By the Numbers · Genesis 5 & 11
Why Did They Live 900 Years?
Adam lived 930 years. Seth 912, Methuselah 969, Noah 950. Then, after the flood, the ages fall away: Shem 600, Peleg 339, Abraham 175, and at last Moses at 120. The modern reader has three quick verdicts ready. It is legend. It is symbolism. It is a mistranslation, and “years” once meant months.
Before choosing one, look at how the ages behave. They do not scatter like legend; they fall along a curve, and the curve is choosy about what could have made it.
1 · The raw data
The ages, as written
Here is the whole line from Adam to Moses, with each recorded lifespan. The figures follow the Septuagint, the Greek Old Testament this site uses throughout for its chronology; the traditions agree on the pre-flood totals almost exactly and part company after the flood, which is the subject of §4. Tap any name for its verse.
Before the flood · Genesis 5
After the flood · Genesis 11
After Abraham · the recorded deaths
Every figure checked against the text. Ages follow the Septuagint in the reconstruction defended by Henry B. Smith Jr.; the pre-flood totals are the same in the Hebrew and Greek (only Lamech differs), while the Genesis 11 totals diverge and the Greek carries a second Cainan. The full source note is in the written argument.
Read down the column and the first thing to say is the honest one: nobody lives 900 years. That is exactly why the quick verdicts are tempting. So let us not argue about whether the ages are possible. Let us ask a different question, one the numbers can actually answer: whatever these figures are, how do they behave?
2 · The shape
It really is a curve
Plot every age in order, Adam on the left, Moses on the right. If these were the ages of legend you would expect either a flat line of uniform hugeness or a random scatter of impressive numbers. Instead:
With JavaScript on, the ages listed in section 1 are drawn here as a curve: a high pre-flood plateau around 900 years, a sharp drop at the flood, a second decline through Genesis 11, and a levelling near 120 at Moses.
One long descent, in two steps, from just under a thousand years to sixscore. Enoch, who did not die, is marked apart. This is not the shape of legend. It is the shape of a process. The rest of the page asks which processes could have drawn it, and which could not.
3 · The test
What the shape rules out
Here is the same curve, with the ages drawn as pale points. Lay each of the three quick verdicts over it in turn and watch what each one would have to look like. Two of them cannot fit on the chart; the third fits, but not the way its friends hope.
With JavaScript on, this chart overlays three explanations on the ages. In brief: the Sumerian King List’s pre-flood reigns run to tens of thousands of years and will not fit on the same axis; random myth produces scatter, not a descent; and a simple halving model crashes to the human floor in three or four generations, far faster than the recorded ages, which take a dozen.
So legend is the wrong scale and myth is the wrong shape. The simple dilution curve is the interesting failure: it is close to the mechanism this site actually argues for, a real lineage thinning into the surrounding population, and yet the naive version falls too fast. The real decline is gentler than raw halving allows, which is a genuine constraint on any explanation, and one the written argument takes up rather than hides.
A quicker fix, and why it fails on either text
One popular rescue keeps the numbers and reinterprets the unit: perhaps “years” once meant lunar months, so every figure should be divided by about twelve. Methuselah’s 969 becomes a reasonable 78. It is usually refuted from the low fathering ages of the Hebrew text, where dividing turns Enoch into a father at five. But this site reads the Septuagint, whose fathering ages are a century higher, so that particular refutation does not bite: Enoch’s Greek figure of 165, divided by twelve, is a boy of fourteen, unusual but not impossible. The theory still fails, and on the figure the two traditions share:
| Fathered his son at | Years (as written) | If those were months | |
|---|---|---|---|
| Enoch | 165 | 13.8 years old | strained, but possible |
| Adam | 230 | 19.2 years old | possible |
| Shem | 100 | 8.3 years old | impossible |
| Nahor | 79 | 6.6 years old | impossible |
Shem is a hundred years old when he fathers Arphaxad, and the Hebrew and the Greek agree on that hundred exactly. Read as months, it makes him a father at eight. The divisor that tames the 900-year lifespans starves the fathering ages below childhood, and it does so on both texts. You cannot pick one factor that keeps every number human at once, which is what a genuine change of unit would require.
4 · The cross-check
Two traditions, one structure
Suppose a scribe did invent these ages, composing them as a numerical pattern laid over the genealogy. If he did, the pattern lives in the exact figures he chose, so a rival version of the text that changed those figures should carry a different pattern, or none. That is a test we can actually run, because two traditions of Genesis survive and they genuinely disagree. After the flood the Hebrew and the Greek differ on almost every number: the Masoretic figures for the Genesis 11 fathering ages run about a hundred years lower, the Hebrew lacks a second Cainan the Greek preserves, and its post-flood lifespans are shorter. So if the shape of the decline were an accident of one set of numbers, the other set should not reproduce it. Switch the curve between the two traditions and watch, not the individual heights, but the two places where it drops.
With JavaScript on, this chart switches between the two traditions. Both place the same two cliffs at the same generations, Noah to Shem and Eber to Peleg, even though the Greek and Hebrew figures differ almost everywhere after the flood.
| The two drops | Masoretic | Septuagint |
|---|---|---|
| Noah to Shem | 950 to 600 (down 350) | 950 to 600 (down 350) |
| Eber to Peleg | 464 to 239 (down 225) | 504 to 339 (down 165) |
| Average lifespan, Shem to Eber | 484 years | 532 years |
| Average lifespan, Peleg to Moses | 174 years | 204 years |
| The collapse across the break | down 64% | down 62% |
The two traditions descend from a common original, so agreement in structure is partly inheritance; the real test is that centuries of divergence moved nearly every figure and still did not move the breaks. The first drop is identical to the year in both. The second lands at the same name in both, and in both it opens a gap the series never recloses: every life from Shem to Eber exceeds 430 years, and no one from Peleg onward reaches 350 in the Greek or 240 in the Hebrew. A pattern that lived in the exact figures would not survive the figures changing. The structure is older than either set of numbers.
5 · The cliff and the break
One cliff, and a second break
The ages do not slide down a single ramp. They hold high for nine generations, fall off a cliff at the flood, and then, part-way down the long descent that follows, they step down again at Peleg and never come back up. The two features are not the same kind of thing, and it is worth being exact about each. The first is a cliff by any standard: a 350-year fall off a level plateau, identical in both traditions. The second is a break in the middle of a decline already underway, and calling it a second cliff would overstate it; other single steps in the series are proportionally almost as large. What marks Peleg out is something else: his generation is where the series changes level for good. Both features sit where a reader of this site would expect the conditions of life to break: the flood, and the dispersal “in Peleg’s days” when “the earth was divided” (Genesis 10:25).
The cliff
Noah 950 → Shem 600
−350 years
Off a nine-generation plateau, identical in the Hebrew and the Greek. The generation that passed through the flood.
The break
Eber 504 → Peleg 339
−165 years
The largest single fall after the flood in both traditions (225 in the Hebrew), and the last of its size: from Peleg on, no recorded life returns to the old range. The generation of the dispersal.
Between and after the two breaks the ages decay smoothly, the way a fitted exponential decays. Before Peleg, every post-flood life exceeds 430 years in both traditions; from Peleg on, none reaches 350 in the Greek or 240 in the Hebrew. A plateau, a cliff, a decline, a step to a lower level, a longer decline: that is what the numbers actually are.
This page is about the ages, not the dates. Where these people sit in calendar time, and how the flood and the dispersal are placed against Egyptian and Mesopotamian history, is the work of a different page. If you want the chronology, the years BC and how they are anchored, read How We Date Adam. Here the claim is only this: the ages themselves carry two breaks, and the breaks are real enough to survive translation.
6 · Where it lands
Threescore years and ten
Follow the second decline to its end and it does not fall to zero. It flattens. The fitted curve settles onto a floor a little under a hundred years, and that is precisely where the rest of the Bible, and the rest of human life, already sits.
With JavaScript on, the post-flood ages are fitted as an exponential that levels onto a band of about 70 to 80 years, the span named in Psalm 90:10 and reached by the time of Moses.
The Septuagint fit levels near 82 years; the Hebrew near 57. Either way the curve arrives at the human span the Psalm names.
Long before, at the edge of the flood, the same limit had been spoken: “their days will be a hundred and twenty years” (Genesis 6:3). Moses, who wrote the Psalm, died at exactly 120 with his eyes undimmed (Deuteronomy 34:7), standing at the hinge where the long ages end and ordinary life begins. The curve does not trail off into myth. It lands, on the number a grieving Psalm and a dying prophet both already knew.
7 · What the ages are
Not legend. A record with a shape.
Set the verdicts against what we have seen. Legend is the wrong scale: the pre-flood reigns of Israel’s neighbours run to tens of thousands of years, and the Genesis ages, in their sober hundreds, look nothing like them. Mistranslation fails the arithmetic: no single change of unit keeps both the lifespans and the fathering ages human. Single-tradition invention cannot survive the cross-check: the Hebrew and the Greek disagree on the figures and agree on the structure. And simple dilution, the honest near-miss, falls too fast to be the whole story.
What is left is not a pile of rejected theories but a positive and rather demanding fact: a real, structured decline, in two phases, breaking at the flood and the dispersal, surviving translation, and landing on the span human life still keeps. Whatever produced the ages left a fingerprint, and a fingerprint is the sort of thing you can follow. This page has not proved that the decline is biological rather than, say, a symbolic pattern; that possibility the argument leaves open in honesty. It has shown that the numbers behave like data, and that the easy ways of dismissing them do not survive contact with the shape they make.
Follow that shape far enough, the written argument proposes, and the lifespans stop being an embarrassment and become a signal: a founding lineage thinning into an already-peopled world, its long lives fading generation by generation toward the common lot. God gives the years, and God draws them down. The strangest numbers in Scripture keep behaving this way. Learn how the number behaves, and it starts testifying.
This is one of three investigations in the By the Numbers series. Its companion, on the census that says 603,550 men left Egypt, asks the same question of a different set of numbers: what convention are we failing to read?
The written argument
The Ages as a Demographic Signal
Reading the lifespan decline of Genesis 5 and 11 as a record, not a riddle
The recorded lifespans fall in a two-phase curve, holding near 900 years before the flood, dropping at Noah to Shem, and declining again through Genesis 11 to a floor near the human span. Both textual traditions place the two breaks at the same genealogical positions while disagreeing on nearly every figure. That structure resists the standard dismissals and reads instead as a demographic signal: a founding lineage whose long lives thin into an already-peopled world.
1. The four readings, and where each breaks
Four verdicts have been offered on the ages of Genesis, and three of them are ways of not taking the numbers seriously.
Legend. The obvious comparison is the Sumerian King List, whose antediluvian kings reign for 28,800, 36,000, even 43,200 years apiece, eight of them totalling 241,200 years before its own flood, after which reigns drop abruptly to 1,200 and begin their long descent toward historical lengths (Weld-Blundell 444, per the ETCSL edition). The parallel is real and it is instructive, but it cuts the other way. The King List reigns are one to two orders of magnitude larger than any Genesis age, they are built transparently from the sexagesimal units of sars and ners, and they collapse at the flood in a single step. Genesis works in sober hundreds and declines across a dozen generations. If Genesis were imitating the King List genre it imitated it by discarding everything that makes the genre what it is.
Mistranslation. The lunar-month theory divides every figure by roughly twelve, rendering Methuselah a believable 78. It is usually dispatched from the Masoretic fathering ages, where the division makes fathers of five-year-olds. On the Septuagint figures this site follows, the century-higher fathering ages blunt that particular blade: the Greek Enoch fathers at 165, a plausible fourteen once divided. But the theory dies on the figure the traditions share. Shem is 100 years old at Arphaxad’s birth in both the Hebrew and the Greek; divided, a father at eight. Nahor’s Greek 79 gives a father at six. No single divisor keeps the lifespans and the fathering ages simultaneously human, and a genuine change of unit would have to.
Invention. A scribe composed the numbers as a pattern over the genealogy. This is the strongest of the three, but it founders on the two-tradition test. The Masoretic Text and the Septuagint disagree on almost every post-flood figure; the Masoretic Genesis 11 fathering ages run about a hundred years lower, the Hebrew lacks the second Cainan the Septuagint preserves, and its post-flood lives are shorter. Yet both traditions place the same two structural breaks at the same genealogical positions: the fall from Noah to Shem, identical at 350 years in both, and the collapse at Eber to Peleg, the largest single-generation fall in each tradition’s post-flood series (225 years in the Hebrew, 165 in the Greek), with the average lifespan across that break dropping by roughly three-fifths in both (64% and 62%). A numerical game invented in one tradition and translated into another that altered all its figures would not hold its structure fixed while its numbers moved. The location of the breaks is more stable than the numbers that make them, which is the opposite of what invention predicts.
A record. The fourth reading takes the ages as what they present themselves to be, a register of real lifespans, and asks what process would draw this particular curve. That is the reading the rest of this argument develops. It does not claim to have excluded a deliberate symbolic pattern as a bare logical possibility; it claims that the numbers behave like data, and that the burden has shifted to anyone who would call so structured and translation-stable a signal a mere literary device.
2. The curve, measured
Treated as a series and analysed with the ordinary tools, the decline has definite structure. Two breaks were specified in advance from the two-worlds model, at the Noah-to-Shem transition and the Eber-to-Peleg transition, and then tested; this matters, because a break found by searching a declining series for its steepest step proves nothing, whereas a break predicted before the data are inspected is a genuine test. Within the post-flood phase, from Peleg to Moses, an exponential decay of the form L(n) = A·e^(−kn) + C fits well: in the Septuagint, an asymptote of 82 years with R² = 0.922; in the Masoretic, an asymptote of 57 years with R² = 0.820. The Septuagint’s smoother, better-fitting descent is reported as a finding, not smoothed over: its longer post-flood lives and its second Cainan yield a more regular monotone decline than the Hebrew, whose series is roughened by the flat Peleg-Reu pair and by the small rise at Eber.
That rise is worth naming rather than hiding. In the Masoretic Text, Eber lives 464 years against Arphaxad’s 438: longevity ticks up for one generation, immediately before the Peleg collapse. No simple dilution model predicts a one-generation increase, and this argument does not pretend one does. It is an unresolved feature of the Hebrew data, flagged here because a reading that only reported the figures friendly to it would deserve none of the trust the two-tradition test is meant to earn.
| Phase | Fit (Septuagint) | Asymptote | R² |
|---|---|---|---|
| Pre-flood plateau | not fitted (too few points; described only) | near 910 | — |
| Post-flood decline (Peleg to Moses) | L(n) = 284·e^(−0.192n) + 82 | 82 years | 0.922 |
3. How strong is the second break?
An argument is only as good as its treatment of its own weakest point, so let it be said plainly: the two breaks are not equally strong. The flood cliff is a 350-year fall off a nine-generation plateau, identical in both traditions; nothing else in the series resembles it. The Peleg break is a large step in the middle of a decline already underway, and in proportional terms it is not even unique: in the Septuagint the fall from Eber to Peleg is 33%, while the fall from Serug to Nahor a few generations later is 37%. A reader who checks the ledger will find that, and this page would rather name it than be caught by it.
Two things nonetheless mark Peleg’s generation out, and neither applies to Nahor’s. First, the step at Peleg is the largest absolute fall after the flood in both traditions, 165 years in the Greek and 225 in the Hebrew, landing on the same name in both. Second, and more telling, it is a change of level rather than a wobble in the decline: every recorded life from Shem to Eber exceeds 430 years in both traditions, and no life from Peleg to Moses reaches 350 in the Greek or 240 in the Hebrew. The series steps down and never comes back. Nahor’s steep step, by contrast, happens inside the lower range and changes nothing about it.
A second objection deserves the same candour. Genesis 10:25 itself says the earth was divided in Peleg’s days, in both traditions, so a scribe shaping either text had a motive to place something at that name; and the two traditions are not independent witnesses but divergent copies of a common original. The reply is that the verse marks a territorial event, not a demographic one; it says nothing about lifespans. To manufacture the pattern the ledger actually shows, a scribe would have to depress not one flattering figure at Peleg but the whole subsequent series, six or more figures held below the old range, and the Hebrew and Greek would each have to do it while disagreeing about nearly every individual number. What the common original explains is the location of the break. What it does not explain is why the numbers on either side of it, moved independently by two traditions, keep the two levels apart.
4. Why the simple dilution model fails, and what that tells us
The most natural mechanism for the decline is dilution: a founding lineage of long-lived people marries into a surrounding population of ordinary lifespan, and longevity thins each generation. The naive form of this idea, in which the long-lived fraction simply halves every generation, is testable, and it fails. Halving predicts the ages reaching the surrounding baseline within three or four generations; the recorded series takes a dozen. The observed decline is too slow for raw fifty-fifty dilution.
This negative result is not a refutation but a constraint, and a useful one. A decline this gradual is what one expects if longevity is under some positive selection, or is incompletely dominant, or if the founding clan married substantially within itself for several generations before thinning, that is, if there was a degree of endogamy. Each of these is a live possibility, and each shapes the fuller demographic model set out in the underlying paper. The point for this page is narrower: the shape of the curve is informative enough to rule the simple version out, which is exactly the behaviour of a real signal rather than a decorative one.
5. The reading: a founding lineage in an already-peopled world
Put the pieces together and a coherent picture emerges, the one the wider two-worlds model proposes and this curve was drawn from. The line from Noah enters an already-inhabited Near East as a small, long-lived founding population. Across the generations of Genesis 11 it grows and marries outward, and its extraordinary lifespans fade toward the common human span as its distinct inheritance is diluted, selected against, and absorbed. The two breaks mark the two moments the model names in advance: the flood at Noah to Shem, and the territorial dispersal “in Peleg’s days” at Eber to Peleg. The floor the curve settles onto is not arbitrary; it is the seventy-to-eighty years of Psalm 90 and the hundred and twenty of Genesis 6:3, the span human life still keeps.
Read this way, the long lives are not an embarrassment to be explained away but a datum to be explained, and the same lifespans that puzzle a literal reading become, on the demographic reading, a measurable trace of a real lineage entering real history. The fuller paper carries this further than a web page should, into a demographic projection of the clan’s size and a population-genetic prediction about a Near Eastern Y-chromosome lineage; those are frontier claims, offered as testable and not yet tested, and are kept in the paper rather than asserted here. What belongs here is the modest and sturdy part: the ages make a curve, the curve has structure, the structure survives translation, and it behaves like the record of a process rather than the ornament of a myth.
6. What this page does and does not claim
It does not claim to prove the two-worlds model, nor to have excluded every symbolic reading of the numbers; a deliberate literary pattern remains a bare possibility the data alone cannot kill. It does not settle the chronology, the actual dates, which belong to How We Date Adam. It claims that the ages of Genesis behave like measured quantities: they decline in a fitted two-phase curve, break at two model-specified positions, hold those breaks across two disagreeing textual traditions, resist the lunar-month rescue on the figures the traditions share, and land on the human floor named elsewhere in Scripture. That is a great deal for a set of numbers usually waved away in a sentence, and it is the reason this page asks the reader to trace the curve before judging the text.
Sources and further reading
Ages follow the Septuagint in the reconstruction of Henry B. Smith Jr., “Methuselah’s Begetting Age in Genesis 5:25”, Answers Research Journal 10 (2017), and “The Case for the Septuagint’s Chronology in Genesis 5 and 11”, in Proceedings of the Eighth International Conference on Creationism (2018); Lamech’s figures follow the Masoretic reading Smith judges original, and Shelah’s total is his favoured but tentative reconstruction. Masoretic-tradition Scripture quotations taken from The Holy Bible, New International Version® NIV®. Copyright © 1973, 1978, 1984, 2011 by Biblica, Inc.™ Used by permission. All rights reserved worldwide. Septuagint quotations from Brenton’s translation. Curve fits and the demographic reading from M. D. Russell, “Patriarchal Longevity as Demographic Signal” (working paper, 2026). Sumerian King List figures from the Electronic Text Corpus of Sumerian Literature, t.2.1.1 (Weld-Blundell 444). Scripture also cited: Genesis 6:3; 10:25; Psalm 90:10; Deuteronomy 34:7.