Chiasms, Multiple Sources, Armstrong, and Chasms in Rationality

Chiasms, Multiple Sources, Armstrong, and Chasms in Rationality July 7, 2021

Welcome to another episode of “Brandolini’s Law Exemplified”, your friendly programming that has to put in all the effort where others don’t.

I wrote an article on the issues with the Genesis flood account that did three things: 1) it showed how it was impossible; 2) it is drawn from earlier mythological stories; 3) there are multiple sources – it was not written by Moses.

Dave Armstrong replied:

I have issued yet another refutation of Jonathan’s arguments:

Pearce’s Potshots #39: Ignoring Chiastic Literary Genre in Genesis

This is a continuation of my previous article on the same topic:

Pearce’s Potshots #38: Chiasmus & “Redundancy” in Flood Stories (Also, a Summary Statement on Catholics and the Documentary Hypothesis)

Now, I haven’t read it or them yet. I just want to write this to see if I pre-empt anything, to see whether Armstrong is predictable and whether he has thought everything through. I’d be interested if you people can evaluate for me as I am unlikely to be able to read them for a few days. My theory is that Armstrong says something like “Oh, look, a new theory to me that gets me out of a hole. I’ll take it, and run with it hook, line and sinker.” Thing is, it really doesn’t get him out of the hole. Look, I’m offering my hand to give him a lift up, but Dave’s a stubborn man and refuses the help. He’d rather sink than accept my help, rather surround himself in darkness than see the light.

A chiasmus:

In rhetoricchiasmus (/kˈæzməs/kahy-AZ-muhs) or, less commonly, chiasm (Latin term from Greek χίασμα, “crossing”, from the Greekχιάζωchiázō, “to shape like the letter Χ“), is a “reversal of grammatical structures in successive phrases or clauses – but no repetition of words”.[1]

A similar device, antimetabole, also involves a reversal of grammatical structures in successive phrases or clauses, but unlike chiasmus, presents a repetition of words in an A-B-B-A configuration.[2]

For certain biblical scholars, there is the belief that the Bible has heaps of chiastic structures that do repeat words as well as concepts. Gordon Wenham is a big proponent, who sees the flood narrative written in this way:

Armstrong’s methodological and epistemological approach

There are a number of problems with Armstrong’s position, as summed up in this comment (with the ill phrase added later when I said I’d been in bed for two days):

This is odd because I literally disproved a local flood account that should mean that either 1) it was a global flood or 2) it was a local flood and the claims of Genesis were incorrect.

You literally can’t have a claim that waters were above 5,100m high – Mt Ararat, and it was local. I: mean, he knows how water works, right? Finding its own level and running to the lowest point. Unless the whole Ancient Near East is a 5,100m bowl, I don’t get it.

I also have to say this. It annoys the hell out of me.

  1. I sometimes get thousands of comments a day. I’m truly sorry you don’t. You may get to read all yours. I don’t. I literally can’t. I miss a heck of a lot of them.
  2. I do not exist in order to respond to your articles papers articles posts. I am insanely busy, and things like this take me way too much time than they are probably worth. I have also been ill in bed for two days and am entering a third. This has been tough to write when I can barely open my eyes through the streaming. I am trying to write between 5 and 10 pages of my book a day, blog, read an inordinate amount for research purposes, work (as in a job), be a parent to two insanely demanding twin boys, partner my other half, do regular living things. (I am also disabled with primary progressive multiple sclerosis, and but I don’t really want to pull that card!) That I give you as much time as I do is a naturalistic miracle.
  3. In other words, blogging is something I do on the side. It is a labour of love. I do what I want when I want and am not beholden to you.
  4. You might consider that if I do ignore you, there might be good reasons for doing so.
  5. I will never ignore you if you produced something that, if I read, really challenged me and made me have to genuinely rethink stuff. That is exactly the sort of stuff I would get right on to dealing with. It’s how my brain works.

That aside, I am most interested in the claim that he has dealt with the Documentary Hypothesis qua multiple sources by proposing an alternative explanation for massive repetition in the Pentateuch due to “a thing called chiasmus, that I just discovered a few days ago”.

The huge humdinger of a clanger here is that 1) he admits huge repetition and 2) this is solved by chiasms but yet 3) he only discovered this a few days ago.

Does this mean that the “massive repetition” in the OT that he used to explain by sticking plasters on every single one, weighing down the Bible with a plethora of ad hoc rationalised bandaids is now explained by something else? That every single one of those rationalisations he previously did he now admits was wrong? That the certainty he was administering his blog posts with, that bolstered each and every rationalisation, was entirely misplaced?

I will wait for him to come out and admit this.

Why should we pay any attention to his certainty and rectitude right now if only last week, he had the same level of claimed epistemological rectitude and certainty and it was wrong?

He jumps from one possible defence to another like a rat on speed from a sinking ship.

The chiasmus – a lifesaver?

Let’s look at his new toy: chiasms.

For the biblical scholar like Gordon Wenham, who adheres to the chiastic structures above, the chiasms are great because he thinks, they explain repetition and allow you to maintain a single author.

Except repetition is only one of four problems that the Documentary or Supplementary Hypotheses (DH, SH) solve. These are: contradictions, repetition (redundancy), discontinuity, and stylistic and terminological divergences.

I’m sure Armstrong has considered all of this, but the first problem is that chiasms only solve one of these issues – repetition. They do nothing to solve the other three. Take the proposals for chiasms in the Joseph story, that are famously proposed by J.P Fokkelman (fun fact: I just quoted him in my book on the Exodus). As Joel Baden observes in The Composition of the Pentateuch (p. 10):

…Fokkelman does not address, nor even recognize, the substantive contradiction of the Midianites and Ishmaelites. Rather, he privileges the discernment of formal structure over the narrative coherence of the passage. In short, he does not take into account, either positively or negatively, the main textual difficulty.

Which is to say that it fails to account for or solve the problems that litter the Pentateuch. In this case, the famous and irreconcilable problem for literalists that Joseph is sold to the Midianites and a few verses later they have magically turned into the Ishmaelites. Of course, this is both solved and made complete sense of by the DH or SH.

This is the same for the flood narrative. In his papers “An Examination of Some Attempts to Defend the Unity of the Flood Narrative in Genesis” and “An Examination of Some Attempts to Defend the Unity of the Flood Narrative in Genesis: Part II“, JA Emerton shows that the theorising of Wenham and others fails because chiasms do nothing to solve the contradictions in, say, both the number of animals and the chronology. These are only solved by DH/SH. In part I, page 402-405, Emerton details the contradictions, and also details a lexical analysis of the 150 days vs 40 days contradiction that Armstrong so vehemently denies. Armstrong should read that section to clear up his own “abating” issues. Perhaps I’ll write a piece on that, too.

Emerton looks at several structural theories – those of Umberto Cassuto, FI Andersen, YT Radday, and Gordon Wenham. [e.g., “Cassuto’s attempt to establish that there is an “architectonic structure” in the story of the flood (and even his delimitation of the section) has thus failed. It cannot be used against an analysis of the narrative into sources.” p. 408] The funny thing is always that such approaches are mutually exclusive. Three of these Christians are dead wrong. Anywho, since we are talking chiasms, I’ll jump to Wenham.

I will go to the criticisms of Wenham because he is perhaps the most famous proponent. Interestingly, if Armstrong relies on Wenham, Wenham does not exclude the possibility of multiple sources, though prefers “a simpler and more economical hypothesis” to think “of one epic source which has been reworked by a later priestly editor” ([1] p. 348). Notice here that Wenham advocates two sources: some non-P source and a P redactor.

Oh dear.

Emerton looks at Wenham’s chiastic structure (more detailed than the one above) and picks holes in it straight away. I don’t know if Armstrong at all relies on Wenham, but if he does, well…:

Before viii 1 are fifteen items, A-O, which are balanced by fifteen items in the reverse order after viii 1, O -A . He admits that up “to a point it is not surprising to find one [sc. a palistrophe] in the flood story” (pp. 337-8), since some events happen in the reverse order after the flood (e.g. entering and leaving the ark, and the rise and fall of the flood waters), but argues that there is an element of contrivance in some of the items. Thus, the events recorded in F -A (ix 9-19) do not naturally correspond in order to A-F (vi 10-21); items H and I refer to the same period of seven days, whereas H and I refer to two different periods, and the forty days of L (vii 17) are part of the hundred and fifty days of vii 24. This contrivance shows that an attempt has been made to arrange items in a special order even when the counterparts do not naturally match. Further, he admits that there are some things in the narrative that have no place in the palistrophic pattern, such as Noah’s sacrifice in viii 20 ff. (p. 340). Such items are explicable because the author was dependent on traditions which he did not feel free to change or omit, and because he wanted variety, for “Absolute repetitiveness is dull” (p. 340). Thus, Wenham claims repetition as evidence for his theory, but does not recognize lack of repetition as evidence against it. Is he trying to have it both ways? (Part II, p. 7)

It doesn’t get better for Wenham, because his theory appears to largely accommodate the DH.

Before we examine Wenham’s arguments in detail, it may be noted that, even at their face value, some of the elements in his pattern are compatible with an analysis into J and P sources, as he perhaps recognizes in his remarks on pp. 347-8 (see above). In addition to the fact (which he also recognizes) that the order of some events is natural, it may be observed that some corresponding pairs in the palistrophe come from the same source. Thus, K, L and part of M and their counterparts are found in J; and D, E, F and parts of H and M and their counterparts belong to P.

But I’m sure Armstrong has looked into this.

I think it will be most efficient for me to critique chiasms and Wenham, with regard to the flood narrative, in a bullet-point list. I really hope Armstrong has used Wenham, otherwise this is a very time-costing detour! (D’oh?) All criticisms refer to the palistrophe listed on pages 6-7  and pictured above:

  1. The structure is supposed to be symmetrical, yet each chiasm can vary hugely in length from a short sentence to several verses.
  2. The choice of things in the list/structure appear completely arbitrary (p. 8). This, to me, is one of the biggest issues and utterly smacks of post hoc rationalisation and contrivance and is pretty terminal for me (p. 8):
    Noah is mentioned many times in the narrative. How can we be sure that the mention of his name in A (vi lOa) and A (ix 19) is significant, but not that in, for example, ix 18 between D and C ? Wenham might reply that at least Noah is mentioned at the beginning and end of the unit he is discussing, and the fact is not altered by the observation that the mention in such positions is not unnatural or surprising. More seriously, why is the mention of Ham, Shem and Japheth thought significant in vi lOb (B) and ix 18b (B), but not in vii 13, the only other place in the alleged palistrophe where the names appear? Similarly, the ark is mentioned many times in the story, but why are only the references in vi 14-16 (C) and ix 18a (C) significant
  3. Some of the items in the palistrophe are “questionable” (p. 8):
    …the pairing of J and J is a mistake. It may also be doubted how closely K, vii 16 in which God shuts Noah in the ark, corresponds to K , viii 6b in which Noah, not God, opens a window in the ark.
  4. Emerton details how Wenham’s items are sometimes not distinct from each other, but that Wenham recognises this and bizarrely tries to have his cake and eat it by saying this was intentional. (See p. 9)
  5. Another issue Wenham himself recognises is that some major parts of the story simply don’t feature in the chiastic structure (p. 9):
    If, however, that writer was as skilful and ingenious, and so concerned to compose a palistrophe, as Wenham believes, it is surprising that he did not find a place for some other important events in his list of significant points in the story. The force of the argument is not turned by asserting that, “if he had achieved total and perfect symmetry, the story might have lost some of its interest” (p. 340). The fact that some important items are not fitted into Wenham’s pattern cannot but cast doubt on his theory. [See the rest of this paragraph for much more analysis and examples of bizarre omissions.]
  6. Lastly, p. 10:
    But there are some corresponding passages that do not have a place in the alleged pattern, and their existence casts doubt on whether a palistrophe was intended. We have already seen that Wenham’s failure to find a place in the palistrophe for the entry into the ark in vii 7-9 and the departure from it in viii 18-19 causes a difficulty for his pattern. It also leads one to ask why the author did not fit these corresponding verses into his alleged pattern. Equally striking is the failure to include vii 11 (between H and I), which narrates that the springs of the deep and the windows of heaven were opened, and viii 2 (between P and O ), which records that they were closed again and obviously matches the earlier verse. Further, yeser leb haddddm raC in viii 21 echoes wekol-yeser mahsebot libbo [sc. of hdadddm] raq rac in vi 5, but vi 5 is not in the part of the narrative in which Wenham claims to find a palistrophe, and that raises the question whether it is right to suppose that vi 10 is the place where a unit of the story begins.

The reason I pull heavily on Emerton here is that he does such a good job of showing why jumping onto such a structural analysis of chiastic structures pertaining to the flood story is problematic. Now, the chiasmus does feature in the Bible, just that in this instance, it is not relevant and presents more problems than it solves. As Emerton states: “It is far from being a strong argument against the analysis of the flood story into sources.” (p. 10) Sorry, Dave.

Part II of Emerton’s second analysis (Part II, p. 11-13) deals with Wenham’s fudging of the chronology problems of the flood account. I have spent long enough already, so go check it out in the source. Part III (Part II, p. 14-15) then analyses how Wenham deals with J and P both relying so heavily on the Atrahasis and Gilgamesh myths. I will need to look more closely into this, but if Armstrong is perhaps reliant on Wenham, I wonder if he also accepts the Genesis reliance on the Gilgamesh and Atrahasis myths?


By the way, if you want to look at Emerton’s take-down of all the other thinkers there, please read his posts papers. But, also, it is worth looking into Steven R. Scott who at least takes the theory one step further in seeing chiastic structures in a multiple source theory. See his piece that rolls off the tongue “Chiastic Structuring of the Genesis Flood Story: The Art of Using Chiasm as an Effective Compositional Tool for Combining Earlier Chiastic Narratives”. Here, he notes that both sources used chiasms that were then redacted into a unified chiastic structure. He concludes:

The above analysis provides strong evidence that chiastic structuring was used over several centuries—no matter which dating scheme one uses for J, P, and R. J is usually dated to the monarchy, P to the Exilic Period, and R in the Persian Period. There is a high likelihood that R was fully aware of the chiastic structuring of both J and P, because of the meticulous inter-splicing of the two chiasms. The redactor was careful to preserve both accounts: both versions were likely well established and cherished within various sections of the Jewish community.

Also, contrary to Wenham’s thesis, because both J and P were arranged in chiastic structures, it would not have been overly difficult for the redactor to splice the two stories together into one story. The two-source theory remains the best explanation for the doubling of events and also the two different language styles found in the text as we have it.

The other problems unsolved

All that the chiastic structure does is save someone like Armstrong from dealing with some micro-level repetitions. We’ll come back to this.

As mentioned, it doesn’t remotely solve contradictions, it doesn’t do much for problems of discontinuity, and it stays well away from dealing in any way at all with terminological and stylistic differences. I could list an inordinate number of contradictions that would still remain equally as thorny. See, for example, Schwartz in The Oxford Handbook of the Pentateuch in his chapter on the Documentary Hypothesis how contradictions are so problematic, yet solved by the DH (19.11-19.15 Epub).

Terminological and stylistic differences. These are literally the original cornerstones of DH/SH – the idea that different sources use different terms, vocab grammatical constructions and styles in their writing. Indeed, this is how the sources are teased apart into their distinctive groupings.

For a single author theory of chiastic structures, you just have to blithely ignore all of this. At least with the theories of people like Scott, above, they take this into account. But you can’t get away from multiple sources!

It doesn’t even solve repetition, silly!

This is really important to note. Such theorising is supposed to solve, at least for newfound chiasmus evangelical Dave Armstrong, repetitions. it does this by looking at a chapter, or a set of verses, and sees (lie a Rorschach picture) a pattern in there that explains away the repetitions. At best, this would only work for that set of verses.

The problem, you see, is that repetitions come in many shapes and sizes, and chiastic structures do nothing for macro-level repetitions. As Baruch Schwartz observes in The Oxford Handbook to the Pentateuch (and I produce this long quote here to show just how much Armstrong’s case fails in terms of repetitions – 19.3-19.10 [Epub]):

A case of redundancy in the Torah is essentially an instance of unexplained and unwarranted repetition of what has already been said. In the narrative portions of the Torah, redundancy is present whenever each of two or more passages purports to provide the one and only account of an event that can logically have occurred only once. In the legal sections of the Torah, redundancy is a case in which two or more passages purport to provide the legal stipulation that is to be fulfilled in a given, uniquely defined situation.

This phenomenon is extremely widespread. The creation of the cosmos, of humans, and of animals is described twice (Gen 1:1–2:4a and 2:4b–25); the establishment of the covenant with Abraham is recounted twice (Gen 15:1–21 and 17:1–27); the changing of Jacob’s name to Israel is related twice (Gen 32:28–29 and 35:9–10); the divine name, Yahweh, is revealed to Moses twice (Exod 3:13–15 and 6:2–9), among many others. Redundancy is also rife within individual narratives. For example, in the course of the story of the flood (Gen 6:5–9:17), the narrator twice describes the evil that spurred Yahweh’s decision to bring about the flood (Gen 6:5–6 and 6:11–12); we read twice that Yahweh informed Noah of his decision (6:17 and 7:4); twice we learn that he conveyed his instructions to Noah (6:18–21 and 7:1–3), and more. In the course of the account of Moses’s commissioning (Exod 3:1–4:17), Yahweh twice mentions that he has seen the affliction of his people and has decided to act (3:7–8 and 3:9); Moses twice expresses his objections to having the task imposed upon him (3:11, 13 and 4:1, 10, 13); twice Yahweh responds to his reservations (3:12, 14–15 and 4:2–9, 11–12, 14–16), and so forth. In all these cases and innumerable others, the individual passages provide no recognition that the event itself has already transpired or that it might not be the only such event. Every such narrative, and every similarly duplicated subsection of a repetitive narrative text, presents itself as the one and only account of the event described, as does its counterpart.

Turning to the legal portions of the Torah: twice the Israelites are commanded with regard to permitted and forbidden foods (Lev 11 and Deut 14:3–21), the prohibition of usury (Lev 25:35–37 and Deut 23:20–21), the sustenance of the poor from the produce of one’s field and vineyard (Lev 19:9–10 and Deut 24:19–21), the sabbatical year (Exod 23:10–11 and Lev 25:1–7, 20–22), and more. Three times they are given the laws pertaining to the manumission of slaves (Exod 21:1–11, Lev 25:39–46 and Deut 15:12–18), talionic restitution (Exod 21:22–25, Lev 24:17–22, and Deut 19:21), murder, manslaughter and asylum (Exod 21:12–13, Num 35:9–34, and Deut 19:1–13), and more. They are commanded with regard to the annual festivals four times (Exod 23:14–19, Exod 34:18–26, Lev 23:1–44, Deut 16:1–17; an additional section in Num 28–29, dedicated to the unique sacrifices offered on each festival day, complements the law in Lev 23). Just as in the narrative portions of the Torah, each of these passages is always presented as the sole and complete account of the legislation that it claims to convey, never as an addendum, continuation, or even emphatic reiteration of one or more of its counterparts. They thus compete with one another for the status of the authoritative promulgation of the command in question (see Deut 4:2, 13:1). Furthermore, these competing passages appear in completely different places in the Torah—a fact that cannot be explained reasonably under the assumption that the Torah is a unified work.

Not every case of formal or substantive similarity should be mistaken for redundancy. A single storyteller may recount two similar episodes, if he maintains that they both occurred and there is no categorical impossibility for this to have been so. For example, even if Abram’s wife Sarai was abducted by Pharaoh (Gen 12:10–20), she may also have been abducted later by Abimelech (Gen 20:1–18), and Isaac’s wife Rebecca may have subsequently been abducted by Abimelech as well (26:6–11) since, despite the similarity, the three accounts do not purport to be reports of a single event. Only mutually exclusive competition between two accounts constitutes redundancy.

The most conspicuous and serious instance of redundancy is not limited to two or three competing passages but is woven through the entire Torah. This is the account of how Israel received its laws. The story of the proclamation of the Decalogue and the establishment of a covenant at Horeb (Exod 19:2b–9a, 16aα2–17, 19; 20—23; 24:3–8, 11bβ–15a; 32:1–8, 10–25, 30–35; 33:6–11; 34:1, 4, 28) relates that the laws were written down and that the covenant that Yahweh made with the Israelites was concluded “on the basis of these words” (Exod 24:8), i.e. the written text of the laws. With regard to these laws the people said: “All that Yahweh has spoken (i.e. Exod 20:19–23:33) we will faithfully do” (Exod 24:7), and the story concludes with no expectation of additional laws to be given at some future time. This account thus purports to be the sole report of the lawgiving. Nonetheless, the reader is also presented with a second story of a covenant made at the same time, in the course of which Moses ascended a mountain—Sinai, according to this account—to hear the attributes of Yahweh’s mercy (Exod 19:9b–16aα1, 18, 20–25; 24:1–2, 9–11bα; 32:9, 26–29; 33:1–5, 12–23; 34:2–3, 5–27). Here too, a corpus of laws is given to Moses (Exod 34:11–26), he is commanded to record them in writing, and it is they that are referred to in the statement: “In accordance with these words I have made a covenant with you and with Israel” (Exod 34:27). This second account shows no signs of continuing, adding to, affirming, replacing, or denying the first; it too is presented as the one and only story of the conclusion of a covenant between Yahweh and Israel, in the course of which Yahweh conveyed his laws to Moses.

Interspersed between these two stories and extending over the long text that follows, a third account emerges, according to which Moses is told that the lawgiving will commence only after Yahweh’s portable dwelling, the tabernacle, has been constructed at the foot of Mount Sinai. Only then, by means of divine speech emanating from between the cherubim on the cover of the ark, will Yahweh communicate to Moses “all that I will command you concerning the Israelite people” (Exod 25:22). This plan too is then carried out exactly as promised (see Schwartz 1996a). Just as neither of the other two stories offers any intimation that the legislation it contains is only part of a larger body of laws and that more legislation will follow, this third story contains no indication that the legislation it contains (which extends throughout Leviticus and Numbers) is intended to supplement what preceded. All three accounts ignore each other’s existence entirely, and the author of one cannot be the author of either of the other two.

The same goes for the account of the lawgiving given in Moses’s second valedictory oration. Moses affirms (Deut 5:19–6:3) that the full body of Yahweh’s commandments was given to him at Horeb “on the day of the Assembly” (Deut 9:10; 10:4; 18:16), that is, on the same day that the Decalogue was proclaimed for the entire Israelite people to hear, but he goes on to relate that he did not convey this legislation to the people at the time but has rather kept it to himself until the present, four decades later (see Weinfeld 1991, 236–327; Nelson 2002, 73–85; Vogt 2006, 113–159). This thus constitutes a fourth independent and complete report of how and when Yahweh’s laws were conveyed to the Israelites.

Not only do we possess four independent accounts of the time, manner, and location of the lawgiving, each alleging to be the only such account, but each of the four also includes its own version of the laws themselves, each version purporting to be the laws and statutes commanded by Yahweh through the agency of Moses. The existence of four mutually ignorant legal corpora on the one hand, and of four mutually exclusive stories functioning as distinct narrative frameworks for them on the other, is incontrovertible evidence that the writings of several authors have been incorporated in the Torah.


Armstrong’s case doesn’t work. But also, it doesn’t work. Take the last two sentences and repeat them. A lot.

Moral: Don’t jump from the saucepan and into the fire. But if you do, I’m always here to help. Remember, I’m not the one burning you, I’m just the fire marshall pointing it out.|


[1] G. J. Wenham, “The coherence of the flood narrative”, VT 28 (1978), p. 336-48


terminological and stylistic differences.

"No wake up we need more of your classic rants to cut & paste to ..."

God: Some Unholy Questions for You
"I’ve just taken a brief look at your post, and read about Noah’s Ark. I’ll ..."

Biblical Hermeneutic Circularity
"Demonstrates why guns have no place in a civilised society."

Why I Am Atheist and Not ..."
"I remember him making almost exactly this same comment once under his skl handle."

Why I Am Atheist and Not ..."

Browse Our Archives