The Re-Emergence of Source Criticism: The Neo-Documentary Hypothesis
The Documentary Hypothesis, abandoned in much pentateuchal scholarship of the last 40 years, is making a significant resurgence, although in a new and more precisely argued form. It is once again taking its place as a significant theory of the composition of the Pentateuch.
The Composition of the Pentateuch: Renewing the Documentary Hypothesis (Anchor Yale Reference Library; New Haven: Yale University Press, 2012).
By Joel S. Baden
Assistant Professor of Old Testament
Yale Divinity School
It is perhaps overly dramatic to refer to the existence of two competing schools of thought as a crisis. It is, after all, difficult to think of a single aspect of biblical scholarship in which there are not dissenting opinions, contradictory readings, or diverse historical reconstructions. The various opinions regarding the origins of the Dead Sea Scrolls, for instance, do not constitute a crisis in Qumran studies. Such instances of scholarly disagreement are points of dispute, or on which a consensus has yet to be reached. Yet references to current pentateuchal research frequently use the term crisis to describe the state of the field.
This is in part because the divide between the two schools of thought in pentateuchal scholarship is particularly wide. Geographically, it is essentially the size of the Atlantic Ocean, with one school located almost entirely in continental Europe, and the other mostly at home in the United States. Theoretically, it is just as large: the fundamental basis for each schools approach is essentially rejected by the other. Moreover, the shift in European scholarship away from the Documentary Hypothesis happened very quickly: one can almost draw a line at the publication of Rendtorffs Problem of the Process of Transmission in the Pentateuch in 1977,1 with pre-Rendtorff scholarship being largely documentary and post-Rendtorff scholarship being almost entirely non-documentary.
For scholars trained in the Documentary Hypothesis, as most Americans still are, theres a feeling that if you blinked at the wrong moment, you missed the sudden emergence of a radical new wave of scholarship. And given the profusion and depth of European scholarship, it can be hard to catch up. It is also the case that, as there are very few American scholars teaching the current non-documentary view, American scholars have not had a systematic and thorough introduction to the newest pentateuchal research. The result is that, for many, what is known is that the documentary model we learned in school has been discarded, but we have not accepted its ostensible replacement.
Add to this the centrality of the question of the composition of the Pentateuch for all sorts of biblical scholarship, past and presentincluding, but not limited to, the early history of Israel, the development of Israelite religion, biblical theology, the study of oral traditions, and the dating of biblical textsand it becomes more apparent why pentateuchal scholarship represents a crisis. If, as an American scholar, you want to speak or write about any topic that even lightly depends on the analysis of the Pentateuch, you may find yourself at a loss: either you continue using the older scholarship you learned, but which you know is now considered passé, or you have to undo what you learned and fully immerse yourself in the newer European approach. Few are willing to do the latter; we here largely still learn and teach the documentary model, but are subsequently unable to apply it in any productive manner without fear of being labeled out of date. As a result, scholarly output from America in pentateuchal studies has sharply diminished in the last forty years. This may represent a legitimate crisis.
Today, however, this situation is changing. Serious, thorough, textually and methodologically rigorous documentary scholarship is being produced. In one sense, in the theoretical sense, this deepens the crisis: now there are, indeed, two fundamentally conflicting views on the composition of the Pentateuch. In another sense, however, in the practical sense, the crisis may be improving: those scholars who have been adrift in the great pentateuchal sea may find that their long-held views have a home again. In short, the Documentary Hypothesis is regaining its place as a viable, productive, and current approach to the Pentateuch.
This is not to say, however, that the Documentary Hypothesis of the early twentieth century is simply being restated. On the contrary: one of the main contributions of more recent source-critical work has been the identification and correction of the methodological problems that plagued earlier scholarshipthe very problems which, it is fair to say, contributed significantly if not primarily to the move away from the Documentary Hypothesis in Europe in recent generations. Two of these methodological problems are worthy of particular attention: the question of stylistic differences as a means of distinguishing among the sources, and the question of similarity of content among the various sources.
For generations, the most commonly cited basis for differentiating one source from another has been the issue of style and terminology. Lists are produced: lists of words that belong to one source or another. I dont mean to suggest that there arent words or phrases that are unique to a single source, because there certainly are. The problem is that these lists are taken as prescriptive: when you come across a particular word, you check the list to see which source it belongs to, and then you know which source youre looking at. Such a procedure seemingly ignores the basic notion that all of the biblical authors knew Hebrew, and were perfectly capable of using any word from the language that they wished. We must therefore be cautious in how style and terminology are used in source-critical work. These features are not prescriptive: they do not tell us which source a passage belongs to. They are, rather, descriptive: once we have isolated the sources in their entirety throughout the Pentateuch, then, and only then, are we at liberty to say what, if any, words or phrases are unique to a given document.
The other major methodological problem with classical source-critical analyses was the implicit belief that the sources must have told the same stories in the same manner. Although there are some episodes that are obviously unique to a single documentthe Akedah, for examplein many cases scholars believed that more than one source must have told a given story. Even in ostensibly unified passages, scholars went to great lengths to pull verses out and attribute them to another source, solely because they required evidence of another source. There is no a priori reason to think that the sources must be so very similar, however. Indeed, given their theological differences, why should we be surprised by differences in plot as well?
Methodological problems are not the same as theoretical problems, however. The Documentary Hypothesis, it must always be remembered, is precisely that: a hypothesis. It is an attempt to explain the literary phenomena of the Pentateuch: clear narrative contradiction, repetition, and discontinuity. It posits that the best explanation for these features is the existence of four independent documents that were combined into a single text, basically the canonical Pentateuch as we now have it. It is the literary solution to a literary problem, no more and no less. Scholarly claims regarding stylistic criteria or similarity of narratives are not inherent parts of the theory; they are aspects of the methods used to argue for the theory. If they do not succeed, the theory does not of necessity fail; the methods do. The theory may simply need to be argued on different grounds. Thus the very correct criticisms of anti-documentary scholars from the earliest days of the theory until our own time are not necessarily grounds for dismissing the whole hypothesis; they are, rather, a call to refine and revise the methods employed by scholars when describing and applying the hypothesis. When such refinements and revisions are undertaken, as they have been recently, the Documentary Hypothesis regains its place as the most economical, comprehensive explanation for the literary phenomena of the canonical Pentateuch.
David Wright has termed the recent source-critical approach the Neo-Documentary Hypothesis, a label which is gaining some use among its adherents and others. What, then, is the shape of the Neo-Documentary Hypothesis? How does it differ from its earlier incarnation; how does it take into account the methodological problems of classical documentary scholarship; how does it help us to better understand the composition of the Pentateuch?
First: Whereas classical scholarship more often than not took stylistic and terminological markers as the starting point for the division of the text, the Neo-Documentary Hypothesis recognizes that these elements are valuable only as secondary, supporting criteria. Instead, we place at the forefront of the analysis plot and narrative continuitythe events that occur, the sequence in which they occur, cause, and effect. The mark of an author is his creation of and adherence to a distinctive and definable set of narrative claims: who did what, when, where, and how. Where these claims are contradictory, we must consider that a different author is at work; where they are the same, there is no need to pursue any source division. It is no small irony that in current non-documentary scholarship, style and terminology have re-emerged triumphantly as the fundamental basis for analysis, and with far greater demands for linguistic similarity than classical source-critical scholarship ever required.
Second: The belief that the documents must have all told the same stories in the same way, a hallmark of classical scholarship, is discarded in the Neo-Documentary Hypothesis. In its place is the recognition that there is no reason whatsoever that each source could not and indeed should not tell the stories however it wished. The sources in fact tell very different stories within the same larger framework, with different episodes, in different orders, and with very different viewpoints. This recognition allows for literarily unified passages to remain so, and also allows for simpler source divisions.
Third: For generations now, the Documentary Hypothesis has been considered synonymous with Wellhausens reconstruction of the evolutionary growth of ancient Israelite religion. The source division and the placement of the sources in a straight line of development from earliest to latest, from naturalistic to legalistic, has been taken as the fundamental claim of the hypothesis. This is demonstrated by the attempts in scholarship to debunk the Documentary Hypothesis by arguing against Wellhausens view of Israelite religion, as if the former is dependent on the latter. On the contrary, however, it was Wellhausens source division in his Composition that allowed for his historical reconstruction in the Prolegomena.2 In the first book, he addressed only the literary evidence; in the second, he addressed only the historical questions. The Neo-Documentary Hypothesis returns to the first stage, and leaves the second unconsidered. The literary question is primary, and is in fact the only question that can be answered by the documentary theory. Even if one disagrees with or disproves the arguments of Wellhausens Prolegomena, the literary analysis of the Pentateuch stands on its own merits.
Fourth: In the classical model, the sources were understood as representing discrete historical periods, and were therefore dated accordingly. The order J-E-D-P was almost universally accepted, and was made the basis for much of the analysis. In the Neo-Documentary Hypothesis, the absolute dating of the sources is not a topic of investigation. There is little in the sources themselves that allows for any absolute dating. What is possible is relative dating, though only in one particular case. The relationship of D to E and J makes clear that D was written after the other two non-priestly documents. Yet whether J or E came first, or how P fits into this picture, are questions for which the literary data simply do not provide evidence. Nor does the theory rest on any specific dating of the documents: if all four were written within twenty years of each other, the literary evidence would not change; if J were written in the tenth century and P in the Middle Ages, the literary evidence would not change. The dating of the sources has no impact on the Neo-Documentary Hypothesis.
Fifth: The classical source theory was often concerned with identifying the various strata that made up the individual sources, positing J1, J2, etc. Further, the presence of strata in the sources was used to solve supposed internal discrepancies within the sources. The Neo-Documentary Hypothesis is concerned only with the penultimate form of the text: what the compiler had at hand when he put the four documents together. This approach allows for far greater clarity in addressing the question of how the Pentateuch came to be this way, for it goes back only a single step. It is crucial to note, however, that the Neo-Documentary Hypothesis does not deny the internal growth of the sources; it is simply unconcerned with them. Like so much else, how each source came to look as it does is a secondary question. The Neo-Documentary Hypothesis does not deny that each source has a history; nor does it deny that the Pentateuch itself has a history after the compilation of the documents. It is a restricted answer to a restricted question.
Sixth: The classical approach posited at least three redactors for the Pentateuch, each responsible for one stage in the evolutionary growth of the text. The Neo-Documentary Hypothesis posits a single, almost mechanical compiler, who was responsible for the combination of all four sources. The compilers work was entirely literary: it was no more than the combination of the four documents into a single story, with the rare small adjustments and insertions that contributed to that process. Literary activities that do not participate in the process of combining the source documentsglosses, secondary additions, theological revisionsthese are not part of the compilers work, and are not attributed to the compiler.
Seventh: The classical theory began as a fairly simple proposition: four independent documents, combined into a single Pentateuch. Over time, however, it expanded dramatically, so that even within a generation or two of Wellhausen the analysis of the Pentateuch required innumerable sigla, regular divisions of the text into half-verses and even single words, and highly complex theories about redaction. The unwieldiness of this theory inevitably led in part to opposition, as it could no longer be said that the Documentary Hypothesis was a particularly simple or elegant solution to the problems of the pentateuchal text. Ironically, of course, the newer analyses coming out of Europe are, if anything, even more complex than the most tortuous classical source-critical work. The Neo-Documentary Hypothesis restores the simplicity of the earlier scholarship. It requires precisely four sources and one compiler.
The Documentary Hypothesis, in general and in its particulars, is a literary solution to a literary problem, and no more than that. It does not begin with the search for sources in the text: the sources are the conclusion of the theory, not its beginning. It begins with the canonical text, and the literary problems that require explanation. Why the Pentateuch is incoherent: that is the driving question of all critical enquiries into the composition of the text, and the Neo-Documentary Hypothesis is the most comprehensive and economical answer to that question.
1 Sheffield: Sheffield Academic, 1977.
2 Julius Wellhausen, Die Composition des Hexateuchs und die historischen Bücher des Alten Testaments (Berlin: G. Reimer, 1885); Prolegomena to the History of Israel (Atlanta: Scholars Press, 1994).
Niels Peter Lemche
Niels Peter Lemche
Is this accurate? My impression is that there is a predominantly literary approach to the final state of the text which cares little for the hypothetical sources because many of the criteria originally claimed to be decisive in recognising the extent and nature of those sources have been shown to be spurious (one think also of Whybray's The Making of the Pentateuch) and then the various source hypotheses which not only claim that it is possible to recover sources but that there is also value in doing so.
"we place at the forefront of the analysis plot and narrative continuity—the events that occur, the sequence in which they occur, cause, and effect…"
This is certainly an improvement over the methodology underpinning earlier source theories, but ISTM that it still faces a number of problems. First, the analysis presumes a set of norms against which the text is measured. Whybray demonstrated that the previous methodology placed anachronistic expectations on the biblical text. How can we be certain that the Neo-DH is not guilty of the same error?
Second, the underlying presupposition is that the final editor was an idiot. Rather than combining sources into a coherent narrative, they've produced a rather slapdash conglomeration of material. But if editors can be idiots, can't authors also be idiots? Why attribute narrative features which we think are indicative of disparate sources to editorial incompetence rather than authorial incompetence? Or perhaps, in light of the first point, why not attribute them to the incompetence of the readers?
For some considerable time the DH was treated as a given, but — as you've noted — it fell from grace (and with good reason). I suspect that's the danger with any hypothesis that attempts to define the nature of sources without any physical evidence of the source.
Perhaps the book addresses these issues in full and I should read it, and then perhaps wait for David Clines to update "New Directions in Pooh Studies"!
Also, you allow for the internal growth of sources (J1, 2, 3, etc.). But what about mutual influence between sources?
What methodological control do you give to empirical examples of scribal transmission in the ANE lit, DSS, LXX, etc?
There are, of course, many more theories of pentateuchal composition than just the two I mention in this piece, but it is safe to say that among active pentateuchal critics, at least as evident in conferences, volumes of collected essays, etc., these are the two most prominent theories in play in the field today.
As to the potential for anachronistic expectations of narrative, I think that this is not really a significant concern. We are not importing modern notions of coherence into the biblical text - the Bible itself is replete with examples of narratives, from small to large, that are perfectly readable in what we might consider a "modern" sense - consider the books of Ruth or Esther, for example. What makes the case of the Pentateuch stand out is precisely its deviation from the biblical norm.
I am most interested, however, in responding to your claim that the compiler was an "idiot." Nothing, to my mind, could be further from the truth. That would be the case only if we imagine the compiler was attempting to create a perfect, contradiction-free story; yet it is eminently clear that such is not the case. If, on the other hand, the compiler was trying to preserve his sources as much as possible while still retaining the basic chronological sequence of events, then he did a magnificent job. And since the literary evidence suggests that this is indeed what he was trying to do, then we have no reason to imagine otherwise.
I will conclude this response, at least, by noting that the observation of literary difficulties in the Pentateuch can hardly be attributed to the incompetence of the reader (i.e., the modern scholar) - the classical rabbis and the medieval Jewish commentators, to my mind the best readers of the text ever, responded to and dealt with virtually every single one of the problems that modern scholars see. Indeed, some of the contradictions in the Pentateuch are clearly at issue even for later biblical authors. We are neither inventing nor anachronistically imagining the problems in the Pentateuch.
I certainly agree that textual criticism is a necessary step in the analysis of any and all biblical writings. That said, one of the advantages of the narrative-based analysis of the Pentateuch that I am describing here is that minute differences in the text, of the type that commonly creep in during transmission, do not really affect the broad historical claims of the narratives, especially as these claims are evident across multiple texts and are deeply interconnected with other such claims, forming a web that remains strong even if one small link is cut.
The Pentateuch is, of course, one of the least text-critically problematic portions of the Hebrew Bible. But even where there are significant text-critical issues, such as in the Tabernacle chapters in Exodus 25-31, 35-40, the text-critical analysis doesn't affect the source-critical analysis.
As for the scribal model, I should say that, at least from my perspective, no model is "assumed" - when the Pentateuch is analyzed as I am describing, the four documents emerge, and their very presence in the Pentateuch, in the manner that they are present, requires at the minimum one person to have combined them. That is to say, the scribal model develops from the literary evidence, rather than the other way around, as I think is indeed proper.
The examples of scribal transmission and textual growth in other ANE, biblical, and post-biblical books are undoubtedly of great interest, but I do not think it proper to import those models into the Pentateuch, any more than it would be to import the pentateuchal model into other texts. The model of transmission and growth must be derived from the literary evidence on a case-by-case basis.
That said, I think that there may be at least one reasonable parallel to the model of pentateuchal composition that I am describing here: the Samaritan Pentateuch. As is well known, Sam will occasionally conflate multiple versions of a single event that is narrated differently in different places; for example, in Exod 18:13-27, Sam takes the MT story of Moses appointing judges in Exod 18:13-27 and combines it with Deuteronomy's version of the same, from Deut 1:9-18. The scribal impulse here is clear enough: these two descriptions of the same event should be brought together into a single account. Yet the newly-created version in Sam has narrative difficulties that result, understandably enough, from its conflation of two distinct versions of a single story.
The same phenomenon, it should be observed, occurs also in the DSS and elsewhere in post-biblical literature. For details, I recommend the article by Jeffrey Tigay, "Conflation as a Redactional Technique," in idem, Empirical Models for Biblical Criticism (Philadelphia: University of Pennsylvania Press, 1985), 53-95.
Now to your question about mutual influence between sources: quite simply, I think that there is influence when the literary evidence demands it - when there is clear dependence or citation. Thus I think it indisputable that D was heavily influenced by E, and to a lesser extent by J. I do not see any evidence at all that either J or E knew the other, or that P knew any other source or that any other source knew P. (This is slightly complicated by what seems to be a good case that H, a later stratum of P, was indeed familiar with both D and E, but here we may be getting a little ahead of ourselves.) In any case, the mutual relationship of the sources, like their processes of internal growth, dating, social location, etc., is a question that can be answered only after the sources have been isolated and identified. Fundamentally, if it were somehow to be proven that P knew every other source, or vice versa, it shouldn't affect the division of the sources in the text, as that division emerges from internal literary analysis.
My thanks to you and to everyone I just responded to seriatim for interesting and engaging questions, and especially for the unusually and very welcome civil and respectful manner of discussion - may it always be thus!
Thanks for these helpful replies.
As a follow-up to the SamP example above. Does the resulting conflation leave enough of the original material from EACH story in tact so as to preserve the narrative coherence/semblence and distinctive profile of each earlier source in the proto-MT (esp. if we didn't have those texts)? If so, then I feel that the analogy holds for what you describe going on in the Pentateuch. But if not, then on analogy, how can we use narrative coherence as a criterion for determining earlier sources whose original coherence was lost in a process of conflation?
Word for word, simply interwoven at the most appropriate chronological point, resulting in a text that both preserves the distinctive profile of its sources and, though chronologically plausible, is internally contradictory on a number of substantive narrative grounds.
Do we face 'non-documentarism' as soon as anyone says that, though there may have been multiple sources for the text we have, there were attempts to harmonise them before they were finally compiled?
Are these hypotheses falsifiable? What, if anything, would qualify as showing that the earlier documents did not exist?
Use the form below to submit a new comment. Comments are moderated
and logged, and may be edited. You must provide your full name.
Inappropriate material will not be posted.