Internal complexity, chiasmus and 'Hebraisms'

Whether the book's intricacy (names, geography, chronology, chiastic structures, Hebrew-like syntax) exceeds what a dictating author could manage.

What the Church claims

The Book of Mormon's internal consistency and literary structures, especially chiasmus, could not have been produced by Joseph Smith and point to ancient Hebrew authorship.

Once the evidence is in, this claim: Contested. The apologetic inference from complexity to antiquity fails, and the chiasmus tallies are inflated, but critics have not matched the feat with a controlled comparator, and Alma 36 is a genuine design.

The record

  • Apologetic tallies: 337 proper names (188 unique), 175+ individuals, 125+ places, 425+ geographical movements, 430 proposed chiasms, over 100 names for deity; three calendar systems and 'dozens of editorial promises'.
  • John W. Welch identified chiasmus in Alma 36 in 1967-69. Boyd and Farrell Edwards (BYU Studies 43:2, 2004) computed that an eight-element Alma 36 chiasm has a chance likelihood of about 0.00000049 and, allowing for opportunities across the book, P about 0.00018; Mosiah 5:10-12 and Helaman 9:6-11 also scored low. Chiasms in D&C 88 (P = 0.52), Abraham 3 (P = 0.98) and a database manual introduction (P = 0.66) were consistent with chance.
  • Earl Wunderli ('Critique of Alma 36 as an Extended Chiasm', Dialogue) argued that the chapter's repetitiveness lets analysts pick among multiple appearances of key words and adjust boundaries, and that Alma 36 fails Welch's own literary criteria; the Edwardses replied (Dialogue 39:3) that statistical validity does not require those literary standards and put the ten-element chance at under one in 100,000.
  • Wunderli's 'An Imperfect Book' (Signature, 2013) argues from the text alone: Nephi's 55 years take 20% of the book while the next 415 years take 5%; Nephite writers quote scripture written after their time; prophecy is detailed up to Joseph's day and fails after; the theology, idiom and names are uniform across supposed authors and read like one 'white European-American'.
  • The Mosiah-first dictation order (Mosiah to Moroni, then 1 Nephi to Words of Mormon) is accepted by Welch and is standard on both sides; it means the book's opening, with its explanations of later matters, was composed last.
  • Internal slips exist: the 1830 text had King Benjamin alive at Mosiah 21:28 where Mosiah must be meant, corrected in later editions; Gardner cites the Benjamin/Mosiah error as evidence of memorized detail, FAIR lists it among textual changes.
  • Bruce and Brian Dale (Interpreter 2019) assigned Bayes factors of 0.5, 0.1 or 0.02 to 131 'positive correspondences' between the Book of Mormon and Michael Coe's 'The Maya', multiplied them, and reported combined odds of about 10^-132 in favor of historicity even after six negative correspondences; they acknowledge subjective weighting and an independence assumption.
  • Davis's oral-composition model explains the summary headings, resumptive repetitions and outline structures as products of 'laying down heads', and believing reviewers concede that the text shows planned composition.

The moving target

  • 2016-11-01: Hundreds of names woven into a complex set of events without notes is beyond a plowboy; even Fawn Brodie granted the book's coherence.

The exchange

John W. Welch; Edwards and Edwards (BYU Studies 2004; Dialogue 2006)

Alma 36 and other passages are deliberate chiasms with vanishingly small chance probabilities, a Hebrew form unknown to Joseph in 1829.

The answer: The Edwards analysis shows that Alma 36's repeated key words are unlikely to fall in mirror order by chance; it does not show ancient origin, and the authors say so. Parallel and inverted repetition is native to biblical and revival preaching, which Joseph absorbed, and to oral composition generally. Their own method finds that Joseph's D&C chiasms are chance-level, which cuts both ways: a speaker who produces one deliberate chiasm need not produce them everywhere. Wunderli's point that the chapter's repetitiveness lets an analyst select the mirrored terms remains a live objection to the 430-chiasm tallies, if not to Alma 36 itself.

What they get right: Alma 36 does look designed, and Edwards and Edwards were candid that chance cannot explain it while design does not establish antiquity.

Brian C. Hales (Interpreter 31); Scripture Central

The volume and consistency of names, movements and chronology across a text dictated out of order is beyond unaided memory.

The answer: The tallies are impressive but uncontrolled: no one has counted the same features in a comparable oral epic or a long dictated novel. The consistency is also imperfect (Benjamin/Mosiah; Wunderli's catalog of proportions, anachronistic citations and uniform voice). Long-form storytellers routinely keep hundreds of names straight, and Joseph had six years and one lost draft to fix the frame. The Mosiah-first order is a problem only if the author had no plan; on the oral-composition model the plan is the point.

What they get right: Critics have not produced a documented single-session dictation with this density of cross-reference; the closest comparators (A. J. Davis, Pearl Curran) are less intricate.

Bruce E. Dale and Brian Dale (Interpreter 32, 2019)

Bayesian accumulation of 131 Mesoamerican correspondences yields overwhelming odds for historicity.

The answer: Multiplying 131 hand-assigned likelihoods that are assumed independent, drawn from a single popular book, with the correspondences selected by the analysts, produces any number one likes; the authors concede the weights are personal judgments. The method has not been accepted by any statistician outside the apologetic community that these notes could locate. Six 'negative' items (horses, elephants, steel) each weighted as the strongest possible evidence still lose to a hundred weak positives, which shows the arithmetic, not the history.

What they get right: The authors invite reweighting and ran control comparisons on View of the Hebrews and the Spalding manuscript; the paper is transparent about its assumptions.

Both cases at full strength

The skeptics' case

Complexity is evidence of a capable author, not of an ancient one. The best statistical work on chiasmus, by believers, concludes only that Alma 36 was designed; the larger tallies rest on criteria loose enough to find chiasmus in a software manual, and Wunderli's critique of the selection problem has not been answered on its own terms. The name and geography counts have no control group. Against them stand the text's own tells: a modern proportion of narrative, prophecy that goes blind after 1830, one theological voice across a thousand years, and slips like a king alive after his death. The oral-composition model turns the remaining consistency from a miracle into a method.

The apologetic case

Nobody has shown a modern text dictated in one pass with this many interlocking names, journeys and flashbacks, and a reversed order of composition. Alma 36's structure survives its critics on the numbers. The forms that Ostler and Welch identify (covenant renewal, prophetic lawsuit) fit ancient Israel better than Palmyra. Complexity does not prove antiquity, but it raises the cost of every naturalistic theory.

What would change this

A controlled comparison showing that skilled oral performers cannot sustain this density of cross-reference; or verification of any Book of Mormon name or structure in pre-Columbian sources.

Closely related

Sources

An independent project. Not affiliated with The Church of Jesus Christ of Latter-day Saints or with any organization cited here. Summaries are written here; short quotations are attributed and linked. Every page names its sources so that it can be corrected, and every page can be challenged.