On Making Manuscripts, Genre, and the Boundaries of Ancient Jewish Literature

Using literary qualities as variables rather than categories allows us to reimagine and redescribe hybridity in our archive. The use of taxonomic-style categories can imply a kind of closure of inquiry, or elimination of alternate possibilities. So, an apple is a fruit, which means to the casual grocery-store shopper that it is not a vegetable. But the flummoxed shopper might be faced with a tomato, both a fruit and a vegetable. Categories might promise to order complex systems, but ultimately cannot reduce diversity without remainder. But, variables, as I interpret Zahn to propose, represent descriptors of difference that do not end the process of exploration. We now have the theoretical tools to handle metaphorical ‘tomatoes’–entities for which multiple classifications are operative. Speaking in terms of ancient Jewish literature, and Zahn’s proposal, we might think of texts which contain both law and prophecy, or first and third person voicing, or the rewriting of one source alongside another source, or ‘sectarian’ alongside ostensibly non-sectarian material.

This newly describable hybridity can teach us about ancient understandings of coherence. To use a phrase by Annette Yoshiko Reed, we can now discover ancient practices of “constructing continuity” when we’re equipped with the variables proposed by Zahn (Reed 2017).

One thing that I want to bring into the conversation is some material evidence for ancient ‘groupings,’ possibly hidden from view because of the field’s legacy of imagining literary categories in a certain way and editing scrolls accordingly.^[1] I will present a few places in which scholars have, on literary grounds, created multiple manuscripts out of fragments that could, materially, be sorted into a single manuscript. The idea of a work in Second Temple Judaism is complicated by the reality that many of our works are constructed—fragments from Qumran are joined into manuscripts according to certain material, but often literary, considerations. But, if we reimagine the ways that literature might be taxonomized—with groupings, and alternative groupings—we might find ourselves confronted with new works.

Speaking to Variable 1, Literary Form: A key example is 4QS^e (4Q259), a manuscript of a work known as the Community Rule, or Serekh ha-Yahad, which (among other things) provides a series of social and religious regulations meant to help structure the life and practice of an ancient Jewish group. Text from 4QS^e resembles and occasionally overlaps with text belonging to other manuscripts also classified as manuscripts of the Community Rule, like 1QS.

The initial editor of 4QS^e, J.T. Milik, recognized that the fragments now sorted into 4QS^e (4Q259) belonged with fragments now sorted into another manuscript called 4QOtot (4Q319).^[2] In his reconstruction, they represented different columns in the same scroll. It is also advanced by the publishers of these fragments in the DJD volumes that they were written by the same scribal hand, and based on this and other material factors, belonged to the same scroll. ^[3] Nevertheless, the fragments were assigned to two different manuscripts, as 4QS^e was subsequently published in a different DJD volume, and assigned a different number, than 4QOtot. The grounds for this assignment of the fragments to two manuscripts, or the date at which such assignment was undertaken after Milik’s initial treatment, are not always entirely clear.^[4] But at least part of the editorial decision might lie in literary factors, as 4QS^e is a recognizable ‘serekh’ text, whereas 4Q319 better resembles Qumran’s calendrical and mishmarot texts.

I want here to reflect on how the boundaries of manuscripts guide our conception of literary works. It is not necessarily the case that one manuscript houses one, and only one, literary work. Some manuscripts, like 11QPsalms, lend themselves readily to thinking about a scroll as a collection, a gathering-site of literary entities, and it is reasonable to suggest that manuscripts at Qumran contained multiple works identifiable as such to ancient readers. In some cases, vacats or spacing might be used to indicate the end of an identifiable work and beginning of another.^[5] If 4QS^e-4QOtot is one such manuscript (though we do not have the transition point between the ‘two’ preserved), the separate numerical classifications granted to its fragments would signal literary hybridity of a kind, though not their material separability. And so, even though a manuscript need not house a single work or literary entity only, perceived literary hybridity has nevertheless resulted in the creation of multiple manuscripts.

For another way to think about it, although the editing of separate manuscripts or giving of separate numbers (e.g. 4Q259 & 4Q319) does not necessitate a corresponding hypothesis of separate works, in practice it often prompts the modern reader of the Scrolls to think in such a manner. A sense of a boundary line is signaled in the editing of these fragments separately. What remains to be reclaimed is the extent to which these fragments studied together represent evidence for the crossing of boundary-lines, and material practices of constructing continuity. We are here confronted with a production in which serekh-style and calendrical text were constructed into a continuous whole. At some point in the editing of these fragments, observable hybridity led to the editing of separate manuscripts. If the fragments are reunited, we can learn about one instance of constructed continuity within Zahn’s first variable.

Speaking to Variable 2, Voicing: Another illustrative example comes from the manuscript history of 1 Enoch. The initial editor, J.T. Milik once again, identified the same scribal hand behind 4Q207 (a fragment of the work known as the Book of Dreams) and fragments belonging to a copy of of Aramaic Levi (4QAramaic Levi^d, or 4Q214).^[6] In other instances, Milik used the same scribal hand as a key principle in the creation of manuscripts. For example, Milik edited the manuscript known as 4Q204, in which he gathered fragments corresponding to four different literary works: the Enochic Book of the Watchers, Book of Dreams, and Book of Noah (these three all eventually collected together in the Ethiopic collection now called 1 Enoch), alongside a work he called the Book of Giants. Outside of Qumran, the Book of Giants is never found in a collection with the aforementioned Enochic works—there is no external precedent by which to imagine a literary work bounded in such a way. Instead, material consonance (the same hand, similar layout and spacing) met Milik’s own understanding of consonance in content, to create his understanding of an Enoch compilation manuscript comprising the Book of Giants alongside the other Enochic works.^[7] Other scholars, working after Milik, treated the fragments belonging to the Book of Giants under the heading of a separate manuscript numbered 4Q203, and have not always shared Milik’s opinion that 4Q203 and 4Q204 belong to the same manuscript.^[8] Nevertheless, from Milik on down, there has been a boundary line drawn between the Book of Giants and the other Enoch works, even though material judgments concerning the scribal hand might encourage sorting their fragments together.

One might wish to be wary of hanging too much on the hypothetical identification of scribal hands at Qumran, as it is a matter of opinion whether paleographic difference should be interpreted as, say, one scribe altering their style, or two scribes, or multiple scribes operating within a similar stylistic framework, and so on.^[9] But even with these methodological concerns, the point is that the literary imagination of Scrolls scholars has often had as much to do with editing practices as the materiality of the scrolls themselves.

To return to the Enoch and Levi material: the Enoch manuscript we are exploring now, 4Q207, is actually a single fragment, so there is hardly enough material evidence to speak to the physical construction or layout of the scroll to which it belonged. This minimal evidence makes it difficult to enact an informed comparison with 4Q Aramaic Levi (4Q214). But, as my phrasing indicates, Milik assigned these fragments to two different manuscripts, most probably in keeping with the conviction that the Animal Apocalypse and Aramaic Levi represent separate works, though he did not offer explicit comment. Against this assignment, the identification of the same scribal hand could also suggest a single scroll collecting Enoch and Levi pseudepigraphy. This possibility is not a necessity—a single hand does not always or necessarily signal the creation of a single manuscript—but it is a real possibility that has not been previously addressed in scholarship on these fragments. We might imagine a manuscript shaped like the Genesis Apocryphon, itself a construction of multiple pseudepigraphic perspectives, redacted into a whole. Or we might imagine a manuscript shaped like our Psalms collections, in which the distinction between units seems to be preserved, even as they are collected alongside one another. With such poor preservation of our materials, in this instance, it may not be possible to decide which style of collection might be operative here. But, even with the exact relationship of the materials collected together undecided, perhaps we should still look for deeper literary affinities than simply interest in pseudepigraphic voicing—the focus on the Temple in the Animal Apocalypse might be consonant with the sacrificial instructions attested by 4Q214 in particular, and Aramaic Levi more generally.

These fragments are not well preserved enough to sufficiently provide guidance to their editor, and it seems the documents have been edited separately primarily on literary grounds. But if our practices of literary mapping shift to accommodate greater hybridity, these fragments might be rebound into a new unity, one which speaks to certain continuities in ancient understandings of pseudepigraphy.

Speaking to Variable 3, Observable Relationship to Known Texts: We now arrive at a more complicated example, because of the greater volume of fragments implicated in the discussion. One area that has already known its share of controversy on its legacy of editing, in conversation with changing evaluations of the literary work or works found therein and their relationship with known texts, is the fragments edited into manuscript(s) labeled 4Q385-4Q390. We can summarize these with reference to the much-debated assignment of fragments to two manuscripts—4Q385 (Pseudo-Ezekiel) and 4Q385^a (Apocryphon of Jeremiah C)—and an ongoing discussion about how this assignment ought to take place.

The major grouping enacted among these materials is a division between two previously unknown works: the Apocryphon of Jeremiah, and Pseudo-Ezekiel. The material considerations according to which the fragments are assigned have been much debated.^[10] According to the initial editor, Devorah Dimant, relying upon John Strugnell’s early notes, all the fragments now sorted into 4Q385 and 4Q385a came to the Palestinian Archaeological Museum in a wad, a datum that is often a sign that the fragments belong to the same scroll. And Dimant notes, “[m]ost of the fragments initially assigned to 4Q385 display a very similar, even identical, hand and resemble each other closely in material appearance” (Dimant 2001, 129). We have discussed above the extent to which the same hand has been a binding principle in the creation of Dead Sea manuscripts by scholars confronted with fragments. But Dimant understood the literary difference between material engaging the text and legacy of Jeremiah, and material engaging the text and legacy of Ezekiel, to be so stark as to mandate the division of the fragments into two works—4Q385, and 4Q385a. Monica Brady has argued that these fragments should not be subdivided into two works, capitalizing on the physical and material similarities between the fragments assigned to 4Q385 and 4Q385a, maintaining instead that Ezekiel and Jeremiah traditions could be pursued under the heading of the same work (Brady 2001, 2005). In response, Devorah Dimant argues that viewing the fragments as remnants of a single work results in a “strange and mixed assemblage…such a combination is unknown in any of the apocryphal and pseudepigraphic compositions” (Dimant 2011, 21).

I will not make a judgment on which reconstruction is the more plausible, nor will I comment on different evaluations of the material factors at play. What I do think important is that this is a site in which scholars are debating the extent to which we want to allow hybridity with regards to Zahn’s third variable, the observable relationship to known texts, to create boundaries. If the fragments are assigned to separate manuscripts, as suggested by Dimant, such a reconstruction is in consonance with a literary understanding of pseudepigraphy as generally bound to engagement with the legacy of an individual prophet. But if the fragments are understood together, as suggested by Brady, this would speak to practices of constructing continuity across the legacies of individual prophets. Although I find this case much more opaque that my other examples, given the tangled history of construction and reconstruction of manuscripts, and the ambiguity of the material data, it is nevertheless a site that might benefit from Zahn’s new mapping metrics.

These have been, briefly, cases in which perceived hybridity using the variables detailed for us by Zahn may have represented a force of division in the history of scholarship. Serekh and calendars were assigned to two separate manuscripts (4Q259 and 4Q319). Enoch and Levi were assigned to two separate manuscripts (4Q207 and 4Q214). And, though this is a controversial case, Ezekiel and Jeremiah materials have been variously assigned to 4Q385 and 4Q385a, while other scholars reject such an assignment. These are all cases in which literary considerations have sometimes encouraged the siloed and separate editing of documents, and hypotheses of separable works. But I have suggested that we might capitalize upon material commonalities, like the identification of the same scribal hand across certain fragmentary sets, to explore new possibilities in reconstruction that might work in concert with our developing imagination of Second Temple literature. It might be possible to see these same manuscripts as sites in which continuities are constructed, and from which we might learn more about the way that ancient Jewish readers understood coherence and categorization in their own archive. Zahn’s variables describe valuable sites of literary investment in the Second Temple period, but also are flexible enough to allow us to spot hybrid instantiations of these variables and learn about certain meta-coherences in Second Temple literature.

And so, I have here attempted to demonstrate one way in which Zahn’s remapping of literary variables—and an openness to these markers as variables, rather than confining categories—leads to a reimagination of the very works in our archive.

Acknowledgments

Thanks to Molly Zahn for opening the conversation, to Seth Sanders and James Nati for inviting me to participate, to Eibert Tigchelaar and my anonymous reviewer for their helpful comments, and to the organizing community of the BRANE Collective for making this exchange possible.