Does the worldwide diffusion of flood myths imply a datable Flood horizon?

Type: supporting argument (off-timeline) · Assessment: weak — as the claim that the spread of flood traditions across culturally separated peoples yields, or constrains, a date for Noah's Flood. The distribution is real and its bearing on whether the narratives share a historical kernel is weighed elsewhere and comes in plausible; this page inherits that verdict and asks only the narrower dating question — and there the argument fails. Grant even that the flood narratives descend from one remembered event: a scatter of stories carries no internal clock. It cannot say when the remembered flood happened, because the same distribution is equally produced by independent invention (destructive floods are a universal human experience), by cultural diffusion along known routes, and by shared memory carried in migration — none of which fixes a horizon, and the migration route points to deep time rather than a historical-era deluge. The one flood text that is datable is still only the Mesopotamian strand, which dates the record, not the event. Weak — scoped to the dating claim; the spread remains consistent with a real regional flood behind the Mesopotamian tradition, the kernel reading weighed on its own page and left untouched here.

The claim

The reckoning arguments on this timeline fix the interval from the genealogies; the ground arguments test whether a datable stratum can be the Flood. This one reaches for a date from a third kind of evidence entirely: the sheer ubiquity of flood stories. The observation is old and genuinely striking — nearly every ancient culture, on every inhabited continent, preserves a tradition of a world-destroying flood survived by a remnant in a vessel. The inference weighed here is the further one: that this near-universal distribution does not merely attest a flood somewhere, but points to one event within human memory, dispersing from a common source — and that the common-source reasoning implies a datable horizon, because the scattered peoples must all have carried the memory from before they separated, so the event predates the dispersion and can in principle be dated by it.

The narrow question here is only the dating one. Whether the narratives share a historical kernel at all — the corroboration question — is weighed on the material-record page and inherited, not re-argued: there the prediction that a real flood tradition should seed convergent stories comes in plausible, on the strength of oral traditions independently checked against the geology, while the stronger claim of a single worldwide deluge fails. Granting the milder, plausible verdict for the sake of argument, this page asks the question left over: does the distribution of the traditions supply a usable date?¹

Who advances it

The distribution-implies-a-single-datable-event inference is a staple of the modern creationist case, argued explicitly as a body of evidence rather than a passing remark. John D. Morris of the Institute for Creation Research puts the reasoning in its sharpest form: "When two separate cultures have the same 'myth' in their body of folklore, their ancestors must have either experienced the same event or they both descended from a common ancestral source that itself experienced the event" — and concludes that "the only credible way to understand the widespread, similar flood legends is to recognize that all people living today … have descended from the few real people who survived a real global flood on a real boat that eventually landed on a real mountain." Walter T. Brown, tallying "more than 230 flood legends," reasons the same way: "The many common elements in more than 230 flood legends suggest a common historical event that left a vivid impression on survivors of that catastrophe." The dating payload is implicit but real: a common source before the dispersion places the event before the peoples scattered — for the young-earth writers, before the Babel dispersion of Genesis 10–11, which their chronology in turn seats within a few centuries of Ussher's Flood.²

The evidence for

The distribution is real, and broad. This is the argument's genuine strength, granted in full. Flood narratives are a major structural feature of the mythologies of Eurasia, Oceania and the Americas; the catalogues run to two hundred and more; and the shared skeleton — a warning, a vessel, a remnant, a mountain, a world remade — recurs in cultures with no plausible route of contact. That much is not in dispute, and it is exactly what a real remembered flood would leave behind. The corroboration page grants more still: oral tradition has been shown, in a controlled test, to carry the memory of a real inundation faithfully across millennia. If the stories descend from a remembered event, the reasoning that the event predates the dispersion is not absurd on its face.

And a shared source would, in principle, order the events in time. The advocate's move has a real logic: if two separated peoples carry the same memory, the memory is older than their separation, so a genealogy of the traditions would be a rough clock — the more widely a motif is shared, the deeper it must sit. Pressed hard, that is an argument that the Flood lies at or before the earliest node from which the traditions fan out. The inference is not empty; it simply has to deliver an actual date, and it is there that it comes apart.

Adversarial assessment

What holds. The distribution is genuine, the shared skeleton is real, and — inherited from the corroboration page — a flood tradition with a historical kernel is a plausible reading of it. None of that is denied here. The argument's premise, that peoples the world over remember a flood, is sound; what fails is the leap from that premise to a date.

A scatter of stories has no internal clock. This is the decisive point. Even granting that the flood narratives all descend from a single remembered event, their distribution cannot say when that event occurred, because distribution is not a dating method — it records that a story spread, not when the thing it reports took place. The same worldwide pattern is produced, with no single datable deluge, by at least three mechanisms the argument cannot exclude. Independent invention: destructive floods are among the most universal of human experiences, so cultures on separate continents can arrive at flood-and-remnant stories without any shared event — humanity did not need one global deluge to fear drowning, only rivers, storms, coastlines, and memory. Cultural diffusion: the Mesopotamian flood tradition demonstrably travelled, its recensions reaching Megiddo and Ugarit by the fourteenth century BC, so shared features can mark a shared story rather than a shared event. And migration-era memory: Anna Birgitta Rooth's analysis of some three hundred North American creation myths found detail-motifs recurring in Eurasian myth so specifically that, on her reading, they "must be considered to have a common origin" — but a common origin carried across the Bering land bridge places the shared node in the Pleistocene peopling of the Americas, tens of thousands of years deep, not in a historical-era Flood. Each mechanism yields the same map; none yields a date; and the migration route, taken seriously, points away from any horizon Genesis can reach.³

The only datable flood text is still just the Mesopotamian one. Strip away the undatable oral distribution and ask what actually carries a date, and the answer is unchanged by the argument: the oldest date-bearing flood witnesses are the Sumerian King List, Atra-ḫasīs, Gilgamesh XI, and — as a transmitted number — Berossus' antediluvian kings, all second- and third-millennium BC Mesopotamian compositions or later reports of them. These date the record, the moment the tradition was written down, not the event the tradition remembers, which is exactly what the whole page is trying to fix. The worldwide distribution adds no independent date on top of them: it multiplies the witnesses to the story without supplying a single new anchor for the flood. So the distribution argument, at its best, returns the reader to the Mesopotamian strand it started from — and that strand dates a text, not a deluge.⁴

And the one date the argument does name is read off Genesis, not off the myths. The young-earth version supplies a horizon by tying the dispersion to Babel and Babel to the post-Flood genealogies of Genesis 10–11 — but that is circular as a use of the myth evidence. The date is not extracted from the distribution of traditions; it is imported from the same genealogical reckoning the short- and long-chronology arguments weigh on their own pages, and then the traditions are arranged to fit it. The distribution does no dating work of its own here: remove the assumed biblical dispersion timeline and the myths yield no number at all. An argument that can only date the Flood by first assuming the date it wants is not an independent dating argument.⁵

Where the traditions can be dated, they date to many floods, not one. The strongest single result in this whole field cuts directly against a single datable horizon. Patrick Nunn and Nicholas Reid checked twenty-one Aboriginal Australian coastal traditions against the reconstructed curve of postglacial sea-level rise, dating them, by the water depth each story's details require, across a span from about 7,250 to 13,070 years ago — a range of nearly six millennia, describing a gradual global rise experienced locally at many different times. The Mesopotamian silt bands behave the same way, real flood layers at Ur, Shuruppak and Kish laid down in different centuries. So the one part of the corpus that is datable does not converge on a horizon; it spreads across thousands of years of many local inundations, which is the direct disconfirmation of the single-dateable-event the distribution argument needs.⁶

Scope, and the supernatural. The verdict here is evidential, not metaphysical: it does not rest on ruling out a global Flood or on denying that God could act in one. Grant a real Flood freely — the objection is not "a worldwide flood is impossible" but "the worldwide spread of flood stories is not a clock, and cannot tell you which century the remembered flood belongs to." A miracle can be granted without repairing that, because the defect is in the inference, not in the premise: possibility is not proof, and a real event does not become datable merely because many peoples remember it.

Assessment

Weak — as a way of dating Noah's Flood. The premise the argument rests on is sound and is granted: flood traditions really are distributed across the inhabited world, they really do share a narrative skeleton, and — on the corroboration page's inherited verdict — they are plausibly the memory of real floods. But the distribution of traditions is not a dating method. Even granting a single remembered event, a scatter of stories carries no internal clock: independent invention, diffusion along known routes, and migration-era memory each reproduce the same worldwide map without fixing any horizon, and the migration route points to the deep Pleistocene rather than to a historical Flood. The only witnesses that actually bear a date are the second- and third-millennium Mesopotamian texts, which date the record and not the event; and the one horizon the young-earth version names is not read off the myths at all but imported from the Genesis genealogies and the Babel dispersion, which makes it circular as a use of the myth evidence. Where the traditions can be independently dated — the Australian coastal corpus, the Mesopotamian silts — they resolve into many floods at many dates, the opposite of the single datable event the argument requires.

The verdict is scoped, not sweeping. It falls on the dating claim — that the spread of flood myths yields or constrains a Flood date — and not on the reality of the distribution or on the corroboration reading, which stands on its own page: a real flood tradition with a historical kernel is neither reached nor disturbed by anything here. The honest residue is that the worldwide spread of flood stories is consistent with a real regional flood behind the Mesopotamian strand of the tradition — but consistency is not a date, and the distribution, for all its breadth, cannot tell us when the flood it remembers occurred.

Sources

The corroboration question — whether a historical flood tradition predicts convergent independent flood narratives, and whether those narratives share a source — is weighed and rated plausible (as a tradition with a historical kernel; weak if pressed to prove one worldwide deluge) on the material-record page at "A historical Flood tradition predicts convergent independent flood narratives — do they exist?", which this page inherits rather than re-argues. The decisive datum there — Patrick D. Nunn and Nicholas J. Reid, "Aboriginal Memories of Inundation of the Australian Coast Dating from More than 7000 Years Ago," Australian Geographer 47/1 (2016): 11–47 — establishes that oral flood memory can survive millennia, which is the premise this page grants before asking the narrower dating question. ↩
John D. Morris, "Why Does Nearly Every Culture Have a Tradition of a Global Flood?" Institute for Creation Research (1 September 2001), and the same argument in "Traditions of a Global Flood" (31 October 2014) — the "same event or … a common ancestral source" dilemma and the conclusion that all peoples "have descended from the few real people who survived a real global flood." Walter T. Brown, In the Beginning: Compelling Evidence for Creation and the Flood, "130. Flood Legends," creationscience.com (Center for Scientific Creation; full text also at archive.org) — "more than 230 flood legends" whose "many common elements … suggest a common historical event." The young-earth dating payload — the dispersion as the Babel event of Genesis 11, seated within a few centuries of Ussher's Flood — is developed from the same genealogies weighed at the Masoretic short chronology. ↩
That destructive floods are a near-universal human experience, and that independent invention, cultural diffusion, and shared psychological structure each explain the ubiquity of flood myths without a single datable deluge, is surveyed by Tim Brinkhof, "Why are flood myths so common in stories from the ancient world?" Big Think (3 February 2023). The demonstrable travel of the Mesopotamian flood tradition — its recensions reaching Megiddo and Ugarit by the fourteenth century BC — is set out on the material-record page at the Mesopotamian flood corpus. For the migration-era mechanism, Anna Birgitta Rooth, "The Creation Myths of the North American Indians," Anthropos 52/3–4 (1957): 497–508 (reprinted in Alan Dundes, ed., Sacred Narrative, 1984, 166–181), analysing some three hundred North American creation myths and finding detail-motifs shared with Eurasian myth that "must be considered to have a common origin" — a shared node the Bering-land-bridge peopling of the Americas places tens of millennia deep, not in a historical-era Flood. ↩
The oldest date-bearing flood witnesses are the Sumerian King List (ETCSL 2.1.1), Atra-ḫasīs, Gilgamesh XI, and Berossus' ten antediluvian kings (preserved through Eusebius and George Syncellus) — all second- and third-millennium BC Mesopotamian compositions or later reports, treated as datable witnesses on the summary timeline. They date the composition of the record, not the event it remembers; the distinction is the same one drawn on the Ur-silt argument (a real silt band dates a deposition, not necessarily the Flood). ↩
The young-earth horizon is reached by identifying the dispersion of peoples with the Babel event of Genesis 11 and dating Babel from the post-Flood genealogies of Genesis 10–11 — the same figures weighed, and found not to fix a Flood date, at the Masoretic short chronology and the Septuagint long chronology. The circularity is that the myth distribution supplies no number of its own: the date is imported from the genealogical reckoning and the traditions then arranged to fit it. ↩
Nunn and Reid (n. 1) date twenty-one Aboriginal Australian coastal traditions across c. 7,250–13,070 cal yr BP by the water depth each narrative's details require against the postglacial sea-level curve — a span of nearly six millennia of local inundations, not a single synchronous event. The Mesopotamian silt bands at Ur, Shuruppak and Kish are likewise of different centuries: C. Leonard Woolley's Ur deposit and the comparison of the layers in David MacDonald, "The Flood: Mesopotamian Archaeological Evidence," Creation/Evolution 8/2 (1988), with M. E. L. Mallowan, "Noah's Flood Reconsidered," Iraq 26/2 (1964): 62–82. Both corpora, where independently datable, resolve into many floods at many dates — the direct disconfirmation of the single datable horizon the distribution argument needs. ↩