Book of Mormon DNA literature review

The sources behind the Book of Mormon and DNA article — official Church statements, the ancient-genome literature, the Allen Ancient DNA Resource, and the method papers — with what each argues and whether it was opened.

Scope

This bibliography tracks four separate questions, which the article keeps apart: what the Book of Mormon and Latter-day Saint leaders actually claimed about Lamanite ancestry, and with what authority; what ancient genomes show about who was in the Americas and when; how much pre-Columbian genome-wide data exists, in which regions and periods; and what an admixture test could detect in principle.

Status key. Opened Aug 8, 2026 means the source was read, or re-read, in the session that wrote this bibliography and the article's quotation ledger. Opened Aug 4, 2026 means it was read in the drafting session that produced the article and ran research/aadr-american-genomes.py. Abstract only and Not opened mean exactly that, and the article's footnote says so.

Official and semi-official Latter-day Saint sources

  • "Book of Mormon and DNA Studies", Gospel Topics Essays (2016). Opened Aug 4 and re-checked Aug 8, 2026. The Church's official position and the document the article answers. Its reason is population-genetic rather than theological: a small founding group's contribution falls below detection, uniparental lineages are lost to drift, founder effects mean a founder's own lineage need not look like his homeland's, and the post-contact bottleneck removed rare lineages. All four of those arguments are correct, and the article concedes each in turn. Its most striking feature is that it disciplines its own side — "arguments that some defenders of the Book of Mormon make based on DNA studies are also speculative" — which is unusual for an institutional document and is credited as such.
  • Detailed Summary of Approved Adjustments for the 2013 Edition of the Scriptures, page 21. Opened Aug 4 and re-checked Aug 8, 2026 from the local PDF. The Church's own record of the "principal ancestors" → "among the ancestors" change, in its own words. This document catalogues the change; the essay dates it, to 2006. Keeping those two roles distinct is what fixes the error the series has retired.
  • Introduction to the Book of Mormon, current edition. Opened Aug 4, 2026. The current wording. The 1981 wording is quoted from the adjustments document rather than from a 1981 printing, which was not opened.
  • Spencer W. Kimball, "Of Royal Blood", Ensign, July 1971. Opened Aug 4, 2026, read in full. The clearest single statement of the hemispheric claim by a senior leader, with a population figure and an explicit definition of who counts as a Lamanite. Its authority tier — an individual apostle in a correlated Church magazine — is stated in the article rather than inflated.
  • "Indian Student Placement Program" and "Lamanite Identity", Church History Topics. Opened Aug 4, 2026. The first supplies the numbers that turn the hemispheric claim from rhetoric into institutional action; the second is the Church's most cautious current statement. Both are the Church's own history writing, which is why the article uses them rather than a critic's account of the same events.
  • Joseph Smith, "Church History" (the Wentworth Letter), reprinted in the Ensign, July 2002. Opened Aug 4, 2026. Used for the hemispheric claim in the Church president's own published account. The 1842 Times and Seasons original was not collated, and the reprint carries editorial spelling and bracketed insertions; the article's footnote says so. The Joseph Smith Papers source note was retrieved and confirms the publication data, but the document transcript did not render and was not read.
  • Doctrine and Covenants 28, 30 and 32, and the Book of Mormon passages quoted in "What the text says." Opened Aug 4, 2026 at churchofjesuschrist.org and re-checked Aug 8, 2026. Scripture is cited inline by chapter and verse in the ordinary way; footnotes carry the wording where a specific reading is load-bearing.
  • Dallin H. Oaks, quoted in the Gospel Topics essay. Not opened. The original is not identified in the essay and was not traced. The article's footnote labels the attribution second-hand, which is the right treatment for a quotation reached only through another document.

Ancient genomes

  • Eske Willerslev and David J. Meltzer, "Peopling of the Americas as inferred from ancient genomics," Nature 594 (2021): 356–364. Abstract only — the full text is paywalled and was not obtained. The abstract supplies the two sentences the article uses, both of which are conclusions rather than data, and nothing is attributed to the body. This is a real weakness: a review this central to the article's framing should be read in full.
  • Morten Rasmussen et al., "The genome of a Late Pleistocene human from a Clovis burial site in western Montana," Nature 506 (2014): 225–229. Opened Aug 4 and re-checked Aug 8, 2026 — read in full from the PMC author manuscript, PMC4878442. Its reason for mattering here: Anzick-1 is directly radiocarbon dated on the bone, carries founder-clade uniparental lineages, and is closely related to all Indigenous American populations, which places the founding ancestry ten thousand years before Lehi.
  • Morten Rasmussen et al., "The ancestry and affiliations of Kennewick Man," Nature 523 (2015): 455–458. Opened Aug 4 and re-checked Aug 8, 2026 — PMC4878456. The single most important sample in the argument, because X2a was the one mitochondrial thread a Book of Mormon reader could pull on and Kennewick Man carries it at the root of the clade thousands of years too early. The correction owed on the date has now been made. The reference list was read directly: reference 2 is Owsley, D. W., and Jantz, R. L., eds., Kennewick Man: The Scientific Investigation of an Ancient American Skeleton (Texas A&M University Press, 2014), and reference 1 is Taylor et al., Science 280 (1998): 1171–1172. The article's footnote now says plainly that the 8400–8690 cal BP redating reaches it second-hand through Rasmussen et al. and names Owsley and Jantz as the underlying source.
  • Douglas W. Owsley and Richard L. Jantz, eds., Kennewick Man: The Scientific Investigation of an Ancient American Skeleton (Texas A&M University Press, 2014), 669 pp. Not obtained. Searched on Aug 8, 2026: archive.org holds no copy under any title-plus-editor query, the publisher serves no readable text, and the Google Books API was rate-limited out. This is the study that produced the 8358 ± 21 ¹⁴C BP redating, so its absence is the one place where a load-bearing number in the article is two steps from its measurement. The mitigation in the article is to print the independent 1998 determination — 8410 ± 60 BP, calibrated 7250–6390 BC — from the AADR annotation beside it, which was opened and read.
  • Ashot Margaryan et al., "Population genomics of the Viking world," Nature 585 (2020): 390–396. Abstract only — paywalled. It is the publication of record for all 23 Greenland Norse individuals the article counts out of the AADR, which is why it is cited at all; nothing beyond the abstract is attributed to it.
  • Ida Moltke et al., "Uncovering the Genetic History of the Present-Day Greenlandic Population," American Journal of Human Genetics 96 (2015): 54–69. Opened Aug 8, 2026 — PMC4289681, abstract and results read. This is what turns the Greenland Norse from an anecdote into a control experiment. Its reason for mattering: it is a direct test, on 4,674 living Greenlanders, of whether a documented pre-Columbian European colony left descendants in the population that followed it, and the answer is no — the ~25% European ancestry present in most Greenlanders traces to recent Danish contact, not to the Norse.
  • Viridiana Villa-Islas et al., "Demographic history and genetic structure in pre-Hispanic Central Mexico," Science 380 (2023): eadd6142. Opened Aug 8, 2026 — PMC12087575 author manuscript. Used as an external check on the census: its twelve genome-wide individuals all date AD 600–1351, and its ten Mexican individuals appear in the AADR release counted here under Mexico_Medieval. Its twenty-seven mitogenomes do reach 320 BC, which is the honest qualification the article now makes — central Mexico is not unsampled in the Book of Mormon window, it is unsampled for the genome-wide data an admixture test needs.
  • Douglas J. Kennett et al., "South-to-north migration preceded the advent of intensive farming in the Maya region," Nature Communications 13 (2022): 1530, and Cosimo Posth et al., "Reconstructing the Deep Population History of Central and South America," Cell 175 (2018): 1185–1197. Both downloaded Aug 8, 2026 and stored in the library; neither is cited in the article. Kennett's Belize individuals are the closest ancient genomes to a Book of Mormon heartland, and every one of them predates 2000 BC, which is consistent with the census but adds nothing the census does not already say. They are recorded here so the next session does not re-download them.
  • Brendan D. O'Fallon and Lars Fehren-Schmitz, "Native Americans experienced a strong population bottleneck coincident with European contact," PNAS 108:51 (2011): 20444–20448. Opened Aug 4 and re-checked Aug 8, 2026 — PMC3251087, read in full. Used against the article's own critical instincts: it establishes that the post-contact contraction in effective population size was about half, which is a real bottleneck and not the tenfold one that critical arguments usually assume. Distinguishing census size from effective size is the whole point of citing it.
  • The haplogroup X2a divergence-date literature. Collected but not read. The article says so in the text and states nothing about X2a beyond what Kennewick Man directly shows. This is the most conspicuous deliberate silence in the article and it is the right one: dates quoted from unread papers are the failure mode this series exists to avoid.

The data layer

  • Allen Ancient DNA Resource, release v66.p1, published 8 June 2026 (Harvard Dataverse, doi:10.7910/DVN/FFIDCW, file v66.p1_1240K.aadr.PUB.anno, file id 13994515). Now stored in the library at sources/data/mallick-2026_aadr-v66p1-1240k.anno. Counted directly, not read. The AADR is the field's curated compilation of published genome-wide ancient data, which is what makes the census in this article possible at all: reviews describe the record qualitatively, and this article counts it.
  • research/aadr-american-genomes.py and research/aadr-american-genomes-output.txt. The script's inclusion rules and its two data-quality corrections — dropping Northern Mariana Islands samples filed under the political entity "USA," and treating bare superhaplogroup labels such as "R" as unresolved rather than as West Eurasian — are documented in its docstring and summarized in the article's footnote. Every figure in the census table, the haplogroup tallies, the five anomalous mitochondrial calls, the Greenland Norse group, and the two Mesoamerican individuals in the narrative window were re-checked against the output file on Aug 8, 2026 and all match.
  • The known limit of this data layer: the AADR is a compilation of published data, so its regional and temporal gaps are the field's publication gaps. The finding that only two Mesoamerican individuals fall in the 600 BC – AD 400 window is a statement about what has been published and released, not about what has been excavated.

Population genetics: the sources behind the reasoning the article concedes

The article's longest unevaluated section states the Latter-day Saint case in its strongest form, and that case is population genetics. These four sources are what keep it from being asserted as general knowledge.

  • Douglas L. T. Rohde, Steve Olson, and Joseph T. Chang, "Modelling the recent common ancestry of all living humans," Nature 431 (2004): 562–566. Opened Aug 8, 2026 from the library copy, read in full. This is the source for the single most important idea in the subject: genealogical ancestry and genetic ancestry diverge, and past the identical-ancestors point a person inherits "little or no actual genetic inheritance from the vast majority" of their genealogical ancestors. Its reason is mechanical rather than statistical — DNA is inherited in large recombining segments, so the number of ancestors who contribute any DNA at all stops growing long before the number of ancestors does. Note the limit: the paper's dates are simulation outputs under assumed migration rates, and the paper says so, so the article cites the mechanism and flags the dates as model outputs.
  • Agnar Helgason et al., "A Populationwide Coalescent Analysis of Icelandic Matrilineal and Patrilineal Genealogies," American Journal of Human Genetics 72 (2003): 1370–1388. Opened Aug 8, 2026 — PMC1180299, read in full. This is the Icelandic result the Gospel Topics essay alludes to, and it is stronger than the essay's summary of it. Its reason for mattering: Iceland's genealogical records are complete enough that uniparental lineage loss can be counted rather than modelled. Of the potential matrilineal ancestors born in 1698–1742, only 6.6% have a matrilineal descendant among Icelanders born after 1972; for patrilines, 10.3% — in three hundred years, in a population that grew fivefold. This is why the article retires the "Sariah's mitochondrial line is missing" argument rather than repeating it.
  • Noah A. Rosenberg and Magnus Nordborg, "Genealogical trees, coalescent theory and the analysis of genetic polymorphisms," Nature Reviews Genetics 3 (2002): 380–390. Opened Aug 8, 2026 from the library copy. Cited for two textbook facts the article would otherwise be asserting on its own authority: the definition of genetic drift, and the coalescent relation between effective population size and the expected depth of a genealogy. The house rule is that a standard population-genetic fact still needs a named source.
  • Alexander Koch et al., "Earth system impacts of the European arrival and Great Dying in the Americas after 1492," Quaternary Science Reviews 207 (2019): 13–36. Opened Aug 8, 2026, CC BY copy from UCL Discovery. Cited only for the ninety-percent census figure and the 60.5 million pre-1492 population it is computed against, because the article's own correction depends on separating that census number from the effective-population-size contraction O'Fallon and Fehren-Schmitz measured. The paper's own subject is the carbon cycle, and the article does not use it for anything else.
  • Gil Atzmon et al., "Abraham's Children in the Genome Era," American Journal of Human Genetics 86 (2010): 850–859, and Marc Haber et al., "Continuity and Admixture in the Last Five Millennia of Levantine History," American Journal of Human Genetics 101 (2017): 274–282. Both opened Aug 8, 2026 — PMC3032072 and PMC5544389. Together they are the reference-panel problem stated with sources: modern Jewish populations carry "variable degrees of European and North African admixture" and so are a poor proxy for an Iron Age Levantine, while the ancient genomes that are the right proxy — five Canaanite individuals from Sidon — were only published in 2017.

Latter-day Saint scholarship on geography

  • John L. Sorenson, "Digging into the Book of Mormon," Ensign, Sept 1984, 27. Opened Aug 8, 2026, read in full. Its value is its date. It sets out the limited Mesoamerican reading in the Church's own magazine sixteen years before the DNA argument became public, and traces the reading back to a 1939 geography by the Washburns. That is what makes the objection "Latter-day Saints only adopted limited geography after the DNA results" overstated. Authority tier is stated in the article: an individual scholar in a correlated magazine, disclaiming official standing in his own words.
  • "Book of Mormon Geography," Topics and Questions (2023). Opened Aug 8, 2026. This corrected an overclaim in the article's own draft, which had said Mesoamerica is where "every serious Latter-day Saint geography now places the story." The Church's current statement records North American, Central American and South American readings side by side and takes no position among them, so the article now says most Latter-day Saint scholarship and cites both this page and Sorenson.

Method

  • Éadaoin Harney, Nick Patterson, David Reich, and John Wakeley, "Assessing the performance of qpAdm: a statistical tool for studying population admixture," Genetics 217:4 (2021): iyaa045. Opened Aug 4 and re-checked Aug 8, 2026 from the local PDF. The only source in the article that quantifies what an admixture test can actually resolve, which is what turns "no Israelite DNA has been found" from a rhetorical claim into a testable one. The conversion of Harney's standard errors into a 3% and a 10% detection threshold is the article's own arithmetic, stated as such in the footnote. That step should be checked by someone who works with qpAdm before publication; it is the one number in the article that is a step removed from its source.

Not obtained, and what each would settle

  • The full text of Willerslev and Meltzer 2021. Would let the peopling framework be described from the review rather than from its abstract.
  • The haplogroup X divergence-date literature. Would let the X2a section say something quantitative instead of resting on a single individual.
  • Owsley and Jantz, Kennewick Man: The Scientific Investigation of an Ancient American Skeleton (Texas A&M University Press, 2014). Would let the Kennewick date be cited to the study that produced it. Searched Aug 8, 2026 and not found — see the entry above for what was tried.
  • Mesoamerican Preclassic and Classic demography. Still the largest gap in the article. Koch et al. 2019 and Denevan (through Sorenson) give hemisphere and regional figures for 1492, which is two thousand years too late to fix the parameter the founder-fraction arithmetic scales with. Until a Preclassic figure is sourced, the article gives a range of illustrative values and says so in the footnote rather than picking one.
  • The Dallin H. Oaks original. Would replace a second-hand quotation with a first-hand one.
  • The 1842 Times and Seasons text of the Wentworth Letter. Would replace a modernised reprint with the printed original on a load-bearing hemispheric-claim quotation.
  • An Iron Age Levantine ancient-genome reference panel, and a qpAdm run against it. This is not a bibliographic gap but a research gap, and it is the thing that would actually move the question. The article's own conclusion is that the decisive test — dense, directly dated, high-coverage ancient genomes from Preclassic and Classic Mesoamerica — is available in principle and has barely been run in the right place and time.
  • A documented pre-2006 Church statement, at First Presidency level, teaching the limited-founder model before the genetic results were public. No such statement was found, and the article's central charge on the independent-motivation question depends on its absence. A single counterexample would blunt that charge substantially, so the search should be pushed harder than it has been.

The simulation that was not run

research/model-plan.md scopes a founder-survival simulation: what parameters it needs, what data would supply them, what it would output, and — importantly — what it would and would not prove. It was written as a plan, and no simulation has been run and no simulated number appears anywhere in the article. The article says so in its own text. The two arithmetic results the model would refine, the founder-fraction calculation and the qpAdm detection threshold, already point the same way, so the qualitative conclusion does not depend on the model; what the model would sharpen is the ancient-DNA power calculation, which is the question most worth answering next.