Another look at the Documentary Hypothesis. An alternative proposal.

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

by Neil Godfrey

Were the first books of the “Old Testament” composed by “redactors” piecing together stories from different sources (that has long been the predominant view) or is it possible that they were composed by a single author or “school of authors”? The former is the documentary hypothesis. For a background on the documentary hypothesis refer to the post Who wrote the Bible? Rise of the Documentary Hypothesis. The topic has resurfaced with some recent posts (Plato and the Bible; Genesis to Kings, Authorship) that question the validity of the DH, proposing a “unitary authorship” of the Pentateuch and more. I am intrigued by the new alternative but have not yet sold my soul to it. I am not always overly enthusiastic about some of the arguments for a single authorship. So consider these posts as exploratory and informative. (But that’s what most of my posts are here, anyway.

Here I cite but one scholar’s criticism of the DH and proposal for the Flood Story in Genesis being composed from scratch as a single narrative. That is, the narrative is not a patched quilt of priestly and Yahwist sources after all. Here is Thomas Brodie’s take on the story (quoted words are in dark azure; the rest is my summary and comment.)

The deluge account contains much repetition and variation, and so some researchers have suggested that it is composed from two sources — J and P . . . .

(Brodie, 181)

Some of the examples of the repetition and variation:

  • Sometimes Yhwh, other times Elohim, are used for God. The sections with Yhwh have been attributed to the J source, Elohim sections to the P source.
  • At one time we read that two of each kind of animal was brought into the ark; later we read there were seven pairs of clean animals but only two of each unclean. The former has been attributed to the P source; the latter to J.
  • We first learn of forty days and nights of rain; later of the flood cresting after 150 days. J, then P.
  • The command for Noah to enter the ark is duplicated: 6:18–20 (P) and 7:1–3 (J).
  • Twice we read of Noah and his family entering the ark: 7:7 (J) and 7:13 (P).
  • Twice we read of God’s promise to never again destroy the earth: 8:21 (J) and 9:11 (P)

Brodie’s faults the documentary hypothesis as being based on contradictory arguments:

The final product is not an unimaginative collection of material drawn from distinct sources, but an artful unified composition arranged chiastically around the central affirmation in 8:1 that “God remembered Noah.”

This explanation, however, contains radical problems, problems that are both general and specific.

In general: the explanation is not coherent. It implies two opposite procedures — mechanical juxtapositioning (regardless of overlap or divergence) and artistic unifying. One procedure is slavish, the other imaginative. Behind these procedures are two opposite attitudes — scrupulosity and freedom. The contradiction in the explanation is far deeper than the contradictions in the text.

More specifically, the theory is not supported by the details. There are several problems (see esp. Ska, 1994; 1996, 259–260):

1. The purported J story is seriously incomplete. It contains no account of making the ark or leaving it. Such lack of completeness is not explained by a desire to avoid repetition. The author had no problem with repetition as such; some minor details occur twice.

2. The so-called “double” entry to the ark (7:7–9, 13–16) does not require two sources. The doubled text is part of a single coherent repetitive style — similar to the repetitiveness of Genesis 17 (17:23–27). Essentially the same is true of other so-called doublets: they are part of a coherent repetitive style. Repetition is a basic feature of narrative, especially of biblical narrative (Alter, 1981, 88–113). Repetition results from various techniques (Niccacci, 1994). The issue then is not whether there is repetition but whether it is possible to discern the repetition’s variation or purpose. The two commands about entering the ark (6:18–20; 7:1–3), for instance, have several variations of context and content, but they are sufficiently similar to build something important: the sense — amid a collapsing world — of momentum and continuity. Nor do the two diverse types of bird (the raven and the dove, 8:6–12) mean two sources. In Tablet XI of the Epic of Gilgamesh, the Noah-like Utnapishtim sends out three diverse birds—a dove, a swallow, and a raven (Brichto, 1998, 114) — but that does not mean three sources.

3. The language of some purported J material, especially as regards order and sacrifice, would normally be reckoned as late or priestly.

For these and other reasons, scholars such as Blenkinsopp (1992, 77–78) and Ska (1996, 259) have moved to the idea of a single (priestly) account, which was later retouched. In other words, rather than dividing the text fairly evenly in two, they attribute most of it to a single author and reserve just a small percentage to an editor.

So some scholars are finding the different pieces fit better as if the story was composed as a single unit from the start. Brodie lists other indicators of what he believes is a unitary narrative:

1. Chiasm. Kselman speaks of the flood story as artful, and highlights one artistic feature — the chiastic arrangement (91).

2. The inextricability of J and P. There is significant continuity of the flood story with diverse aspects of Genesis 1–5, especially with Genesis 1 (see above, “Relationship to Preceding Chapters”), and this continuity strengthens the idea that J and P are inextricable. For instance, within 8:1–9:17 continuity with Genesis 1 (attributed to P) is not limited to so-called P material; it includes texts usually attributed to J — for instance, the opening of the window, the appearance of a fresh plant, the emergence of the dry ground, and especially the promise of the seasons (8:21–22, a variation on the arrangement of the seasons in 1:14–18). Thus, the alleged J material cannot be separated from dependence on P.

And alleged P material cannot be separated from dependence on J. For instance, the emphasis on violence and not shedding blood (6:13; 9:5–6, two P texts) fits best not with anything otherwise attributed to P but with material attributed to J — the violence of Cain and Lamech (chap. 4). This confirms other data of a similar kind: the alleged P material of chapter 5 builds on alleged J material in Genesis 2–4 (see Chapter 18, “Chapter 5 as a Conclusion to all of Genesis 1–5”).

The effort to disentangle two hypothetical sources sometimes reaches incredible proportions. For instance, “in a pericope assigned mostly to P [7:8–9] but with inexplicable intrusions from J, we are asked to believe that an intrusion from P is present in the J intrusion” (Brichto, 1998, 134).

At times both divine names are so embedded in a single seam of narrative that they belong together. For instance, as the commentary on 6:9–7:16 indicated, the various concluding statements about Noah’s response to God’s word (6:22; 7:4, 10, 16) contain a precise diminution:

“Noah did this; according to all that God commanded him, so he did.”
“Noah did according to all that Yhwh commanded him.”
“ . . .as God had commanded Noah.”
“ . . .as God had commanded him.”

Within this carefully-wrought picture of diminution—itself part of a larger pattern depicting the fading of God’s word — both divine names are integral.

3. The increasing intelligibility of using two divine names. As a general principle, the use of two names does not require two diverse sources. In the David story, for instance, the variation of name — “David” and “King David” (e.g., 2 Sam. 7:17–18) — does not mean two sources.

On the variation in the references to God Brodie cites an observation by Brichto that there appears to be logical sense to the variants:

In Gen. 1:1–4:16 the gradual change—God, Yhwh God, Yhwh—was seen to accord with an increasing involvement of God in human problems. The first mixing of divine names (4:25–26) occurred in the panel that first mixed literary forms (4:17–26). And in introducing the deity into the deluge story (6:5–8), the use of “Yhwh” rather then “God” makes sense: “Yhwh” helps to distance the deity from “the sons of God” in 6:1–4 (Brichto, 1998, 140).

It is possible that something similar applies in the repeated references to God’s commands: God commanded, Yhwh commanded, God commanded, God commanded (6:22; 7:4, 9, 16). The variation between God and Yhwh may perhaps be linked partly to structure: obedience to the commanding role of God forms a conclusion to the three main sections of panel one (6:22; 7:9, 16). Obedience to the commanding role of Yhwh, however, concludes a subdivision (7:5).

On the question of repetition combined with variation Brodie finds a coherence that suggests a unitary composition:

The phenomenon of repetition and variation is understandable not only in light of later chapters (e.g., Genesis 17), but especially in light of the chapters that precede; the many repetitions and variations in the deluge story are a more complex form of the repetitions and variations that occur in Genesis 1–5.

Genesis 1 is laden with repetition and variation. The days follow one another with a smooth rhythm, yet there are variations: the omission of “it was good” on the second day; the fifth day’s failure to increase in quantity; the subdivision of days 3 and 6 into two parts. The deluge story has more of the same, but without the serenity. Genesis 2–3 brings further and greater kinds of variation and repetition. The genealogies (Gen. 4:17–chap. 5), by mixing both literary forms and divine names, prepare the way for the more complex mixing that occurs in the deluge story.

Brodie further argues for a purposefulness in the contradictions in the narrative 

The three “virtual” contradictions:

  • God’s attitude to evil;
  • Number of animals (2 or 2X7);
  • Number of days (40 or 150)

. . . . but these apparent contradictions, far from breaking up the story, bring out its meaning: to make peace with this troubled world, to have the capacity for forgiveness and reconciliation, it is necessary for God, and implicitly for humans, to rise above superficial calculation, superficial counting.

In conclusion, the idea of the unity of the flood story is not new. Already in 1978 Wenham (347–348) was clear: “The syntax, literary structure, chronology and Mesopotamian parallels all point to the unity and coherence. . . of Genesis vi–ix.” As time passes Wenham’s essential thesis is increasingly justified.

I have to say that I do not find Brodie’s rationale here convincing. To say that the contradictions in numbers is a sign we are to not care about numbers but only care about God’s plan sounds like theology more than serious literary or textual analysis. I would like to check out other views such as those of Wenham.

The theory of two combined sources lacks internal coherence; it contains contradiction. Nor does the theory correspond to the external data. The story’s structure is chiastic and unified. Despite the text’s openness to repetition, the purported J source is seriously incomplete. The criteria for distinguishing the sources are inconsistent. Materials allegedly belonging to distinct sources are inextricably bound into a unity. The text’s pattern of repetition and variation is part of a larger coherent phenomenon that is found elsewhere in other forms (for instance, in Gen. 1 and 17). The variation in divine names is becoming increasingly intelligible. And even the most puzzling features, the apparent contradictions, fit well with the basic story: forgiving humankind means letting go of some normal calculations.

The theory of artistic unity is initially difficult — it means entering into a world of art — but, piece by piece, it begins to make sense. The elaboration of the artistry is not complete, but with each added detail of understanding there is confirmation that this is the direction in which to search for the coherence and meaning of the text.

Brodie, T. L. (2001). Genesis As Dialogue: A Literary, Historical, and Theological Commentary. Oxford England ; New York: Oxford University Press.


The following two tabs change content below.

Neil Godfrey

Neil is the author of this post. To read more about Neil, see our About page.


  • john dauria
    2018-11-25 14:17:52 UTC - 14:17 | Permalink

    poor ol Brodie seems to be going in for a bit of apologism hisself:)

  • Pingback: Another look at the Documentary Hypothesis. An alternative proposal. — Vridar | James' Ramblings

  • Austendw
    2018-11-26 15:59:58 UTC - 15:59 | Permalink

    I’m not sure why a chiasm should be thought of as proof of a single author. Why couldn’t a redactor add or combine material either by organize his material in a new chiastic structure or fitting it into an existing chiastic structure?

    Textual criticism has moved on quite a bit since Wellhausen, and it is quite possible to deviate from his or Friedmann’s analysis without having to jump to a single-author theory. I agree with Brodie when he says; “Nor do the two diverse types of bird (the raven and the dove, 8:6–12) mean two sources.” In that particular case, I don’t think that any of the birds are part of the P-version.

    But equally, it is hard to deny that Gen 6:19-20 and Gen 7:2 give completely contradictory statements on the number of pairs of animals on the ark. Moreover this distinction isn’t only a question of content (which is the impression one sometimes gets from critics of Documentary/Supplementary theories): the Hebrew words for male/female differ in the two passages, supporting the notion that they come from distinct accounts.

  • 2018-11-26 20:35:58 UTC - 20:35 | Permalink

    The J story of the Noah has Yahweh tell Noah to build an ark and then it says “Noah did according to all that Yahweh commanded him” in 7:5. That’s exactly what happens in P as well, Elohim tell shim to build an ark and then “Noah did it” in 6:22. It’s true that J doesn’t say that Noah left the ark, but it does say that he pulled back the covering to see that the land was dry.

    Using Richard Friedman’s JEPD model, in J, Isaac prophecizes that Esau would break Jacob’s “yoke” from his neck, which would seem to indicate it was written after the Edomites won independence from Judah in 848 (Gen. 27.40) and that Simeon and Levi being dispersed but none of the other tribes, which he takes to mean the other tribes must have still existed, putting the source before the fall of Israel to the Assyrians in 722 (Gen. 49.5). Presumably, a prophecy like that would be made soon after it happened so that it would be still relevant to its first readers. There are a lot of parallels between Solomon and king Jehoram, including both ruling 40 years and their mother and high priest having identical or near-identical names. King Jehoash (r. 836-796) is said to have “repaired” Solomon’s Temple but the description provided sounds a lot more like he was building it from scratch. Interestingly enough, Friendman’s interpretation of J ends with Solomon, and many of the early proponents thought J was written during Solomon’s time. Given the many parallels, Jeroboam I, likewise, appears to be a mythical creation based on Jeroboam II.

    I would describe myself as a New Testament minimalist but the problems I see with Hebrew Bible minimalism is that these links I mention here are hard to justify as coming from a later time because at what time did anyone really ever care about King Jehoash or Jeroboam II? If the minimalists could show links like these to the “current” political dramas of the Ptolemaic and Seleucid, then that would certainly help, but I would still like to the kind of linguistic confirmations that Friendman’s work has gotten from Hebrew language scholars.


    • Austendw
      2018-11-27 06:42:52 UTC - 06:42 | Permalink

      In the traditional source criticism, the J-version doesn’t contain Yahweh’s instruction to Noah to build an ark. J is said to consist of 6:5-8, and then to resume at 7:1 when Yahweh tells him to go into the ark that he has already built. But there appears to be no J-version instruction to build it.

      However, I have a suspicion that there actually is a fragment of J’s version of Yahweh’s instruction to Noah. Gen 6:14aβ-b comes before what looks like the introduction to the details of the construction in 6:15 (“And this is how you shall build it”) suggesting that this intrusive passage has come from a different account. If we repoint the Masoretic qinnum (rooms/nests) in verse 6 as qannim (reeds), it is possible that this passage is in fact the J-version – a reed ark caulked with tar – that has been combined with the P-version – a three-storey gopher-wood ark.

      I’m not sure if anyone else has suggested the possibility that this short passage is from J.

  • Scot A. Griffin
    2018-11-26 21:08:59 UTC - 21:08 | Permalink

    The two ideas, (1) that the Pentateuch includes prior materials that were edited together and (2) that the Pentateuch was produced as a unitary work, are not mutually exclusive. I’d argue that it is quite probable (even highly likely) that the Pentateuch includes portions of prior works that were edited into a larger, newer unitary narrative.

    The Documentary Hypothesis relies too much on the belief that doublets are indicia of different authorship, which is based on the assumption that the Bible is a history as we understand that term in modern use.

  • Leave a Reply

    Your email address will not be published. Required fields are marked *

    This site uses Akismet to reduce spam. Learn how your comment data is processed.