August 28th, 2007
My Trouble with the Documentary Hypothesis, Part 1
In these fleeting last days before the fall semester starts, I have begun working through Richard Elliott Friedman’s book The Bible with Sources Revealed, which visually shows the results of the Documentary Hypothesis by color-coding all the text of the Pentateuch according to how Friedman attributes each passage to the underlying sources. To put it briefly, the DH says that the first five books of the Bible (Genesis, Exodus, Leviticus, Numbers, Deuteronomy), or the Pentateuch, traditionally attributed to Moses, are actually composite works edited together from a variety of source documents. The source documents are identified by the letters J, E, D, and P, for Jahwist, Elohist, Deuteronomist, and Priestly Source, respectively, according to a distinctive characteristic of each source.
According to Friedman, there are seven main lines of evidence in support of the DH:
- Linguistic: the hypothetical sources seem to come from different periods in the development of the Hebrew language: early, middle, and late.
- Terminology: the sources are distinguished by their choice of words. For example, 14 out of 15 occurrences of “plague” are said to be in the P source, whereas only the D source uses the phrase “with all your heart and with all your soul.”
- Consistent Content: This is broken down into several sub-categories. I will give one or two examples from each category.
- Revelation of God’s Name: According to J, humans knew God as YHWH from very early (Genesis 4:26b), whereas P seems to claim that the name YHWH was first revealed to Moses (Exodus 6:2-3)
- Sacred Objects: The Tabernacle is a special concern of P, but not of J or D. Only P and J mention cherubs, E and D never do.
- Priestly Leadership: In P, only the sons of Aaron are priests, whereas in D, all the Levites are priests.
- Numbers: P reports dates and numbers precisely; the other sources do not.
- Continuity of Texts (Narrative Flow): According to Friedman, when the sources are distinguished from each other, they each result in a highly continuous narrative in their own right. This is critical since each of the sources is supposed to have been an independently existing text before they were edited together.
- Connections with Other Parts of the Bible: D exhibits parallel wording with passages in Jeremiah; P parallels Ezekiel; J and E have connections with Hosea.
- Relationships Among the Sources to Each Other and to History: J is hypothesized to have originated during the time of the divided kingdom (ca. 10th-8th centuries BCE) in Judah, the southern kingdom. E was written approximately during the same time in Israel, the northern kingdom. That makes it noteworthy that in J, Abraham lives in Hebron, which was the capital of Judah. Further, the ark was located in Judah, not Israel, and correspondingly, we see the ark emphasized in J but not in E.
- Convergence: This, Friedman believes, is the strongest evidence for the DH: that several of these lines of evidence converge in particular passages. When you separate out the sources, you find distinctive terminology, different stages in the historical development of the Hebrew language, and continuous narratives all together.
The DH is certainly plausible. The tradition that ascribes the whole Pentateuch to Moses is a very old one, but it’s only a tradition, and even if it is the case that a substantial portion of the books were written by Moses himself (or a single other person), we can’t rule out the possibility that bits of other material were added later. There are apparent discrepancies in the text that would be explained if they come from different sources. If the evidence is as clear as he says it is, especially if by attributing different bits of text to different sources all kinds of other distinctive features come to light (convergence of evidence), the DH ought to be accepted by all.
As I say, I’m only beginning to work through this material, but I’m continually bothered by what look to be ad hoc ascriptions of a text to J or to P not because there’s anything in the text that would indicate such a source but because it’s necessary to the hypothesis. Let me give an example from the doublet telling of the separation of Abraham and Lot in Genesis 13. Doublets are double-tellings of an incident and are especially important in this kind of historical inquiry because they’re prime candidates for a text that was spliced together from two different original sources, both of which had versions of the one event. Here’s the text in question:
(5) And Lot, who was going with Abram, also had a flock and oxen and tents. (6) And the land did not suffice them to live together, because their property was great, and they were not able to live together. (7) And there was a quarrel between those who herded Abram’s livestock and those who herded Lot’s livestock. And the Canaanite and the Perizzite lived in the land then. (8) And Abram said to Lot, “Let there be no quarreling between me and you and between my herders and your herders, because we’re brothers. (9) Isn’t the whole land before you? Separate from me: if left then I’ll go right, and if right then I’ll go left.” (10) And Lot raised his eyes and saw all the plain of the Jordan, that all of it was well-watered (before YHWH’s destroying Sodom and Gomorrah) like YHWH’s garden, like the land of Egypt, as you come to Zoar. (11) And Lot chose all the plain of the Jordan for himself, and Lot traveled east. And they separated, each from his brother. (12) Abram lived in the land of Canaan, and Lot lived in the cities of the plain. And he tented as far as Sodom. (Genesis 13:5-12, Friedman’s translation)
According to Friedman, this is a doublet, comprised of the J and the P source of Abraham and Lot’s separation. All of it is J except for vv. 6 and 11b-12a:
And the land did not suffice them to live together, because their property was great, and they were not able to live together (verse 6). And they separated, each from his brother. Abram lived in the land of Canaan, and Lot lived in the cities of the plain (verses 11b-12a).
Is this actually a doublet? Do the two verses that he claims are from the P source actually retell the rest of the incident or are they just summary statements? It’s not easy to tell. Go back and read the whole passage again and see if it looks like two tellings of the same incident from different sources.
The P source, as Friedman has identified it, does show narrative unity. The immediately preceding text in P is found in Genesis 12:4b-5:
And Abram was seventy-five years old when he went out. And Abram took Sarai, his wife, and Lot, his brother’s son, and all their property that they had accumulated and the persons whom they had gotten, and they went out to go to the land of Canaan. And they came to the land of Canaan. (Friedman’s translation)
Which leads straight into the abbreviated account of Abraham and Lot’s separation we just saw:
And the land did not suffice them to live together, because their property was great, and they were not able to live together. And they separated, each from his brother. Abram lived in the land of Canaan, and Lot lived in the cities of the plain. (Genesis 13:6, 11b-12a, Friedman)
If the alleged P bits in Genesis 13 are really only summary statements from an original unity, then it should come as no surprise that there’s narrative continuity in P. If every account has both a summary and the details, and you separate out all the summaries and then join them together, of course you will have a continuous narrative. So in this case, narrative continuity is actually not evidence for the DH against the traditional view that Genesis 13 comes from a single author.
Even if we conclude that this is a doublet, why should we attribute verses 6 and 11b-12a to P and not some other source? The rest of the account would clearly have to be attributed to J, since it says that “Abram invoked the name YHWH there” (13:4b), but why suppose verses 6 and 11b-12a are P and not E or some other source? The only distinctively P trait is the word “property” in verse 6, but how much evidentiary value does that have? The conclusion that only P uses the word “property” can only be reached after the text units have been separated in their sources, so it seems that there is no reason to choose P over E, except the desire to preserve narrative continuity. Since P is believed to leave off with an account of Abraham’s journey to Canaan (in 12:4b-5, above), attributing the doublet in chapter 13 to P, rather than something else, preserves the continuity of P. That seems reasonable, but at the same time it also lacks evidentiary weight. There’s no reason to multiply entities by postulating a different source here, but nevertheless, attributing the doublet in chapter 13 to P is simply performed ex hypothesi: the hypothesis says that these independent texts should be continuous, so Friedman assigns the doublets in such a way as to ensure that they are. Or so it seems. I’ve noticed a lot of other source attributions that seem to be carried out ex hypothesi.
What’s more, separating out the text of Genesis 13 in this way creates new problems. The text of P is supposed to run thus:
(12:5) And Abram took Sarai, his wife, and Lot … all their property … and the persons whom they had gotten … And they came to the land of Canaan. (13:6) And the land did not suffice them to live together …
But to whom does “them” refer to in 13:6? From the context of chapter 13, only Abraham and Lot are in view: “And Lot, who was going with Abram, also had a flock and oxen and tents” (13:5, J source), but if we accept Friedman’s hypothetical reconstruction of P, the “them” seems to refer to all the people Abraham took with him to Canaan.
And another thing: Genesis 13:11b (supposedly P source) says, “And they separated, each from his brother,” clearly referring back to 13:8b, which reads, “‘Let there be no quarreling between me and you and between my herders and your herders, because we’re brothers.’” The problem is that 13:8b is ascribed to the J source! It looks to me like 13:6 and 11b-12a are summary statements from a single author, not units from a different source.
This isn’t really a big deal, though. It certainly isn’t fatal for the DH. Even if we say that all of Genesis 13 comes from a single source, there is still narrative continuity in the hypothetical P source, which next appears in Genesis 16:3: “And Sarai, Abram’s wife, took Hagar, the Egyptian, her maid, at the end of ten years of Abram’s living in the land of Canaan …” My point is only that a lot of the attribution decisions seem to be calculated to support the hypothesis. In other words, they’re deductive when what we really need is something inductive, something in the text itself that doesn’t just work with the hypothesis but actually has explanatory value. Words and phrases of a general nature like “property” won’t do it, and I’m not seeing a lot of other kinds of evidence converging here either.
My impression at this stage of my study is that the DH is strong in a few passages (and perhaps even true there!) and most of the rest is ad hoc and calculated to support the hypothesis. I’ll keep working on it as I have time. I want to give it a fair shake.
No related posts.