, such as scholarship on the medieval West and the early modern period.

The “Rewritten Bible” in Ancient Judaism

Christians were, of course, not the first to treat their own Scripture as a site of rewriting. Jewish literary history is particularly rich with paraphrases, and these appear from an early point. Deuteronomy, the ultimate expression of the Mosaic law in the Pentateuch, is largely a rewritten systematization of legal material from Exodus and Numbers. [38] A few centuries later, the postexilic author of the book of Chronicles not only drew material from the Pentateuch, Isaiah, Jeremiah, Ezekiel, and the Psalms, but clearly rewrote substantial sections of the earlier histories of Samuel and Kings, adding, subtracting, and summarizing according to the ideologies of Second Temple Israel. [39] The successful reception of Chronicles as a rewriting is attested, of course, by its subsequent inclusion in the biblical canon, but, interestingly also by its Greek name in the Septuagint, Παραλειπομένων (“the things left out” or “omissions”), a title which suggests that readers in the third to second centuries BC already recognized its unique relationship to Samuel and Kings. Thus, within the biblical tradition itself, rewriting was not seen as a banal or opportunistic activity but could be acknowledged as a legitimate, even “canonical,” form of literary endeavor. [40]
Just a few generations before the canon of the Hebrew Bible became fixed (c. 1st century AD), two trends emerged in its interpretative history. {78|79} One trend was to codify the accepted books (Deuteronomy and Chronicles included, of course) and to comment on the text externally, thus attaching a protected status to Scripture: this is the trend that ultimately resulted in the formation of a “Masoretic” canon as well as influencing the development of rabbinic exegesis. [41] (The mode of rabbinic midrash, i.e. lemma + commentary, also attests to this trend.) Another trend, however, was to continue to mix commentary with received text and thus to perpetuate the interpretive habit of “Rewritten Bible” established by the authors of Deuteronmy and especially Chronicles. [42]
Within this latter tradition, the copyists and commentators of the Dead Sea Scrolls community chose to rewrite biblical books according to the sectarian eschatological vision of their Teacher of Righteousness. The manuscripts found at Qumran are overwhelmingly biblical in their orientation: only one major text (the Copper Scroll) is not a biblical manuscript or a work based on Scripture. And every book of the Hebrew Bible was found there, either complete or in fragmentary form. [43] But the biblical texts are not identical with the Masoretic versions: they show a tremendous amount of variation, even between themselves. [44] In addition to these individual changes (both conscious and not) to the biblical text, a striking feature of the Qumran exegetical literature is its extensive interweaving of Scripture and comment on the page, to the degree that often the commentary seems to become Scripture. The fragmentary Genesis Apocryphon, a very loose paraphrase, is outstanding in this regard. [45] Surely, this technique (called pesher in its standard Qumranic form) is where some of the scribes’ own biblical interpolations originated, but what is striking is how pervasive the habit of paraphrase seems to have been at {79|80} Qumran across the board. The scribes of this community clearly took a cognitive/interpretative position on Scripture different from those who were simultaneously working to make the Bible inviolable. At Qumran the received text of the Bible was a book susceptible to modification and elaboration, rather than the monolithic code it (more or less) became in rabbinic circles.
It is important also to keep in view the parallel history of the ancient translations of the Hebrew Scriptures, for these translations were at times paraphrases in their own right and can point to how the tradition of rewriting was received. [46] To take the most celebrated example, the translators of the Septuagint (hereafter lxx) significantly modified the original Hebrew text, adding large sections to certain books, despite the claim to accuracy put forward in the legendary Letter of Aristeas. [47] These changes were significant in particular because early Christian writers, including the authors of the books in the New Testament, used the lxx almost exclusively, thereby extending the life of the translators’ rewrites.
A recognition in antiquity of the changes made by the lxx translators is evidenced by the three important attempts in the first two centuries AD to bring the lxx back into line with the Hebrew. These are the so-called “minor versions” or simply “the Three”: Aquila’s literal translation; Theodotion’s less strict revision of the lxx; and Symmachus’ translation in fluid Greek. In addition to these three, Dominique Barthélemy published in 1963 the fragments of a slightly earlier Greek translation of the Twelve Prophets, which were found at Nahal Hever in the Judean desert. These fragments, dating to the first century BC, are part of a literal revision of the lxx that seems to have subsequently influenced Aquila. [48] The importance of these fragments lies in their showing that a decision to revise the lxx, because of its inaccuracies and elaborations, came even earlier than previously thought. [49] {80|81}
There is some scholarly disagreement, however, as to how the translation of the lxx itself should be understood in the context of these subsequent rewrites and translations. Barnabas Lindars, for instance, has argued that a translation of the Hebrew Bible, in whatever form, should be kept distinct from its rewriting in commentaries and the like:
The Septuagint is essentially a translation and not a targumizing paraphrase. There is a sense in which every translation is a commentary, or contains what might be called linguistic exegesis, because it represents the translator’s understanding of the text, and this is inevitably colored by the presuppositions of the time. But this is not the same thing as deliberate modification of the text for the sake of interpretation (which might be designated content exegesis). The aim of the translators of the Septuagint was to give a faithful rendering of the Hebrew. [50]
Lindars’s formula, however, depends on a strict a definition of translation; what is “essential” to any translation was contested in antiquity and continues to be so today. [51] If Lindars means the lxx Penteteuch alone, which is more literal than the other books, then he has some room for argument, but it is important to note that most scholars accept that the lxx is conceptually a paraphrase, and it was understood to be so in antiquity. It is because of this ancient understanding that I think it can be argued from a historical point of view that “content exegesis” and translation should not be so artificially separated. As already suggested, it is clear from the subsequent history of the Greek translations that the lxx was considered too loose and needed to be brought back into line, presumably because the “content” had been altered.
Despite the lxx’s prominence among Christian writers—even to the level of Luke’s imitation of its literary style—later Jewish translators were not satisfied with the text, and Origen, idiosyncratically sympathetic to the Hebrew original, famously put the later Greek versions in parallel columns with the Hebrew and the lxx in his Hexapla. [52] To quote a modern editor of the lxx: {81|82}
It is clear from the very arrangement of the Hexapla that to [Origen], being a scholar, not the lxx, but the original text was the primary authority, for he put the original text first, and then had next to it the translations of Aquila and Symmachus, since they furnished the most accurate renderings of the original text. [53]
In addition, Origen felt comfortable correcting the lxx when he saw fit, and Hexaplan variants have come down to us that are clearly Origen’s own interpolations. [54] Thus, certainly among biblical scholars in antiquity, there seems to have been a keen sense of the inadequacy of the lxx alone with regard to its accuracy—that is to say, some clearly did not consider it a “faithful rendering,” in Lindars’s phrase.
A very different approach to the lxx emerged concurrently with the more literal and idiomatic Greek translations. Jews writing in Greek in the first century AD, such as Philo and Josephus, continued the tradition of rewriting (expansively) their received Greek Scriptures. Louis Feldman has made a sweeping study over several years—in separate articles now collected in one volume (1998b) and also rewritten into a monograph (1998a)—of the rewriting, primarily of the lxx itself, made by Josephus in his Jewish Antiquities. Feldman observes that Josephus reworked biblical stories out of concern for certain factors, including style and narrative quality, the assumptions of his intended readerships, and historiographical tropes—though apparently not out of concern for the accuracy of the translation. Through his rewriting, Josephus emerges, according to Feldman, as “no mere copyist or compiler,” but, instead, “his own views—historiographical, political, religious, and cultural . . . are consistently seen throughout the Antiquities, particularly in the changes which he has made in his paraphrase of the biblical text.” [55] Josephus thus took a comparatively liberal view toward the lxx, introducing his unique vision of the history of the Jews within the biblical text itself, so that, like the Bible of the Qumran community, his paraphrase is an inseparable intertwining of text and commentary. Josephus in his Antiquities presented virtually a new Bible, at least in its historical account, and it is paradoxical that, while the original is all but invisible, to appreciate the argument, irony, and wit of his new text, Josephus’ readers even today must be very well acquainted with the original Scripture, in Greek at least, if not also in Hebrew and Aramaic. {82|83}
Despite his expansive inventiveness, however, Josephus staunchly defended the accuracy of the lxx (interestingly, through his recasting of the Letter of Aristeas) and claimed that he himself was only repeating what was in Scripture—both striking comments if one considers how central paraphrase was to his historiographical method. [56] Acknowledging this apparent contradiction, several scholars have pointed out that creativity with the biblical text does not generally entail a lack of respect for the Bible; on the contrary, the opposite is most often the case. The Aramaic targumim, for example, are paraphrasing translations, collected in late antiquity and the middle ages, which were originally made from the Hebrew, though which took on an authority of their own in Aramaic. [57] Unlike midrash, the targumim did not cite the original text but included interpretive material in the text itself, so that the reader or listener would hear only the recast version of the Hebrew: hence, like in Josephus, it is impossible to reconstruct the original text from the targum alone. [58] This is in direct contrast to the lemma + commentary mode of the midrash.
Despite the conceptual distinctiveness of paraphrase, Josephus and authors of the targumim depended on what they considered to be a stable, {83|84} authoritative text for their own rewritings. [59] The text Josephus used (primarily) was an interpretative, sometimes paraphrasing, translation in its own right, but it provided a textual “site” where Jewish writers of the Hellenistic and Roman East habitually played with the history and literature that they had inherited and, thereby, tried to make it accessible to a broader audience. In his Heritage and Hellenism, Erich Gruen has explored in depth this pervasive characteristic of Hellenistic Jewish writing:
For Hellenistic Jews writing in Greek, the Scriptures provided stimulus for ingenuity and creativity. The concept of a fixed and unalterable tradition had not yet taken hold. No scriptural “canon” existed. Composition and interpretation proceeded concurrently, and the idea of established texts was still in process of formation. The fluidity of the tradition may frustrate modern scholars. But it gave impetus to writers eager to reshape and revivify narratives long familiar but conveniently adaptable. [60]
Although Gruen is speaking here of a specifically Hellenistic context, the practice of Jewish paraphrase was at least as old as Deuteronomy and continued to be employed in the Roman and late antique periods. Furthermore, as I shall explain in the next section, there is ample evidence that the Christian tradition of biblical paraphrase emerged from this Jewish literary milieu.
However, before proceeding to the Christian paraphrases, there is one more group of Jewish texts that warrants attention—the Jewish novels. The works I have mentioned so far are primarily Scriptural in orientation, and I have suggested that this technique of paraphrase took its inspiration from Scripture—both from the canonical models of paraphrastic writing and, of course, from the literal source material with which the paraphrast worked. [61] Moreover, scholars do not normally see these paraphrases as attempts to {84|85} replace Scripture; rather, they represent a kind of homage to Scripture and its imaginative worlds. The extant Jewish novels, while more self-consciously fictional than historical paraphrases like Chronicles, were nevertheless read alongside the interpretative genres in the Hellenistic period and, when they deal with biblical scenes, can arguably be seen as a kind of expanded biblical paraphrase. [62] And the novels themselves went through numerous changes, as is attested by the different recensions that have survived, [63] with the result that the rewriting of the rewriting only further compounds our sense of the pervasiveness of paraphrase and textual elaboration or modification in ancient Jewish literature.
From the texts and fragments that have survived, the Jewish novel seems to have been a particularly successful medium for refashioning biblical stories. James Kugel has argued that, like biblical translators and paraphrasts, Jewish novelists rewrote the stories of the Bible (and added new ones) in response to specific difficulties they found in the text. [64] While this interpretation serves as a productive matrix through which to examine scriptural elaboration, it is probably just as viable to argue that Jewish novelists were inspired by a general flowering of fictional narrative in the Hellenistic Diaspora: works such as Tobit, the Greek Esther, Judith, and the novelistic extensions to Daniel (Bel and the Dragon and Susannah) belong to this tradition and were widely known in the late Hellenistic period. [65] These novels and their successors—Joseph and Aseneth, Artapanus’ On Moses, Third Maccabees—interacted with the canons of Greek literature more directly than biblical commentary and can perhaps be seen as cross-fertilizing the Greek Romance, which emerged concurrently. There is no doubt that Jewish novels owe a great deal to the Bible itself, but the {85|86} latter’s influence on the novel was less compartmentalized than on standard biblical paraphrase, such as that found in the targumim or Josephus’ Antiquities. Nevertheless, Josephus himself is the conveyer (in the Antiquities) of two historical novels—the Tobaid Romance and the Royal Family of Adiabene—and his juxtaposition of these with biblical paraphrase points directly to the crucial interpenetration of translation, paraphrase, and the novel in ancient Jewish literature.
An extensive treatment of the novels’ elaborations is not warranted here, but it should be noted that the novelistic literary style, on display above all in the five major Greek Romances—Chariton, Xenophon of Ephesus, Longus, Achilles Tatius, and Heliodorus—emerged in a potent climate of literary cross-fertilization, a climate in which a key player was Jewish fictional writing in the Hellenistic and Roman periods. [66] Furthermore, this style was, as I have suggested, intimately connected in Jewish literature to biblical rewriting, both from an authorial and an interpretative point of view. [67] Finally, the connection between these literary modes is perpetuated and expanded by Christians in their own tradition of biblical paraphrase. It is to this Christian evidence that we shall now turn.

Textual Elaboration in Early Christian Tradition: From Bezae to Homerocentones

The Gospels amidst Jewish paraphrase

Christians began rewriting their scriptures from the very beginning. The now standard “two-source” theory of gospel composition posits that the authors of Matthew and Luke both used Mark and “Q” (a lost “sayings-source”)—in addition to their own material—to construct the narratives of their Gospels. Mark and Q were, in the parlance of New Testament scholarship, “sources” or Quellen for the authors of Matthew and Luke. However, modern Quellenforschung (or the atomistic separating-out of these strands) seen in the context of the great amount of imaginative rewriting going on in Jewish circles in the first century AD appears, as a methodology, simply stultifying and one-dimensional. By contrast, I would like to try to see Matthew and Luke {86|87} as Christian examples of a habit of biblical rewriting that permeated Judeo-Christian literature in antiquity. [68]
Rather than looking for their sources and origins, it may be more helpful, considering the tremendous amount of evidence for ancient Jewish paraphrase, to see the Gospels as historical “sites” of rewriting where the authors were appropriating a recognizable method of literary activity within their immediate cultural and religious milieu. To be sure, this approach involves a shift of perspective, but it is one that pays dividends. This is true especially when looking at the way late antique prose narratives, such as the Life of Thekla, treat earlier Christian literature. The canonical Gospels and Acts became models for how Christian literature was supposed to be written—in language, style, and religious discourse generally—and, despite (or in conjunction with) the persistent influence of classical Greek literature through the educational system, these earliest Christian narratives took on for many later writers a mimetic authority. We have already glimpsed this in the Life’s invocation of Luke at the beginning of its paraphrase (see above pp. 18–21).
Looking more closely, however, Luke and Matthew are demonstrably not paraphrases—at least not in the traditional Jewish form exemplified by Chronicles, the Greek Esther, or Josephus. Bypassing summary and elaboration, these writers instead reorganize, moving snippets of Mark and Q around like puzzle pieces. Now that some Jewish examples have been produced above, this method can be brought into relief, especially for the sake of comparison with later Christian literature. On the surface, Matthew and Luke seem to be doing something different, but from a cognitive point of view, I argue, they are treating their source texts in much the same way, or at least producing similar effects on the reader.
Matthew and Luke do not approach Mark as a traditional paraphrast might because they do not see the first gospel as an ancient tradition: to put it differently, not only do we know them as the part of the first generation after Jesus, they recognize themselves as such. As Luke says in his prologue to Theophilus: {87|88}
Since many have undertaken to set down an orderly account of the events that have been fulfilled among us (πολλοὶ ἐπεχείρησαν ἀνατάξασθαι διήγησιν περὶ τῶν πεπληροφορημένων ἐν ἡμῖν πραγμάτων), just as they were handed on to us by those who from the beginning were eyewitnesses and servants of the word, I too decided, after investigating everything carefully from the very first (ἔδοξε κἀμοὶ παρηκολουθηκότι ἄνωθεν πᾶσιν ἀκριβῶς), to write an orderly account for you, most excellent Theophilus, so that you may know the truth concerning the things about which you have been instructed. [69]
I have already acknowledged above, in the Prologue to this chapter, the uniqueness among the evangelists of Luke’s creation of his own audience. What is also significant is that he claims to have (re)investigated the details again without using the “many” (Mark and Q?) who came before him. [70] He acknowledges to Theophilus that he currently has the investigative opportunity to return to “the very first”—presumably he means Jesus’ early life, which, of the four evangelists, only he discusses in detail.
By way of contrast, Josephus, in writing his Antiquities, recognized the temporal distance between himself and his textual site of rewriting and seems to have felt compelled to emphasize that he was changing nothing that he had received:
At the outset, then, I entreat those who will read these volumes to fix their thoughts on God, and to test whether our lawgiver [Moses] has had a worthy conception of His nature and has always assigned to Him such actions as befit His power, keeping his words concerning Him pure of that unseemly mythology current among others; albeit that, in dealing with ages so long and so remote (καίτοι γε ὅσον ἐπὶ μήκει χρόνου καὶ παλαιότητι), he would have had ample license to invent fictions (πολλὴν εἶχεν ἄδειαν ψευδῶν πλασμάτων). For he was born two thousand years ago, to which ancient date the poets never ventured to refer even the birth of their gods, much less the actions or the laws of mortals. The precise details of our Scripture records will, then, be set forth (τὰ μὲν οὖν ἀκριβῆ τῶν ἐν ταῖς {88|89} ἀναγραφαῖς προϊὼν), each in its place, as my narrative proceeds, that being the procedure that I have promised to follow throughout this work, neither adding or omitting anything (οὐδὲν προσθεὶς οὐδ’ αὖ παραλιπών). [71]
Elsewhere in the prologue, Josephus claims that Scripture narrated the history of “five thousand years” and that his Antiquities “will embrace our entire ancient history and political constitution, translated from the Hebrew records” (1.13). And as he says here, Moses would have had “ample license to invent fictions”: is this an ironic gesture to the knowing readers who would recognize that Josephus did anything but “set forth” Scripture “without omitting anything”? Is he employing historiographical convention to add humor to an already weighty prologue? Perhaps this is the case, since Gruen has demonstrated with numerous examples that self-reflective humor was characteristic of Hellenistic Jewish historians. [72]
What is important for the present argument is that, while Luke and Josephus conceive of their temporal distance from the textual site in very different ways—with implications for how they treat their source material—both use that textual “site,” the textus receptus, to invent a new narrative recognizably different from the original. Prescriptively they are very different but descriptively they are similar. Or, in other words, their approaches, while distinct in conception, nevertheless imply a similar cognitive angle on received texts, an angle which, I would argue, takes its inspiration from contemporary Jewish habits of rewriting more than from Greek historiographical conventions.

Close elaboration of the New Testament

The habit of rewriting penetrated much of early Christian textual activity, even if not in the style of a formal paraphrase: all of the New Testament Gospels betray some kind of recasting of their source material, and, as I have tried to emphasize, the prevalence of this activity reflects a wider Judeo-Christian metaphrastic mindset. Moreover, as might be expected given the evidence from Qumran, the subsequent copying of these early Christian texts was a particularly fervent locus of rewriting as well.
To take one significant and well studied case, the preeminent witness to the so-called “Western” textual tradition of the New Testament, Codex {89|90} Bezae Cantabrigiensis—a circa fifth-century bilingual (Greek-Latin), uncial manuscript—provides a large number of unique readings for the Acts of the Apostles. [73] In fact, the number of variants is so large that the “Western” text of Acts has been called “virtually an alternative version of the book.” [74] In 1966 the New Testament scholar Eldon J. Epp argued that many of these unique readings are conscious attempts to introduce into the text of Acts a rigorous anti-Jewish polemic. For example, the well known “ignorance motif” of the canonical Luke-Acts pair—which intimates that the Jews were not guilty of crucifying Jesus because they were “ignorant” of who he really was (e.g. ἀγνοήσαντες, 13:27)—is consistently written out of the Codex Bezae text. [75] The Jews are specifically held responsible in the rewritten Acts, and Christological terminology is re-designed to intensify the divide between Jews and Christians. [76]
Looking beyond Acts, Epp also points out the prayer of Jesus on the cross at Luke 23:34, Πάτερ, ἄφες αὐτοῖς· οὐ γὰρ οἴδασιν τί ποιοῦσιν (“Father, forgive them; for they do not know what they are doing”), is expunged from the Codex Bezae text. [77] In addition, there is a consistent “devaluation” of the Jewish element in Christianity and a “positive stress” on the uniqueness of the Christian universalism and the Holy Spirit. [78]
In the time since Epp’s seminal study in 1966, several scholars have found in Codex Bezae other examples of other conscious changes in the language, rhetoric, and narrative of Acts. Ben Witherington has delineated an “anti-feminist” strain in the Codex: at points where the faith of women is applauded there appears to be a coincident attempt to remind the reader of a “gender hierarchy”—γυναικῶν τε τῶν πρώτων οὐκ ὀλίγαι (“not a few leading women”) {90|91} at Acts 17:4 subtly becomes καὶ γυναῖκες τῶν πρώτων (“wives of the leading men”). [79] The text Codex Bezae represents is an important example of rewriting (or “close elaboration,” as I have termed it) in that it illustrates perfectly that the concept of textual malleability extended to the Christian copying of sacred scripture. [80] No doubt Codex Bezae was neither the first nor the only biblical rewriting, but its survival helps demonstrate that this activity is evident at a literal level in the codices of the Bible.
Taking inspiration from Epp and others who highlighted “theological” changes in the Bezae text of Acts, Bart Ehrman has attempted to situate these modifications within a competitive cultural milieu. [81] While Bezae itself probably originated in fifth-century Egypt, the text it contains is considered by most scholars to reflect a second or perhaps third-century textual tradition. [82] Ehrman has convincingly argued that the revisionist milieu of the second century offers the best interpretative matrix for the Greek text of Codex Bezae. [83] Especially with regard to Christological terminology and Jewish-Christian relations, Bezae is one dramatic example of a dominant mentality of rewriting that came to the fore in the second and third centuries. However, anti-Jewish interpolators were not the only ones rewriting the New Testament at this time. In response to Docetic, Ebionite, and other forms of Christianity deemed heretical by “proto-orthodox” apologists, the Gospels and Acts were often rewritten to further emphasize, from an orthodox point of view, the doctrinal differences between the heretical and orthodox sides. [84] For instance, against so-called “adoptionist” (e.g. Ebionite) readings of the Gospels that {91|92} argued for the human Jesus’ adoption as God’s divine Son only at his baptism, the well-attested reading of “You are my son; today I have begotten you” (Υἱός μου εἶ σύ, ἐγὼ σήμερον γεγέννηκά σε) at Luke 3:22 was changed by proto-orthodox scribes to read “You are my beloved son; in you I am pleased” (Σὺ εἶ ὁ υἱός μου ὁ ἀγαπητός, ἐν σοὶ εὐδόκησα). The latter is exactly the text of Mark 1:11, with which the scribes harmonized the former, more difficult passage in Luke. [85] This alteration, which soon gained wide support in the manuscript tradition, seems to be an attempt to remove any opportunity for adoptionist Christians to claim Luke 3:22 in support of their theological agenda. Numerous examples of this process occur in the early textual tradition of the Gospels and Acts: difficult verses that, while not necessarily heretical in themselves, left a door open for heretical eisagesis, were rewritten and sometimes significantly expanded (e.g. the variant endings of Mark) to protect orthodox readings of the New Testament.
In her recent book Guardians of Letters, Kim Haines-Eitzen has succinctly described the interpolative tendencies found in early Christian manuscripts:
The discursive debates in the second and third century intersected with textual transcription in the activity of copying and the (re)production of texts and creation of new readings. Intentional scribal changes did not occur in a vacuum, nor were they random in nature; rather, they were constrained by the discursive contexts of the scribes themselves. [86]
Thus significant theological arguments within Christian communities, at the time of copying, very often found their way into the texts, even if the divergences have often been read in the past as mere “variants” in the search for an Ur-text of the New Testament. {92|93}

The emergence of a Christian paraphrase tradition: Gregory Thaumaturgus on Ecclesiastes

I hope to have pointed so far to the fact that the habit of rewriting was a part of Christian literature, perhaps especially Greek Christian literature, from an early point and on a very literal level. Some of the most interesting evidence for Christian rewriting is the recasting of the Gospels and Acts from an orthodox point of view. Heretical groups, such as the followers of Marcion, were often accused of altering the New Testament by early Christian apologists, but no substantial evidence of these alterations has survived. [87] Rather, it is the orthodox changes that can be traced with some precision and testify to a thoroughgoing habit of adjusting the received text at its most difficult points. These altered orthodox manuscripts of the New Testament in turn became received texts in their own right, and even the most altered exemplars of this process, such as Codex Bezae, were still being copied in the fifth and sixth century—although the alterations contained in Codex Bezae appear to belong to a second or third century theological context.
The third century yields a different, perhaps transitional, example of Christian biblical rewriting, this time in the form of the standard biblical paraphrase common to Hellenistic Jewish literature. Gregory Thaumaturgus, the bishop of Neocaesarea in Asia Minor, wrote a lengthy paraphrase of Ecclesiastes that stands out as one of the few patristic commentaries on that elusive book. [88] Originally from a pagan family, Gregory attended Origen’s philosophy classes at Caesarea in Palestine during the 230s, to be converted to Christian theology under his tutelage. Taking up the bishopric of Neocaesarea, Gregory was credited with several writings and labeled a wonder-worker in late antiquity, picking up the title Thaumaturgus sometime in the sixth century. [89] His paraphrase of Ecclesiastes is significant as the earliest surviving Christian exemplar of this genre. [90] The paraphrase is in prose and follows the {93|94} text of the lxx closely. The text shows no sign that Gregory was making reference to the Hebrew, as might be expected from one of Origen’s students. [91] A look at the short preface reveals Gregory’s intentions to recover this work for Christian believers:
Τάδε λέγει Σαλομών, ὁ τοῦ Βασιλέως καὶ προφήτου παῖς ἁπάσῃ τῇ τοῦ θεοῦ ἐκκλησίᾳ, παρὰ πάντας ἀνθρώπους βασιλεὺς ἐντιμότατος, καὶ προφήτης σοφώτατος.
Solomon (the son of the king and prophet David), a king more honored and a prophet wiser than anyone else, speaks to the whole assembly of God. [92]
John Jarick observes in his commentary on the text that, instead of the shadowy Hebrew sage from the original Ecclesiastes, Gregory has named the traditional author of the text, Solomon, and given him his traditional epithet as well, “most wise.” The work is here redirected to a Christian audience through the use of “assembly/church” (ἐκκλησία) and its message is brought into the present tense (λέγει), replacing the lxx’s aorist (εἶπεν). [93] Throughout the Paraphrase there is a conscious effort on Gregory’s part to smooth out both linguistic and theological difficulties: [94]
The recurrent conclusion [in Ecclesiastes] that there is nothing better for a person to do in life than to eat and drink and find enjoyment for himself sounds suspiciously like a certain well known but un-Christian philosophy of life; Gregory tells his readers bluntly that the perfect good does not lie in eating and drinking, and that enjoyment is only granted by God to those people who act righteously. [95]
Gregory replaces the “all is vanity” mantra of the original text with a revisionist comparison between those who “see” spiritually and those who do not: “Most people have given themselves over to transitory things, not wanting to look—with the soul’s noble eye (τῷ γενναίῳ τῆς ψυχῆς ὄμματι)—at anything higher than the stars.” [96] Further, Gregory exchanges the original “the wise person dies, {94|95} just like the fool” for “the wise person never shares the same fate as the fool,” with an emphasis on the moral responsibilities of his Christian congregation. [97]
All of these (and many more) striking changes to the biblical text come in the narrative of the Paraphrase, which is (one must keep in mind) ostensibly only the text of Ecclesiastes itself. Towards the end of his Paraphrase Gregory gives some hints at how he perceived his role as paraphrast:
Δώσουσι δέ τινες τὰ σοφὰ ἐκεῖνα διδάγματα, παρ᾿ ἑνὸς ἀγαθοῦ λαβόντες ποιμένος καὶ διδασκάλου, ὥσπερ ἐξ ἑνὸς στόματος ἅπαντες αὐτοῖς συμφώνως δαψιλέστερον τὰ πιστευθέντα διηγούμενοι.
Some people will pass on those wise lessons which they have received from one good shepherd and teacher, just as if everybody with one voice described in unison and in greater detail what was entrusted to them. [98]
The use of the word “shepherd” perhaps points to Solomon, as the legendary “wisest of all,” or perhaps it signals Christ, who will have taught the faithful through the Paraphrase the Christian “wisdom” that is communicated therein. [99] Most likely the shepherd is simply Gregory, who portrays himself as communicating age-old wisdom to his young Christian flock—who were in turn previously unaware of the riches of this Old Testament manual. He seems here (like Erasmus) to view the Paraphrase as a mode of communicating the deep truths of a difficult text which are not apparent on the surface but which have been nonetheless handed down as pronounced in the chorus of the ages. By adopting the persona of the Koheleth—or “Solomon,” as he names him—Gregory can bring out those truths in a Christian guise and, most importantly, with the authority of the original author.

Cento and the reception of biblical paraphrase in the fifth century

The genre of formal paraphrase continued to be employed in late antiquity, in both prose and verse. Paraphrase in verse found its most talented exponent in the fifth-century Egyptian poet Nonnus of Panopolis, the writer of the lengthy epic poem the Dionysiaca, a hexameter account of the Greek god Dionysus’ mythical conquests in India. That the author of the Dionysiaca would undertake a Paraphrase on the Gospel of John is indeed surprising, and {95|96} many arguments have been marshaled to explain this apparent contradiction in ancient attributions. [100] Some scholars argue that the Paraphrase is simply misattributed; [101] others have hypothesized a conversion (or an apostasy) late in Nonnus’ career; still others claim that the Paraphrase is a distracting exercise, undertaken by (a Christian) Nonnus prior to or even while writing the Dionysiaca. [102]
It is not necessary to rehearse here the debates over authorship: detailed studies have been produced by several scholars, including Enrico Livrea, who is overseeing a new edition of the Paraphrase. [103] It is enough to observe with Livrea that all attempts to rationalize Nonnus’ literary biography suffer from the same lack of internal evidence:
I dati biographici che emergono da tante migliaia di versi sono così parchi e sfugenti da lasciar aperto il campo alle più contraddittorie construzioni, senza peraltro fornire alcuna sicurreza sull’appartenenza di Nonno al Christianesimo né, tanto meno, su una sua presunta conversione o apostasia. [104]
Despite the surprising dearth of self-revelations in the two texts, Nonnus can be seen, in his immediate literary-historical context, as a “wandering poet,” in Alan Cameron’s famous description, competing for literary patronage throughout the eastern empire—after the collapse of the traditional games system during the late 4th century—and composing Homeric verse according to the tastes of his disparate audiences. “It was in search of these patrons that our poets moved from city to city, exploiting in turn each center of learning and fashion.” [105] Cameron has placed Nonnus in a literary world populated {96|97} by scores of “scholar poets” now known to us only by name or anonymously through fragments, many of whom were clearly pagan and also had significant connections to the imperial court. [106]
Following Alan Cameron’s lead, one of his students, Lee Sherry, has postulated the existence of a “Nonnian school” [107] to explain the difficulties of style and attribution. [108] Sherry has argued that the Paraphrase is not by Nonnus at all but by one of his Christian students and is, most interestingly, actually a cento of the Dionysiaca itself. This possibility was first suggested (tentatively) by Joseph Golega in 1930, though revived by Sherry as a “key” to the problem. [109] However, a formal cento of the Dionysiaca (idiosyncratic in the extreme) would clearly have to replicate its verses, and the lack of coherency between the metrical patterns of the Dionysiaca and the Paraphrase is precisely why Sherry attributes the latter to a lesser poet: Golega’s suggestion of a cento seems meant to be evocative of the close relationship between the two poems, more than a genre analysis per se. Moreover, Golega himself argued that the Paraphrase (which he firmly attributed to Nonnus) shows a high level of metrical ability and also concluded that the metrical differences between the poems are due to the paraphrast’s Vorlage, the Gospel of John. [110] Golega has been confirmed by several subsequent scholars, and Sherry’s conclusions have consequently not won wide support. [111] {97|98}
Sherry’s argument about authorship, while widely criticized, has the benefit of suggesting a new way of looking at these texts. In particular, Sherry’s suggestion that a mixing of literary forms (cento and paraphrase) was even possible in this period reaffirms the need for a much wider discussion of the interpenetration of styles and genres in late Greek literature. Centones, typically written directly from the Iliad and Odyssey and not from recent Homeric continuators, are not extremely well attested but seem to have been a literary entertainment akin to the epigram and often appropriated by magical charm writers: lines of Homer pulled from their context in both a bookish and a religious manner. [112] Likewise, formal Christian paraphrase in late antiquity, in its Homeric forms at least, probably developed directly out of the educational system. [113] As Dennis MacDonald has observed in his study on the Gospel of Mark’s imitation of the Homeric epics:
Quintilian supposed his readers would have taken this activity for granted: “I think we shall all agree that this [paraphrasing] is especially valuable with regard to poetry; indeed, it is said that the paraphrase of poetry [into prose] was the sole form of exercise employed by [the rhetor] Sulpicius.” The littérateur Philodemus asked, “Who would claim that the writing of prose is not reliant on the Homeric poems?” [114]
Students learned to write through copying and recopying Homer and other canonical authors, a process which instilled in them both the style of the original and a capacity for rewriting. It is quite right, then, that these genres, cento and paraphrase, could potentially mingle together in the fifth century, despite appearing distinct in earlier literary history.
Both Homeric cento and biblical/Homeric paraphrase presuppose a close attachment to a canonical text, and recomposition is clearly their shared modus operandi. In addition, it appears to be in the fifth century that monastic schools began using the Psalms and certain liturgical texts for basic language instruction, requiring students to memorize large sections of the Psalter and, quite probably, portions of the Gospels as well. [115] That the New Testament texts {98|99} would be subjected to the same project as the Old—and as Homer had been in Greek schools for some time—is not as idiosyncratic to “Nonnus” as Sherry would have us believe in this literary context (which he invoked to begin with). [116] Consequently, any discussion of Nonnian authorship and the development of fifth-century literature should take account of these broader literary and pedagogical movements; the unexplained rise in cento and biblical paraphrase exempla in the fifth century is first of all the result of the extant texts (see below), but it was no doubt also part of the germination of a Christian self-consciousness at this time, a self-consciousness which has been shown to owe a tremendous amount to shifting patterns of education. [117]
These two literary forms, cento and paraphrase, came together in the famous literary endeavors of the fifth-century empress Eudocia/Athenaïs, empress of Theodosius II (421–460). In her centones of the Iliad and Odyssey, Eudocia retells the Christian story of Fall and Redemption and thus follows the narrative line of biblical history, proceeding from the creation of the world to the ascension of Christ. [118] However, she does so in the “patchwork” form of the cento, rewriting the biblical text through Homeric verse; in this striking experiment, she produces what Mark Usher has called “Outsider Art,” a reusing of “discarded material” to create new, “other” literary art, in the same manner as, for example, “the magnificently naïve painting of American folk artists” like Howard Finster. [119] Usher, having recently reedited Eudocia’s Homerocentones (1999) in addition to writing two studies on them (1997; 1998), has brought to the fore some of the complex intertextual questions regarding these works, which, as Gregory Nagy says in his forward to Usher’s study, “presuppose a {99|100} veritable internalization of both Homer and the Bible.” [120] Nevertheless, Usher has not placed these centones in the literary-historical context of biblical paraphrase, an important and necessary juxtaposition, I believe, if we are to understand the full impetus and the cognitive implications of both Eudocia’s and Nonnus’ writings. [121]
In addition to the Homerocentones Eudocia also wrote hexameter paraphrases of Zechariah and Daniel, the Octateuch (in eight books), and the martyrdom of Saint Cyprian of Antioch (in three books). [122] Only the Homerocentones and the paraphrase of Cyprian’s martyrdom survive (the latter only partially), but her corpus attests, again, to the combination of biblical paraphrase with other genres in late ancient writing. Specifically, Eudocia as an author reveals the striking union of (Old Testament) biblical paraphrase, the cento, and the rewriting (in verse) of early martyr acts. There is no reason not to see all three of these literary projects as coming directly out of a Greek Christian education system in the fifth century. [123] As I have already demonstrated in the preceding chapter, the author of the Life and Miracles of Thekla also exhibits the conjunction of these received literary forms in the fifth-century: the influence of biblical paraphrase, the Homeric epics, and early Christian martyr acts. In addition, Photius records a (now lost) verse paraphrase of the Acts of Paul and Thekla by Basil of Seleukeia, the fifth-century bishop once thought to have written the prose Life and Miracles. [124]
Substantial late antique paraphrasing activity in prose and verse did not go unnoticed by other contemporary writers. The historians Socrates and {100|101} Sozomen both comment on the writing of biblical paraphrase, though with contrasting conclusions. These fascinating vignettes on Christian literary history are worthy of being quoted here in full:
The imperial law [of Julian] which forbade Christians to study Greek literature, rendered the two Apolinarii, of whom we have above spoken, much more distinguished than before. For both being skilled in polite learning (ἄμφω ἤστην ἐπιστήμονες λόγων), the father as a grammarian, and the son as a rhetorician, they made themselves serviceable to the Christians at this crisis. For the former, as a grammarian, composed a grammar consistent with the Christian faith (τὴν τέκνην γραμματικὴν Χριστιανικῷ τύπῳ συνέταττε): he also translated the Books of Moses into heroic verse (τά τε Μωυσέως βιβλία διὰ τοῦ ἡρωικοῦ λεγομένου μέτρου μετέβαλεν); and paraphrased all the historical books of the Old Testament (καὶ ὅσα κατὰ τὴν παλαιὰν διαθήκην ἐν ἱστορίας τύπῳ συγγέγραπται), putting them partly into dactylic measure, and partly reducing them to the form of dramatic tragedy. He purposefully employed all kinds of verse, that no form of expression peculiar to the Greek language might be unknown amongst Christians. The younger Apolinarius, who was well trained in eloquence (εὖ πρὸς τὸ λέγειν παρεσκευασμένος), expounded the Gospels and apostolic doctrines in the way of dialogue (ἐν τύπῳ διαλόγων ἐξέθετο), as Plato among the Greeks had done. Thus showing themselves useful to the Christian cause they overcame the subtlety (τὸ σόφισμα)of the emperor through their own labors. But Divine Providence was more potent than either their labors, or the craft of the emperor (κρείσσων ἐγένετο καὶ τῆς τούτων σπουδῆς καὶ τῆς τοῦ βασιλέως ὁρμῆς): for not long afterwards, in the manner we shall hereafter explain, the law became wholly inoperative; and the works of these men are now of no greater importance than if they had never been written (τῶν δὲ οἱ πόνοι ἐν ἴσῳ τοῦ μὴ γραφῆναι λογίζονται). [125]
[Julian] forbade the children of Christians from being instructed in {101|102} the writings of the Greek poets and authors and from visiting their teachers. He entertained great resentment against Apolinarius the Syrian, a man of manifold knowledge and philosophical attainments, against Basil and Gregory, natives of Cappadocia, the most celebrated orators of the time, and against other learned and eloquent men, of whom some were attached to the Nicene doctrines, and others to the heresy of Arius. His sole motive for excluding the children of Christian parents was because he considered such studies conducive to the acquisition of argumentative power. Apolinarius, therefore, employed his great learning and ingenuity in the production of a heroic epic (ἐν ἔπεσιν ἡρῴοις) on the antiquities of the Hebrews to the reign of Saul (τὴν Ἑβραϊκὴν ἀρχαιολόγιαν συνεγράψατο μέχρι τῆς Σαοὺλ βασιλείας), as a substitute for the poem of Homer (ἀντὶ μέν τῆς Ὡμήρου ποιήσεως). He divided this work into twenty-four parts, to each of which he appended the name of one of the letters of the Greek alphabet, according to their number and order. He also wrote comedies in imitation of Menander, tragedies resembling those of Euripides, and odes on the model of Pindar. In short, taking themes of the “circle of knowledge” from the Scriptures (ἐκ τῶν θείων γραφῶν τὰς ὑποθέσις λαβὼν τῶν ἐγκυκλίων καλουμένων μαθημάτων), he produced within a very brief space of time, a set of works which in manner, expression, character, and arrangement are well approved as similar to the Greek literatures and which were equal in number and in force (ἰσαρίθμους καὶ ἰσοδυνάμους πραγματείας ἤθει τε καὶ φράσει καὶ χαρακτῆρι καὶ οἰκονομίᾳ ὁμοίας τοῖς παρ᾿ Ἕλλησιν ἐν τούτοις εὐδοκιμήσασιν). Were it not for the extreme partiality with which the productions of antiquity are regarded, I doubt not but that the writings of Apolinarius would be held in as much estimation as those of the ancients. The comprehensiveness of his intellect is more especially to be admired; for he excelled in every branch of literature, whereas ancient writers were proficient in only one. [126]
The Apolinarius the elder whom both writers cite was the father of the Apolinarius the younger whose Christological teaching was condemned at the first Council of Constantinople in AD 381. A Paraphrase of the Psalms attributed to Apolinarius the elder has come down to us, though the attribution must {102|103} be incorrect due to its dedication to the emperor Marcian (AD 450–457). This text has been analyzed in detail by Golega, who firmly established its date on stylistic grounds to the fifth century. [127]
Clearly Socrates and Sozomen know of even more paraphrasing activity going on in the fourth century, for which we have no texts or fragments, but it is of course reasonable that Gregory of Thaumaturgus’ paraphrase of Ecclesiastes in the third century would have had some immediate successors. The length alone of the vignettes quoted above attests to an interest on the part of Socrates and Sozomen in the literary history of the period, but their assessments of the value of these works are strikingly different.
Theresa Urbainczyk has concluded that the lack of a mention of Eudocia’s Paraphrases in Socrates’ History is a slight against the empress, since he would have implicitly condemned her writings along with those of the Apolinarii. [128] Urbainczyk’s argument seems to assume too much, and the omission has been more successfully and simply explained on other grounds, namely that Eudocia had not published her paraphrases by 439, when Socrates finished his History. [129] One remark of Urbainczyk, however, deserves closer scrutiny: “It seems to me that the subject of the work done by the Apolinarii was probably only remembered in the early fifth century because the empress and her friends were repeating the exercise.” [130] “Repeating the exercise” is precisely the point, I think, and it highlights the disingenuousness of their reporting: clearly there was a much stronger tradition of biblical paraphrase in the fifth century than either Socrates or Sozomen fully acknowledges. [131]
The Apolinarii, far from inventing the genre, were rather perpetuating a long tradition of paraphrase that could claim a famous proponent, Gregory Thaumaturgus, just a century before. Moreover, Socrates and Sozomen set their notices on the Apolinarian paraphrases in the context of fourth-century disputes over education, precisely the region of knowledge from which Christian paraphrases—especially those in heroic meter—seem to have emerged. The Christian tradition of paraphrase to which these vignettes point confirms the argument of the present chapter: that paraphrase and rewriting, even on a very literal level, was more common, and more integral, to Jewish {103|104} and Christian textuality than has previously been recognized, or than, most importantly, is represented by surviving exempla.
Even the divergence between Socrates’ and Sozomen’s histories attests to this habit: Sozomen, writing ten years later (with access to Socrates’ History), chose to include new and different details of the literary reactions to Julian, in addition to providing a startlingly opposing judgment on the value of those reactions. [132] In this strikingly intertextual way, Sozomen shows himself to be a historiographical paraphrast, and his engagement in this exercise, at the very moment of describing other paraphrasts, highlights further the importance (and ubiquity) of paraphrase in Greek Christian literature.

Fifth-Century Metaphrastai: Revisiting Rapp on Antiquarianism

In addition to thriving Homeric imitations, the fourth through sixth centuries was a period when apocryphal Acta from the second and third centuries were being rewritten, extended, and embroidered with facility and vigor. [133] In the late antique East this meant that received texts about famous apostolic personages—like Thekla, the apostle Philip, and the apostle John—were the loci of several individual rewritings and extensions. These latter texts testify, of course, to textual competition and the appropriation of the cults for specific sites—Seleukeia for Thekla, Ephesus and Patmos for John—but, more fundamentally, these rewritings are indicative of an indigenous cultural habit of Christian textuality. To be sure, in late antiquity the apocryphal Acta were not Scripture, and textual critics like Ehrman suggest that rewritings of the New Testament were not still occurring on a large scale in the fourth and fifth centuries (at some point between the third and fourth centuries the {104|105} manuscript traditions solidified and became more or less stable—attitudes had thus changed with regard to the biblical texts). [134] Nevertheless, the apocryphal Acta were often rewritten at this time with the same goal in mind as the earlier biblical revisions, that is, to purge the texts of opportunities for heretical readings, or of heretical material itself. Following this period of reception and rewriting, which helped spawn new forms of literature, writers like Leontius of Neapolis in the seventh century began to collect and to rewrite more recent (fourth- to sixth-century) saints’ Lives in a consciously antiquarian fashion; within a few more centuries, Leontius’ antiquarian tendencies found their preeminent expression in the work of Symeon Metaphrastes. [135]
The perceived historicity of the saints’ “lives and deeds” (βίοι καὶ πράξεις) was of central importance to the rewriters, but they were also not unaware of the fictional, novelistic, and simply imaginative elements of the legends they received and redacted. [136] This process of collecting, culling, and writing was the modus operandi of late antique “hagiographers” and is often described by them in the self-defining sections of their works. Their antiquarian ethos, which has not gone unnoticed by scholars but is still under-emphasized, depends first, I argue, upon the cognitive classification of the traditions of early saints as historical, received, and authoritative. The early saints were more often than not also apostles, and the names associated with the received texts about them—Paul, Peter, Thomas, John, Thekla—added gravitas to the historiographical vocations of the late antique writers who undertook the antiquarian task of discovering, sorting, and publicizing the previously hidden data, the “apocryphal” deeds of the apostles. This same ethos was extended to the lives of saints contemporary with the antiquarian project, and increasingly, to their current, posthumous activities as well.
The contemporary cultural imperative for this kind of literary activity was as crucial as the historical: the latter depended upon the objective existence of a text, a textual artifact, often consciously given the special status of textus receptus; the former depended upon the force of religious habit in late antiquity, the immediacy of sacred, otherworldly holiness in select men and women, and also upon a conscientious respect for the orthodox innovations of the day, notably the ubiquitous cult of the saints and the relics and local stories {105|106} it generated. This project of exhuming the textual past for cults current in late antiquity was fueled by a recognition of the need to preserve the past (and historical present) for the future.
Within this project, however, authors often sought, or felt compelled, to reclassify, reorient, and purify the textual past for the sake of their audiences and readers-to-come. A cathartic imperative such as this betrays an awareness of the dangerous effects to the soul of an improper interpretation of the past: in particular, the elements of the legends of the apostles that signaled heretical conclusions for Christian morality and practice were expunged. The so-called “Encratites,” heretical sectarians who were said to have insisted on (among other things) the necessity of sexual renunciation for salvation, were often accused by late antique heresiologists—antiquarians in their own right—of appropriating to destructive ends what were essentially historical, spiritually nourishing narratives of the apostle-saints. [137] Thus, the “Encratic” elements of early apocryphal acts were removed by late antique rewriters for the welfare of their readers. Interestingly, however, these elements were not seen to have polluted the historical narratives contained in the acts. Subsequent to the purgings, readers were expected to consider the authorized revised versions as true history, and also as beneficial for devotion, prayer, and the Christian spiritual life generally.
It is standard scholarly fare that the earliest Christians, or at least representative writers, considered apocryphal stories concerning Jesus, his family, and the apostles just as factual and authoritative as the canonical New Testament. [138] What scholars of early Christianity have perhaps missed, however, is the inspirational role that apocryphal Acta had on the development of Christian literature. While later generations of writers, particularly in the fourth and fifth centuries, were interested in expunging Encratic elements in these stories—in opposition to the earliest writers who considered such elements authentic?—they were nevertheless enthusiastic about the Acta as received literary tradition. Thus, the apocryphal Acta were not simply bodiless legends about the apostles to be manipulated at will, but they had an {106|107} inspirational role as textual encapsulations of these legends. Consequently, a conscious mimesis of the style, structure, and language of apocryphal Acta is very present in Christian novelistic literature from the fourth and fifth centuries. This fact remains underappreciated by scholars of both early Christianity and late antiquity because most saints’ Lives in this period have no direct early predecessor but instead describe contemporary holy figures. By contrast, the argument of the present study is that a mimetic motivation could potentially stand behind the authorship of some saints’ Lives that have been seen as more or less sui generis. The Life of Thekla is very strong evidence that the tradition of Christian biography (or Christian Romance) represented by the second-century apocryphal Acta was alive and well in the fifth century, a hundred years after Athanasius wrote the seminal Life of Antony.
In addition to these substantial, and apparently frequent, rewritings of second and third-century Acta, new Acta in the style of the earlier ones continued to be written in late antiquity. While the lack of precise dates for the authorship of many Acta prevents scholars from establishing exactly how late this trend continued, it is nevertheless clear that they were still being written and read in tandem with the first late antique saints’ Lives (mid fourth century), and that they were around for a long time after the latter had become widely disseminated. For example, the Acts of Philip, recently re-edited by François Bovon and others, was written no earlier than the fourth century and most likely represents an Encratic community of Asia Minor attempting to provide historical documentation for their position in the face of increasing hostility from the ecclesiastical establishment. [139] This hostility came perhaps even from Cappadocian bishops like Basil of Caesarea and Amphilocius of Iconium who participated in the Council of Gangra in Paphlagonia (c. AD 341 or 355), a Council which condemned the extreme asceticism advocated by Eustathius of Sebaste (c. 300–after 377). [140] In subsequent centuries these apocryphal Acta were still widely read and incorporated into homilies, later saints’ Lives, and chronographies that dealt with the early church. [141] For instance, it appears that apocryphal acts of James, now lost, were incorporated into a {107|108} sermon by Nicetas David of Paphlagonia, who shows a very detailed knowledge of that tradition. [142] The apocryphal Acta thus survived, and surely cross-fertilized, the flowering of the traditional saint’s Life in late antiquity and, moreover, continued to be considered legitimate historical literature concerning the apostles. While it is not clear precisely how late these Acta continued to be written in the style of the second-century ones, they are found as late as the fifth and sixth centuries and, not insignificantly, they were mined by homilists and historians of Byzantium for their own creative writing on the early saints. The Acts of John by Pseudo-Prochorus (fifth or sixth century) stands as perhaps the latest surviving Greek exemplum of this tradition, [143] but the sixth and seventh century translations of the apocryphal Acta into Syriac and Armenian attest to their continued popularity in non-Greek early Byzantium.
In her article “Byzantine Hagiographers as Antiquarians, Seventh to Tenth Centuries,” Claudia Rapp has analyzed the tendency of middle Byzantine authors to rewrite (μεταφράζειν) saints’ Lives, to collect their legends into practical compilations—μενολόγια, συναξάρια, and the like—and to treat the earlier legends as textual artifacts of a distant past. [144] Underlying this tendency was, she says, “the melancholy insight that the age of the saints has irrevocably come to a close” (31). Along the literary-historical continuum of Byzantine hagiography Rapp locates this paradigm shift in the seventh century, during which works like the Life of John the Almsgiver by Leontius of Neapolis and the Miracles of Saint Anastasius the Persian projected onto the saints an innovative antiquarian consciousness: both John and Anastasius are said to have themselves enjoyed reading the Lives of saints (τὸ ἐντυγχάνειν τοῖς βίοις τῶν ἁγίων πατέρων; 35 and n12). Also in the seventh century came the “first flourishing” of μεταφράσεις: Leontius’ “stylistic downgrading” of John’s original Life by Sophronius of Jerusalem and John Moschus anticipated the later, massive project of rewriting by Symeon the Logothete, nicknamed “Metaphrastes” (36–37). [145] {108|109}
As Rapp shows, it was not until the ninth century that hagiographic compilations first appeared: Theodore the Stoudite produced a proto-μενολόγιον, the πανηγυρικὴν βιβλίον, which in the next century was the inspiration for a much larger compilation, the famous μενολόγιον of Constantinople, also by Symeon. This is the literary history that for Rapp is most indicative of an antiquarian tendency: that is, an increasing awareness of the use and spiritual profit of saints’ Lives in the seventh century, then an early interest in compiling in the ninth, culminating in a very substantial intensification of rewriting and compilation in the tenth century, commissioned by Constantine Porphyrogennetos and spearheaded by Symeon Metaphrastes. [146] She has shown how many writers before Symeon were engaged in reconfiguring and preserving older Lives, and she has argued convincingly that Symeon’s metaphrastic collections should thus be seen as the culmination of a tradition rather than as a historically isolated project of “inventorizing” in the tenth century.
Rapp has thus delineated this significant trend from the seventh to tenth centuries, but what I hope to have shown is that μετάφρασις, a constituent element of Greek Christian textuality, was embedded in Christians’ responses to their texts, including the New Testament, from much earlier. Did the Byzantines recognize earlier μετάφρασις? They certainly did with regard to the apocryphal Acta—Symeon is aware of multiple versions of apocryphal apostolic narratives, some of which are closer than others to the revisionary style he advocates and employed in his compilations. [147] On this basis, I would conclude that Rapp’s argument is further strengthened, if significantly revised, by pushing the continuum back into earlier Christian literature. Rewriting and paraphrase are central to Greek literary history (classical, Jewish, and Christian) and provide a cognitive thread which can be traced through the whole first millennium of the Christian era.

Conclusion: Metaphrasis in Late Antiquity and Beyond

The brief survey just presented offers an opportunity to consider synchronically a literary activity that, by the fifth century AD, had been ongoing for a very long time in Jewish and Christian literary traditions. It is probable, though most likely impossible to prove, that the evidence of μετάφρασις in the {109|110} received texts of both religions—e.g. the canonical books of the Chronicles—provided the initial impetus for the receivers (Jews or Christians) to engage in that activity themselves. Scholars would, of course, be arrogant to assume that early Christians were unaware of something of the metaphrastic relationships among the synoptic gospels: the fascinating, if elusive, example of Tatian’s Diatessaron is already suggestive of such an awareness. Through the reception of texts and rewritings of those texts, as well as through the reception of the project of rewriting (as a kind of institution), μετάφρασις became a literary vocation and proceeded to cross-fertilize new and influential texts, such as the disparate group of writings broadly labeled as “hagiography.”
Concerning a topic as big as rewriting there will always be new evidence to cite and new syntheses to be made. However, I have tried in this chapter to point to commonalities among the examples cited above in an attempt to center the scholarly discussion of rewriting on the processes involved. The investment of contemporary ideology or polemic is visible in all of the rewritings, even if not always as pronounced as in the Qumran community’s eschatological anticipations. The presence of the paraphrast in the “hypertext”—sub evangelistae persona—is also a common feature, though often less self-conscious than in Erasmus’ dedication of his Paraphrase on Luke to Henry VIII. Further, Vorlagen could be changed almost beyond recognition, as in the Genesis Apocryphon or Eudocia’s cento of Christian redemptive history; however, subtle changes also point to a similar process of reception and modification, as evidenced by Codex Bezae.
Jack Goody’s model of endless elaboration by oral myth tellers—especially among the “scholars” that elaboration produced—is suggestive in its “textual” outline of a cognitive imperative of rewriting inherent to human textuality or story-telling in general. Ancient paraphrase and rewriting, it has been argued, could provide an extension of his oral evidence. Donald McKenzie has suggested something like this in his controversial attempt to redefine bibliography as a twenty-first-century discipline. [148] In addition to pointing out the importance of ascertaining physical changes to texts through time—as an entrée to pursuing their full meaning diachronically—he has emphasized the synchronic inevitability of a text becoming a “site” for later rewritings.
In other words, not only does the physical history of a text “make up” its meaning, but the dissemination of a text (oral, visual, or written) “lets out” its meaning to be reconstructed by as many as come into contact with it. McKenzie has formulated this argument not to relativize textual meaning as {110|111} much as to historicize it, and to provide a firmer basis for the work of textual criticism in the age of textual deconstruction:
History simply confirms, as a bibliographical fact, that quite new versions of a work which is not altogether dead, will be created, whether they are generated by its author, by its successive editors, by generations of readers, or by new writers. [149]
My argument in this chapter takes inspiration from McKenzie in that I have tried to forge a link between rewriting and paraphrase in practice (as examined in Chapter One above) and the literary history of biblical and apocryphal paraphrase. Such a link is not primarily about authorial intention but about how texts are inevitably changed by their receivers. This may seem at first glance to be a banal point in an age when every phenomenon in human experience comes under the academic designation “text.” However, with regard to the history of Greek literature in late antiquity, the important connections between paraphrase as a literary form, literate education, and the way Jews and Christians read their “Bible” (or the history of their institutions in general) suggest that a link between minute changes to received Scriptures and the wholesale rewriting of formative texts, canonical or apocryphal, needs to be made and explored.
In her analysis of the Paraphrase of Ecclesiastes by Gregory Thaumaturgus, Françoise Vinel has concluded that biblical paraphrase as a genre performs certain tacit operations on its audience. First, within a Jewish or Christian context it conflates interpretation with Scripture to the degree that what originally was text now becomes a kind of prohibitive intertext. “La metaphrasis comme la paraphrase interdit par définition ce mouvement entre le texte original et sa lecture.” [150] More forbidding than in a sermon, an interpretation that is made in a paraphrase sticks with you because it has become the very Scripture you are reading or hearing, and it thereby affects your apprehension of other parts of the Scriptural or imaginative whole.
While it has not been the concern of this chapter to deal with the complex allusions of the paraphrases cited above, the allusive quality of paraphrase must be kept in mind—and perhaps this is a point at which paraphrase, as a genre, diverges from literal rewritings of scriptural texts. [151] However, Daniel Boyarin has argued for the intertextuality of midrash to be seen as inspired {111|112} by, or as the direct result of, the intertextuality of Scripture itself. [152] And one might assume that individual changes to Scripture would also play off of these inherent allusions. As demonstrated in Chapter One, the author of the Life of Thekla appropriates for the contemporary cult the history of the apostles and their personae (particularly Thekla, Paul, Stephen, and Luke). This appropriation, however, requires the author to take seriously any dominant, received legend concerning the earliest period of Christian history.
The seriousness with which the author of the Life of Thekla has taken the received testimony of her apostolic status is evident in every one of the changes that he makes to his source text, the ATh. His cognitive appreciation of that text’s authority is dependent on its received, quasi-canonical status in his contemporary situation. His appreciation is strengthened and intensified by his spiritual relationship to Thekla herself and by her local activities at the shrine in Seleukeia. Both the past and the present thus serve as motivations for the Life and Miracles as a whole—and, as will be shown, so does the future. With this in mind it is time to look closely at the second half of his text in order to see how he transforms Thekla the apostolic saint into Thekla the late antique miracle worker. {112|}


[ back ] 1. Quoted by Vessey 2002a:7; parentheses are his. For Udall, see Craig 2002:316–322 and n22.
[ back ] 2. See Bedouelle 2002 on their translation and reception and Rummel 2002 on Noël Béda’s condemnation of them.
[ back ] 3. For this see Craig 2002 and references, esp. n12 for accounts of the translation of the Paraphrases into English. Interestingly, the Paraphrases were not on the list of proscribed books issued during Mary’s reign: Craig 2002:326–327.
[ back ] 4. Noël Béda was especially resistant to Erasmus’ assertion that some passages of Scripture are not understandable by themselves (Rummel 2002:267); he also condemned the implication that Scripture should be made available in the vernacular (Vessey 2002a:18).
[ back ] 5. For Erasmus’ Paraphrase on Luke, see Phillips 2002.
[ back ] 6. Vessey 2002a:14–15.
[ back ] 7. “Luke-voice” is from Phillips 2002:e.g. 131; the prologue to Luke is expanded into an essay twenty-six times its original length (ibid.).
[ back ] 8. Vessey briefly discusses Erasmus’ knowledge of earlier paraphrasts and cites Roberts 1985, which explores only the Latin side of biblical paraphrase in late antiquity. Bernard Rousell in his contribution mentions a few ancient paraphrasts, such as Gregory Thaumaturgus and Nonnus of Panopolis, but only in passing; his interest lies in the Reformation paraphrasts subsequent to Erasmus (Roussel 2002:59).
[ back ] 9. For the Septuagint as a paraphrase, see below. For the revival of biblical languages in the Reformation and some of their political and social implications, see Goldhill 2002:14–57. Erasmus probably knew the first printed edition of the Targums by Felix Pratensis, who printed them in Venice alongside his four-volume edition of the Hebrew Bible (1517–1518). Pratensis was a Jew who had converted to Christianity and was in the employment of Daniel Bomberg, a wealthy Antwerp native who spent his fortune in Venice printing Hebrew (and Aramaic) books—about two hundred in all. On the early printing history of the Aramaic Targums, see Díez Merino 1994, esp. 80–86.
[ back ] 10. He certainly knew Juvencus, and, “when the Paraphrases themselves came under attack, he repeatedly allied himself with that fourth-century Christian poet” (Vessey 2002b:32). It has been argued, however, that he was unacquainted with Proba’s cento (Vessey 2002b:52n17). There is a good chance Erasmus knew something of Nonnus’ Paraphrase of John, since editions of the latter had been printed at least twice by this time (Roussel 2002:79n3)—one of these was the Aldine edition of 1501–1504; the other was the 1527 edition by Philipp Melanchthon and Johann Setzer. Vessey makes a useful comparison between the rhetoric of Erasmus’ paraphrase program and Jerome’s reflections on paraphrase vs. translation (Vessey 2002b:52n13).
[ back ] 11. For theories of “hypertextuality” in ancient literature, see MacDonald 2000:1–14.
[ back ] 12. For the Renaissance and Reformation, see the standard study of Eisenstein 1980 and, in opposition, Johns 1998; and, for the still disputed significance of our current transition from print to electronic media, see O’Donnell 1998. Both Eisenstein and O’Donnell rely on Goody’s formulations. See also the seminal studies of Marshall McLuhan 1962 and Walter Ong 2002, who rely less on Goody.
[ back ] 13. See his trilogy of major studies, Goody 1977, 1986, 1987, and 2000, which is a convenient summary restatement of his views.
[ back ] 14. Goody 2000:53.
[ back ] 15. Goody 2000:118.
[ back ] 16. Goody cites especially Derrida’s Of Grammatology (1974) but also references discussion of Derrida’s work in Culler 1979. It should be noted that Goody’s student David R. Olson, has discussed in depth the effect of literacy on cognition, especially in the context of linguistic self-location: “Writing is largely responsible for bringing language into consciousness” (1994:xviii).
[ back ] 17. Goody 2000:114–115. Goody’s commitment to the oral has been followed by many outside his discipline. The classicist Gregory Nagy, for example, has repeatedly emphasized the oral vitality of the Homeric epics. See Nagy 1996; and also his and Stephen Mitchell’s new edition of Albert Lord’s Singer of Tales (Lord 2000).
[ back ] 18. Goody 2000:iii.
[ back ] 19. For a thoughtful critique of Goody on this point, see Bloch 1998:131–151. One could argue that Goody has misunderstood Derrida in that, for the latter, “text” is a metaphor more than a mode of communication. Nevertheless, Goody’s seminal conclusions about the cognitive relationship between canon and elaboration still stand.
[ back ] 20. Goody 2000:151: “writing is a prerequisite, a prerequisite for the development of all the technologies with which our intellect engages.”
[ back ] 21. Goody 2000:21.
[ back ] 22. Alter 2000:60. Alter is here admittedly building on Alan Bloom’s The Western Canon (1994), though is critical of its central Oedipal metaphor.
[ back ] 23. Alter 2000:61.
[ back ] 24. Morgan 1998:224. She is right to formulate a pair with these two cognitive activities, but whether there was ever, even in literate societies, a “particular version” of the canonical texts is still debated.
[ back ] 25. On paraphrase in ancient schools, see Morgan 1998:198–226.
[ back ] 26. Zeitlin 2001.
[ back ] 27. See Hollis 1994 and forthcoming.
[ back ] 28. Alan Cameron 1965:482; Suda s.v. “Μαριανός.”
[ back ] 29. According to a TLG search (performed by the author on 28 March, 2005), μεταβολή appears to be a standard Byzantine term for paraphrase from about the tenth century. It only rarely has this meaning in classical and late antique literature (LSJ s.v.). Note, however, that the Suda entry for Marianus just cited calls his works μεταφράσεις.
[ back ] 30. Josephus Antiquities 1.5, 10.218. Interestingly, the verb μεθερμενεύω is also used by Josephus when describing the Greek Septuagint “translation” of the Hebrew Bible (12.20, 48); while this usage may seem like contrary evidence, there is good reason to render it also as “to interpret” or “to paraphrase”; for this see Feldman 1998a:44–45.
[ back ] 31. Writers who use μετάφρασις to mean “paraphrase” include Dionysius of Halicarnassus, Plutarch, [ back ] Origen, and Eusebius; see LSJ and Lampe s.v. “μετάφρασις” and “μεταφράζω.”
[ back ] 32. On the refashioning of myth for contemporary political ideologies, see Veyne 1988.
[ back ] 33. Haines-Eitzen 2000:105–106, citing Foucault 1983:44.
[ back ] 34. McKenzie 1999 [1986].
[ back ] 35. McKenzie 1999:19.
[ back ] 36. McKenzie 1999:23.
[ back ] 37. Cf. Høgel 2002.
[ back ] 38. See Alter and Kermode 1987:92–101 and Alter 2004:xv, 869–877, and passim.
[ back ] 39. See Kugel 1998:2, 6, and the refs at 2n2, esp. Japhet 1997 [1989].
[ back ] 40. The inclusion of rewrites within the Old Testament canon itself must be of fundamental importance for early Christian conceptions of the validity of paraphrase with regard to their own Scriptures.
[ back ] 41. That an importance was attached to the Hebrew text, by the end of the first century AD at the very latest, can be shown from the fact that Aquila’s literalist rendering of the Hebrew into Greek was well received by the Jewish community, over and against the paraphrasing Septuagint preferred by the Christians; see Swete 1900:31–42.
[ back ] 42. The phrase “Rewritten Bible” was apparently coined by Vermes 1975, but others have taken up this concept with vigor. See esp. Kugel 1998, whose conception of the history of Jewish biblical interpretation hinges upon the concept: e.g. “The Rewritten Bible is really the interpreted Bible,” and “The Rewritten Bible (whether one is talking about an extended retelling of whole biblical books, or the ‘retelling’ of a single verse) should be recognized for what it is: the most popular transmitter of biblical interpretation among ancient writers” (Kugel 1998:23).
[ back ] 43. Vermes 1975:39.
[ back ] 44. Only a small proportion of these variations are scribal errors. See Vermes 1998:15 on the “extreme fluidity” of the Qumran Bible(s). On the distinctiveness of the Septuagint’s Vorlage and the Qumran texts, see Tov 1992: “many, if not most of the biblical texts of the third and second centuries BCE were unique . . .” (42–43).
[ back ] 45. See Vermes 1998:448–459.
[ back ] 46. Bernstein 1994:2; Vermes 1975:62–63.
[ back ] 47. As is well known, the Letter of Aristeas records the translation of the LXX by seventy-two Jewish scholars from Jerusalem invited to Alexandria by the king Ptolemy. In a rather frustrating manner the text does not get around to discussing the actual work of the translators until the very end, and, even then, the details of the process are not revealed. However, what the Letter of Aristeas does make clear, through its rhetoric of superiority and self-justification, is that there were competing translations, contemporary with the penning of the Letter (perhaps 1st cent. BC). For the text of the Letter, see H. St. J. Thackeray’s still standard edition in Swete 1900:519–574; see also the translation with introduction and notes by R. J. J. Shutt in Charlesworth 1985:7–34.
[ back ] 48. Response to Barthélemy 1963 has not been completely positive: Grabbe 1992 argues that Barthélemy overemphasizes the influence of this earlier revision (the so-called “kaige recension”) on Aquila’s translation.
[ back ] 49. Brock 1992:303.
[ back ] 50. Lindars 1992:4–5.
[ back ] 51. For competing methods of biblical translation in antiquity, see Brock 1992; for a helpful anthology of essays on modern translation theories, see Schulte and Biguenet 1992.
[ back ] 52. Eusebius Ecclesiastical History 6.16. Some Christians, such as Lucian of Antioch in the third century, made their own Greek translations straight from the Hebrew, as Jerome did into Latin over a century later; for a detailed survey of all the biblical versions, see ABD s.v. “Versions.” However, for most early Christian writers, the approval of Josephus and Philo, in addition to the New Testament, was enough to guarantee the LXX’s authority.
[ back ] 53. Ralfs 1979:lxii.
[ back ] 54. Jarick 1990:6, citing Daniélou 1955:133.
[ back ] 55. Feldman 1998b:539.
[ back ] 56. See Brock 1992:303–310. Josephus Antiquities 12.108–109; Philo Life of Moses 2.25–44, esp. 40: “. . . if Chaldeans [i.e. those who read Hebrew/Aramaic] have learned Greek, or Greeks Chaldean [i.e. Hebrew/Aramaic], and read both versions, the Chaldean and the translation, they regard them with awe and reverence as sisters, or rather one and the same, both in matter and words, and speak of the authors not as translators but as prophets and priests of the mysteries, whose sincerity and singleness of thought has enabled them to go hand in hand with the purest of spirits, the spirit of Moses” (trans. F. H. Colson, LCL Philo vol. 6).
[ back ] 57. Thus, once the rabbis took control of the targumim which they inherited, they “were concerned that targum should be clearly distinguished from Scripture: the same person could not publicly read the Hebrew and recite the targum” (Alexander 1992:330). This is an interesting example of incorporating a paraphrase into a different cognitive system, both to appreciate its teaching as well as to make it submit to a higher textual authority.
[ back ] 58. Some targumim are more paraphrasing than others. For the individual works—Targums Neofiti, Ps.-Jonathan, Onkelos, the Cairo Geniza fragments, etc.—see Alexander 1992, Beattie and McNamara 1994, and Flesher 1995: esp. 40: “This [paraphrasing] approach enables the additions to masquerade as translation, disguising them from all but the most learned. The hidden character of the interpretive material, in turn, enables the targumist to add details, change the meaning, and even rewrite the story without the Aramaic-speaking audience being aware of it. Targum authors, then, provided their audience with a text that adhered to the original Hebrew, but at the same time presented accepted interpretations.” By whom were these interpretations “accepted”? Apparently, Flesher here means “accepted by the targumist” rather than the audience/congregation generally. I have not been able to find a clear answer to the question of whether a standard Aramaic audience would have recognized, before the rabbis instituted the parallel reading of Hebrew, that the targum was in fact a paraphrase. Flesher here suggests they would not have.
[ back ] 59. It should be noted that there are instances where Josephus uses a revised version of the LXX in his Antiquities; e.g. see Ulrich 1978:259, cited by Brock 1992:335n13: Josephus used “a slightly revised form of Old Greek [translations]” for parts of Samuel. As Feldman has shown, there is plenty of evidence that he used Aramaic translations as well, perhaps some of the targumim that have come down to us (Feldman 1998a:28–29). The earliest datable targumim are first century AD from Qumran: Job 37:10–42:11 and some fragments of Leviticus 16:12–15, 18–21 (ibid.:17).
[ back ] 60. Gruen 1998:110.
[ back ] 61. Biblical paraphrase could perhaps be seen as closely aligned to distinct categories or genres, such as commentary (e.g. much of the Qumran material), Jewish historiography (Chronicles and Josephus), or translation (LXX, targumim); however, it is not encompassed by any one of these and ultimately transcends genre.
[ back ] 62. For the salient characteristics of the Jewish novel, esp. in comparison with the Greek Romance (but not with early Christian literature), see Wills 1995.
[ back ] 63. Wills 1995:36: “The Jewish novels appear to be composed and recomposed, without the canon of a fixed text but with the canon of a traditional set of plots and characters. The study of ancient novels thus places the scholar in a difficult position between the analysis of oral and written tradition, oral and literary culture. We are addressing neither oral culture nor written culture but ‘popular written culture’ . . . Comic books, science fiction novels, and drugstore romances occupy a similar position in modern society.”
[ back ] 64. E.g. Kugel 1998:24: “Ancient biblical interpretation is an interpretation of verses, not stories.”; see esp. Kugel 1990 for his well-honed, if somewhat idiosyncratic, methodology.
[ back ] 65. Laurence Wills has produced a helpful one-volume collection of translated Jewish Novels, with introductions, notes, and bibliographies for each (2002); critical texts of these novels are not always available (due to their many recensions), nor easily found if they are—Wills includes a short guide to the disparate texts he used (2002:ix–x). The fragmentary historical paraphrases and novels (such as Artapanus) can be found with text, translation, and commentary in Carl Holladay’s four-volume collection, Fragments from Hellenistic Jewish Authors (1983–1996).
[ back ] 66. For the characteristic style of the Greek Romance, see Reardon 1991. For its influence on Christian literature, see Pervo 1987, 1996, Hägg 1983, and Johnson forthcomingb.
[ back ] 67. On the interaction of novelistic style and Jewish rewriting/interpretation, see Gruen 1998:passim and 2002:part 2, Kugel 1998 (organized according to biblical theme), and, generally, Wills 1995 (a genre-analysis) and 2002 (translations of Jewish novels).
[ back ] 68. Another way of seeing the Gospels in more than one dimension is asking, for instance, what is the relationship between Mark and John? This question of genre has been addressed in detail by Wills 1997 which takes a broader view of the question of influence and which points evocatively to a fluid exchange of literary styles and religious language in the gospel-milieu. In particular, Wills argues for a more inclusive definition of “biography” as a classical genre in order to take account of novelistic treatments of hero cults, e.g. the Life of Aesop. From the point of view of late antique Greek literature—specifically of the influence that the gospel genre had during that period—Wills’s study of the Gospels represents a salutary shift in perspective. Other studies that preceded Wills in this vein are Tolbert 1989, Burridge 1992, and Collins 1992. See Wills 1997:chapter 1 for a thorough discussion of the previous scholarship.
[ back ] 69. Luke 1:1–4 NRSV.
[ back ] 70. If the author of Luke means Mark and Q, then he is not telling the truth, for he relied upon them extensively. If he means other accounts than these, then they have not survived. The third and best option is that this statement is simply a necessary aspect of the rhetoric of historiographical prefaces. For the rhetoric of ancient prefaces (specifically Latin), see Janson 1964.
[ back ] 71. Josesphus Antiquities 1.15–17; trans. H. St. J. Thackeray, LCL Josephus vol. 5.
[ back ] 72. Gruen 2002: chapters 5 and 6.
[ back ] 73. The “Western” tradition is believed to go back to at least the third century: see Aland 1987, cited by Elliot 1996. For Codex Bezae generally, see Ammassari 1996 (the text), Parker 1992, and Parker and Amphoux 1996. For the date and origin of Codex Bezae, see Callahan 1996:57, 64: “[The scribe] worked in the environs of a Roman colony [perhaps Antinoopolis] in upper Egypt between the fourth and fifth century.”
[ back ] 74. Strange 1992:1.
[ back ] 75. Epp 1966:41–64. For the “ignorance motif,” see also Epp 1962 and, in opposition, Conzelmann 1987:104–105, 146–147, and passim.
[ back ] 76. Epp 1966:64: “The portrayal of Jewish hostility toward Jesus and of Jewish responsibility for his death in the [Codex Bezae] reveals a clearly anti-Judaic attitude. On the other hand, the strong positive emphasis on Jesus as Lord and Christ turns the sword in the wound (so to speak), for by presenting Jesus in bold and heightened tones the heinousness of the Jews’ action against him is even more strongly emphasized.”
[ back ] 77. Idem:45. P75 from the third century already contains a truncated version of this verse (Ehrman 1996:111).
[ back ] 78. Idem: 166. The Codex Bezae Acts “seems to ‘out-Luke’ Luke in its emphasis on universalism” (66).
[ back ] 79. Witherington 1984, cited by Haines-Eitzen 2000:116. “Anti-feminist” is Witherington’s; “gender hierarchy” is Haines-Eitzen’s. For this and more examples of the “suppression of women” in early Christian manuscripts, see Ehrman 1995:367–368 and 1996:114–116. “Suppression of women” as a label, however, is perhaps too convenient and anachronistic.
[ back ] 80. Epp makes the important point that Codex Bezae is not a completely new Acts of the Apostles but retains “the bulk of the traditional text” (1966:39); however, it does have enough variants for scholars to consider it an attempt to alter significantly the force of the original work.
[ back ] 81. He rejects Epp’s calling these changes “theological,” “as if they bore no relation to sociopolitical realities” (Ehrman 1993:274). Of course, the term “theological” does not de jure rule out socio-political realities.
[ back ] 82. See Parker 1992:261–78 and Ehrman 1994. In the latter Ehrman demonstrates that the text Heracleon used for his commentary on the Gospel of John in the late second century is “a comparable form of the text that was used for the first eight chapters of John by the late fourth-century scribe of Codex Sinaiticus” and by “the scribe who produced Codex Bezae” (179).
[ back ] 83. Ehrman 1996.
[ back ] 84. Ehrman’s fullest treatment of this competitive milieu is Ehrman 1993. “Proto-orthodox” means, for Ehrman, those in the first through third centuries whose theological and hermeneutical opinions were positively received by those Christians who first called themselves “orthodox” in the fourth century: see Ehrman 1993:11–15.
[ back ] 85. Ibid.:62–67. Attestations to the more difficult reading include Codex Bezae, Justin Martyr, Clement of Alexandria, Origen, Methodius, the Didascalia, Lactantius, Hilary, Augustine, and several Old Latin manuscripts.
[ back ] 86. Haines-Eitzen 2000:116. The strength of her overall argument on this point is undeniable and complementary to the present study; however, I would argue for a slightly more moderate formulation. Clearly many of the textual variants in New Testament manuscripts can be shown to be habitual, standard scribal errors, such as dittography and haplography—to name only the most straightforward—and should not be included in an analysis of the discursive networks behind the scribal project generally. Epp argues persuasively for moderation in reacting to Ur-text New Testament scholarship (1966:15–21).
[ back ] 87. Ehrman 1993:27.
[ back ] 88. Though, interestingly, Origen and his pupils seem to have had a special commitment to the book: Origen, Gregory, and Dionysius of Alexandria all wrote interpretative works on Ecclesiastes, as did Hippolytus of Rome (Jarick 1990:3).
[ back ] 89. On Gregory’s life and the sources for it, see Van Dam 1982. The main source is Gregory of Nyssa’s sermon On the Life of Gregory Thaumaturgus (ed. Heil 1990); see also Gregory Thaumaturgus’ Panegyric to Origen, written on the occasion of his departure from Origen’s school c. 240 (ed. and trans. Crouzel 1969). For Gregory’s later title, see Telfer 1936:240. On the various writings attributed to him, see Crouzel 1969:27–33.
[ back ] 90. Text is in PG 10, columns 987–1018 and is conveniently reprinted with translation and commentary by Jarick 1990.
[ back ] 91. Jarick 1990:310.
[ back ] 92. Jarick 1990:7.
[ back ] 93. Ibid.:8.
[ back ] 94. Jarick 1990:316: “In presenting the Church with this smooth paraphrase of a formerly uncomfortable work, Gregory Thaumaturgus stands firmly at the beginning of a long tradition seeking to remold Ecclesiastes into a more ecclesiastical book.”
[ back ] 95. Jarick 1990:311, citing Paraphrase 2.24, 3.12–13, 8.15–17; cf. 3.22.
[ back ] 96. Paraphrase 1.3; Jarick 1990:9, 359n25.
[ back ] 97. Paraphrase 2.16, 7.25; Jarick 1990:43, 186, 313.
[ back ] 98. Paraphrase 12.11; Jarick 1990:303, 315.
[ back ] 99. Jarick 1990:303–304.
[ back ] 100. The problem as formulated by Livrea 1989 is that the Dionysiaca is “positivamente pagano” and the Paraphrase is “positivamente ammaliato dallo splendore del Logos rigeneratore . . .” (21–22).
[ back ] 101. Sherry 1996; Coulie and Sherry 1995.
[ back ] 102. Livrea 1989; Hollis 1994:58.
[ back ] 103. The standard complete edition of Nonnus’ Paraphrase is Scheindler 1881. Livrea began a new edition (text, translation, and commentary) with the paraphrase of John chapter 18 (1989; cf. Birdsall 1990); since then, he (2000) and his colleagues Domenico Accoriniti (1996; cf. Mary Whitby 1998), Claudio De Stephani (2002; cf. eadem 2004), Gianfranco Agosti (2003; cf. Johnson 2005), and Claudia Greco (2004) have followed with John chapters 20, 1, 5, and 13 respectively. Alan Cameron’s evocative studies of fifth-century literary culture are still benchmarks for historical scholarship on the period, though he does little in the way of actual literary analysis (Alan Cameron 1965, 1982, and 2004). Golega 1930 is still the standard stylistic analysis of Nonnus’ Paraphrase, but see now Hollis 1994 (and forthcoming) in connection with the reception of Hellenistic poetry in the Dionysiaca.
[ back ] 104. Livrea 1989:19.
[ back ] 105. Alan Cameron 1965:485.
[ back ] 106. See Alan Cameron 1982.
[ back ] 107. More specifically centered around Nonnus and his (unknown) students than Golega’s soggenannte Nonnosschüler, which include Musaeus and the Pseudo-Apolinarian Paraphrase of the Psalms (Golega 1960:93–108).
[ back ] 108. This conclusion emerged out of his 1991 Columbia dissertation on the Paraphrase. Note, however, Alan Cameron’s and Sherry’s conflicting estimates of the literary value of Nonnus’ Paraphrase—Alan Cameron 1982:284: “Nonnus (if he it was) treated his model with the utmost freedom, producing an elaborate rhetorical masterpiece in the high style scarcely inferior in its way to the Dionysiaca”; by contrast, Sherry 1996:411, 414: “Why are there so few testimonia for the Paraphrase? I suggest that it is because the poem is not by Nonnus. Since it was not a serious piece of literature and a poem inferior to the Dionysiaca, it did not warrant the same attention from readers and collectors . . . Nonnus was too good a poet to produce so lame a paraphrase.”
[ back ] 109. Sherry 1996:414 and n26. See Golega 1930:143: “Und doch weist die Paraphrase fast noch mehr nonnianische Floskeln auf als Musaios, dessen Epyllion ohne weiteres auch in den Dionysiaka Platz finden könnte. Ja man darf die Paraphrase beinahe als einen Cento aus Dionysiakaversteilen und Evangelientext bezeichnen” (29); and “Die sprachlich-stilistische Übereinstimmung zwischen beiden Gedichten ist so groß, daß die Paraphrase fast ein Cento aus Dionysiakaversteilen in Evangelientext genannt werden kann” (emphasis added).
[ back ] 110. See Golega’s Zussamenfassung (1930:142–144).
[ back ] 111. E.g. Alan Cameron 1982:284; Hollis 1994; Mary Whitby forthcoming; Mary Whitby 1998 (review of Accorinti 1996): “a storehouse of ammunition is accumulated against the cento thesis.”
[ back ] 112. Usher 1998:2 offers the suggestion that the cento is technically not a genre but what he calls simply an “écriture,” which, like parody or pastiche, can take various prose and verse forms (citing Verweyen and Witting 1991:172).
[ back ] 113. See MacDonald 2000:5, with extensive references at 205n14.
[ back ] 114. Trans. ibid.; Quintilian Institutio Oratoria 10.5.4 (cf. 1.9.2–3 and Cicero On Oratory 1.154); Philodemus On Poetry 5.30.36.
[ back ] 115. See Browning 2000:868 and passim.
[ back ] 116. Sherry 1996:420: “The [Nonnian] paraphrase has a unique place in the history of Greek literature. It is not only the sole surviving New Testament paraphrase, but it may well be the only one ever attempted”—a very inaccurate and misleading statement.
[ back ] 117. Averil Cameron 1998:672: “in so far as a Christian consciousness came into being, it was moulded by scriptural patterns, both inside and outside the Christian élite.”
[ back ] 118. There is no consensus on which recension of the centones is Eudocia’s: see the succinct treatment in Mary Whitby 2001. This question has been dealt with in depth by Usher 1997 and 1999, Rey 1998, Schembra 1995, and Whitby forthcoming, all with different conclusions. It is possible that none of the recensions is Eudocia’s, but most scholars have settled on one or the other manuscript tradition, Usher preferring a longer fourteenth-century manuscript from Athos, Shembra a shorter recension incompletely edited by Ludwich 1897, and Rey accepting multiple authorship in the shorter version—see Mary Whitby 2000 for some of the interpretive implications of this debate. If one accepts Usher’s longer recension, then Eudocia’s Homerocentones, at twenty-four hundred lines, becomes by far the longest of the surviving centones. For a list of the other known Homeric centones with references, see Usher 1998:3n3.
[ back ] 119. Usher 1998:16–17. A inspired comparison to be sure, but I hardly think Homer was “discarded material” in late antiquity.
[ back ] 120. Usher 1998:ix–x.
[ back ] 121. This juxtaposition is also suggested in general by Mary Whitby forthcoming—disagreeing with both Alan Cameron 1982 and Urbainczyk 1997, she writes: “One might more cautiously suggest that Theodosius’ [II’s] combination of educational and pious objectives provided an ideal environment for experimentation with this combination in literature.”
[ back ] 122. Photius Bibliotheca 183–184; ed. Henry 1960:2.195–199. The entry for Eudocia in Bowersock, Brown, and Grabar 1999:436 is erroneous in saying that only the paraphrase of the martyrdom of Saint Cyprian has survived, ignoring completely the more significant Homerocentones (a belief, if held, that the latter is wrongly attributed should have been noted and defended).
[ back ] 123. Alan Cameron 1982 has emphasized that the reorganization of schools in Constantinople in 425 should be seen on the background of imperial politics: “After 425 education in Constantinople was in effect the monopoly of a Christian government” (287). This is certainly important, but is it not also possible to see, from a literary-historical point of view, the persistent strength in the fifth century of traditional modes of rhetorical training and biblical exegesis and, then, the contemporary “christianization” of these modes? See n. 121 above.
[ back ] 124. While both Alan Cameron 1982:282 and Sherry 1996:425n58 rightly (though only in passing) cite the Life and Miracles as a comparandum for Nonnus and Eudocia, both appear unaware that its author is not Basil of Seleukeia, accepting the mistaken Byzantine attribution and confusing it with Photius’ notice. For the authorship of the Life and Miracles, see Dagron 1974.
[ back ] 125. Socrates Ecclesiastical History 3.16.1–7; trans. A. C. Zenos NPNF 2nd series, 2:86–87 (translation altered); cf. ed. Günther Hansen 1995:210. Note also how the technical language for paraphrase appears different here, esp. μεταβάλειν instead of μεταφράζειν.
[ back ] 126. Sozomen Ecclesiastical History 5.18.1–5; trans. C. D. Hartranft NPNF 2nd series, 2:340 (translation altered); cf. ed. Bidez and Hansen 1995:221–223.
[ back ] 127. Golega 1960.
[ back ] 128. Urbainczyk 1997:33–34.
[ back ] 129. Alan Cameron 1982:283. Nonnus could have written his Paraphrase prior to 439 since it is possible Socrates would not have known it, and there are no known connections between Nonnus and the court; by contrast, the empress Eudocia could presumably not escape notice.
[ back ] 130. Urbainczyk 1997:33–34.
[ back ] 131. Though Sozomen’s approval of the practice could be read as an implicit acknowledgment.
[ back ] 132. How do we explain Socrates’ harshness in this matter? Besides assuming a distaste for the younger Apolinarius, there is no clear answer. Nevertheless, it is important to note that both historians set the Apolinarii in the same context. They highlight the educational environment from which the paraphrases come and, in their own ways, they obscure the broader tradition of paraphrase through their specific denigrations of Julian’s policies.
[ back ] 133. Bovon 1988:19–20 emphasizes the fact that this vigorous activity was ongoing even in recent times: “At the same time as Konstantin von Tischendorf was preparing his critical edition of the martyrdoms and apocalypses of the apostles, a Greek monk from Palestine [Joasaph of Saint Sabba] was retelling in his own style the same stories which Tischendorf and R. A. Lipsius and M. Bonnet were editing” (see references ad loc.); contrast this observation with the following: “Today no one dreams of publishing interpolated versions of these [canonical] Gospels or of doctoring our holy books” (ibid.)—we have thus inherited a cognitive distinction (formulated sometime between the second and sixth centuries?) between inviolable and violable Christian texts.
[ back ] 134. See Ehrman 1993:17–20: The first attempts to restrict the Christian canon were not voluntary but came only in response to heretical (e.g. Marcionite) canonical definitions. On conceptual distinctions in the second and third centuries between canonical and apocryphal Gospels, see Bovon 1988.
[ back ] 135. See Høgel 2002.
[ back ] 136. It is often the case that the novelistic elements are highlighted by these authors as much as the historical. See Hägg 1983 chapter 6 and Pervo 1996.
[ back ] 137. The term “Encratites” comes from ἐγκράτεια, “self-control” or “continence”; while this label probably refers to various different sects with Gnostic connections, the second century writer Tatian is often said to be their heresiarch. They are described by Irenaeus (Against Heresies 1.28), Clement (Paedagogus 2.2.33; Stromateis (“Patchwork”) 1.15, 7.17), and Epiphanius (Panarion (“Medicine Chest”) 47.2.3–47.3.1), among others. In addition to sexual continence, they were said to have abstained from wine and meat as well, though it is unclear whether abstinence from these two were also necessary for salvation.
[ back ] 138. As does in fact appear to be the case when Origen cites the Acts of Paul in On First Principles 1.2.3 and his Commentary on John 20.12 (Elliott 1999:350).
[ back ] 139. The critical text of the Acts of Philip is Bovon, Bouvier, and Amsler 1999; French translation, Amsler, Bovon, and Bouvier 1996. On the religious community that produced the Acts of Philip see the references at Bovon 2001:140n10, esp. Slater 1999.
[ back ] 140. On the Council of Gangra in the context of the extreme eastern asceticism of the fourth and fifth centuries, see Caner 2002, esp. chapter 3; in addition, see the references in ODCC s.v. “Gangra, Council of” and “Eustathius,” esp. Gribomont 1957, 1980, and Barnes 1989, and, for the text of the Council (20 canons in Greek and Latin) with a French translation, see Joannou 1962–1963:1.2.83–99.
[ back ] 141. On the Byzantine reception of early Christian apocrypha, see Patlagean 1991.
[ back ] 142. Bovon 1999a argues against Lipsius 1883–1890:2/2.233, who said that a later use of this lost apocryphal material on James, by the Byzantine historian Nicephorus Callistus Xanthopoulos, is taken from Nicetas. Bovon argues that they both independently attest these lost Acta, demonstrating that standard versions were in circulation for a considerable time.
[ back ] 143. Critical text: Zahn 1975 [1880]:3–165. See Krueger 2004:216n15 for references to modern translations.
[ back ] 144. Rapp 1995.
[ back ] 145. See Høgel 2002. Between Sophronius and Symeon comes, of course, Nicetas David of Paphlegonia in the ninth century, mentioned above in relation to the lost acts of James the brother of Jesus (Rapp 1995:35–36).
[ back ] 146. It should also be noted that the metaphrastic trend continues for several centuries after Symeon as well: see Talbot 1991, cited by Rapp 1995:36n24.
[ back ] 147. Høgel 2002:89–126.
[ back ] 148. McKenzie 1999 [1986]; see Chartier 1997 for differing reactions to McKenzie’s seminal lectures.
[ back ] 149. McKenzie 1999:37, citing Doctorow, Wittock, and Marks 1978; emphasis is McKenzie’s.
[ back ] 150. Vinel 1987:213.
[ back ] 151. See Pucci 1998 for a recent restatement of the value of allusion in late antique literature.
[ back ] 152. Boyarin 1990:15.