Human Population Growth

"There is more than enough room in the model to allow long periods of time when, because of war or pestilence, the population growth rates were far below the required averages." — Henry Morris, Scientific Creationism, 1974¹

Summary

Henry Morris made the argument famous in 1974.  Start with a handful of people a few thousand years ago, let them grow at a modest rate, and you get today's world.  Start them a million years ago at the same rate and you get a number of people too large to fit in the universe.  So, he concluded, the population data fit a young human race and strain an old one.  He added a second argument:  a long human past means hundreds of billions of dead, and "it is incredible that today there would be so little fossil or cultural evidence of ancient man".¹  Creation Ministries International and the Institute for Creation Research still publish both arguments.^{2 3}

Here is what the arithmetic and the demographic record show.

1. The young-earth arithmetic is right. Eight people in 2348 BC, Ussher's Flood date, become eight billion by 2022 at an average of 0.47% a year. The Septuagint's Flood date of 3298 BC needs 0.39%. Those rates are a fifth of what frontier populations have sustained for centuries: the settlers of French Canada doubled every generation, about 2.4% a year.^{4 5} A skeptic who says eight people could not become billions in 4,400 years is wrong.
2. The argument's key premise is false. Morris judged it "unlikely that the long-time growth rate could have averaged significantly less than (1/2)%".⁶ Every period with data says otherwise. Mainstream estimates put world growth at 0.064% a year from AD 1 to 1750, a stretch of history no one disputes.⁴ Radiocarbon data give 0.04% for most of the Holocene, built from swings more than ten times that size in both directions.⁷
3. Low averages come from crashes, and the crashes are on record. The Black Death cost Europe 30 to 40% of its people.⁴ Central Mexico fell from 6.3 million to 1 million in 57 years.⁴ At 1% a year in good times, a 0.04% average needs a halving every 72 years. The young-earth average can absorb a halving every 132 years. Both fit the kind of history on record. So the average rate cannot tell the two timelines apart.
4. The young-earth model needs its growth front-loaded, and that is where it is tested. Six hundred years at the French Canadian rate gives 14 million people by 1748 BC on Ussher's dates; after that, 0.14 to 0.17% a year reaches the AD 1 estimates. That is demographically possible. Its cost is that, at that rate, the whole world held under a million people until about 1860 BC. Whether Egypt and Sumer allow that is a dating question, settled in Egypt and the Flood Date and The Neolithic Revolution and the Flood Date, not by growth rates.
5. Both sides misuse the numbers. Morris called the 1650 to 1800 rate of 1/3% "the norm" and dated Adam to 6,300 years ago,^{6 6} but at that rate eight people in 2348 BC reach only 8 million by 1800. His own model needs the uneven growth he denied the evolutionist. He also offered "at least 300 million people" by Abraham's day,⁸ more than most estimates for the world of AD 1. On the other side, David Milne's 1984 reply claimed human growth was "negative, zero or near zero over much of times past",⁹ which cannot be true of a species that grew a thousandfold. And a critic who waves the argument away with one plague has not done the sum: at the young-earth average, a 50% loss costs about 150 years of growth.
6. The burial count proves nothing either way. Mainstream demographers estimate about 9 billion births before farming began.^{4 10} Marvin Lubenow, a young-earth author, counted about 4,000 fossil human individuals found by 1976 and called it "an immense amount of material".¹¹ That is one body in 2.3 million. It sounds damning until the loss rate is priced: it fits if a body's odds of surviving halve every 2,400 years. Unburied bones decompose "beyond recognition in 10 to 15 yr",¹² and Ice Age burial was rare.¹³ No one has measured how many prehistoric bodies should survive, so the count tests nothing.
7. Genetics and archaeology agree the old numbers were small. Archaeologists estimate about 1,500 people in all of western and central Europe in the Aurignacian, about 40,000 years ago.¹⁴ The genetic effective size of about 10,000 is a long-run harmonic mean, not a head count, and a short crash barely moves it (Population Bottlenecks). Small ancient populations are what the mainstream model says, and they are why its death toll is billions, not Morris's trillions.

Verdict.  The population argument succeeds as a defence and fails as a proof.  It shows that a young human race is arithmetically and demographically possible, which answers a common skeptical objection.  It cannot show the race is young, because the long-run average depends on crashes that cannot be counted before written records, and the burial argument depends on a survival rate nobody has measured.  Young-earth writers should stop presenting it as evidence for a young earth.  The question it raises is real, and its answer lies in the dating of Egypt, Sumer and the Neolithic.

Part 1: The argument in its own words

Morris, 1974 to 1976

Henry Morris, the founder of the Institute for Creation Research, set out the argument in three places within two years.  His textbook Scientific Creationism opens it with a puzzle:  "If man has indeed been dwelling on this planet for a million years or more, it is strange that only in recent years has population become a problem."¹  He starts "an initial population of two people, the first parents", lets each family have the same number of children, and solves for the family size that reaches the 3.5 billion people of the early 1970s in 100 generations of 40 years.  The answer is about 2.5 children per family, or an average growth "of 1/2 per cent per year", which "is only one-fourth the present rate."¹  The same rates over a million years give a number of people far larger than the number of electrons in the universe, so the evolution model "has to be strained to the breaking point."¹

Two features of this first version matter later.  It starts from Adam and Eve 4,000 years ago, not from Noah's family.  And Morris built slack into it himself:  "There is more than enough room in the model to allow long periods of time when, because of war or pestilence, the population growth rates were far below the required averages."¹  He also conceded that the evolution model "can be modified by various secondary assumptions to fit the known data", while the creation model "fits the data directly".¹

His 1975 article in ICR's Acts & Facts sharpened the premise.  He wrote that "it does seem unlikely that the long-time growth rate could have averaged significantly less than (1/2)%", because "everything that is known indicates the population has steadily increased throughout recorded history."⁶  He took the rise from 600 million in 1650 to a billion in 1800, about 1/3% a year, as "the norm for population growth throughout most of past history",⁶ and from it computed that "the most probable date of human origin, based on the known data from population statistics, is about 6,300 years ago."⁶  Older estimates of ancient populations he dismissed:  "Such guesses are useless, however, because they are based on a discredited model, that of human evolution."⁶  Then he gave the version tied to Genesis:  "The Biblical model for population growth starts with eight people (Noah, his three sons, and their wives) right after the great Flood."  From an Ussher Flood, eight to 3.5 billion in 4,350 years "represents an average annual increase of only 0.44%".⁶  He expected fast growth just after the Flood, and added, "In turn, this means that the rate may have been much less than 0.44% during the long period in between."⁶

The Genesis Record of 1976 put numbers on the early centuries.  Shem, Ham and Japheth had sixteen listed sons;  with as many daughters, the first generation after the Flood was 32 people, "an increase of 533 percent over their original six parents."  Held at that ratio, there were "at least 1,120 mature adults at the time of the Dispersion at Babel", 101 years after the Flood.⁸  By Abraham's call, 367 years after the Flood, eleven generations at "a 500 percent increase" each give "at least 300 million people".⁸  The book's appendix lists "Convergence of population growth statistics on date of Flood" among its evidences for a global Flood.⁸

Batten, 2000, and the bodies

Don Batten's "Where are all the people?" in CMI's Creation magazine is the fullest popular form, still online and updated in 2023.  "With the Flood at about 4,500 years ago, it needs less than 0.5% per year growth."²  He concedes, "Of course, population growth has not been constant," and points to the slow Middle Ages.²  From the Table of Nations he takes an average of 8.53 children per couple in the first two post-Flood generations and gets over 3,000 people within four generations, "a population growth rate of 3.7% per year, or a doubling time of about 19 years."²  He answers the evolutionist's reply directly:  slow growth for most of human history "means that mankind was on the brink of extinction for most of this supposed history.  This stretches credulity to the limits."²

Then the bodies.  A Stone Age of 100,000 years with one to ten million people alive at a time, at 25-year generations, means "4 billion bodies" or "40 billion bodies buried in the earth", and "we would expect the skeletons of the buried bodies to be largely still present after 100,000 years".²  Since they are not, "The number found is more consistent with a 'Stone Age' of a few hundred years, which would have occurred after Babel."²  Morris had made the same point in 1974 with a larger number:  even slow growth to 3.5 billion over a million years means "at least a total of 3000 billion people would have lived and died on the earth".¹

ICR in 2016, and the best modelling

When a 2016 PNAS paper measured prehistoric growth at 0.04% a year from radiocarbon dates, ICR's Brian Thomas replied that "This circular procedure ensures a very slow human population growth," since the rate is a slope against the radiocarbon timescale.  If the slow rate is rejected, "then the old argument for biblical history from human population growth rates still stands tall."^{3 3}

The most careful young-earth work is Robert Carter and Chris Hardy's 2015 computer model in Journal of Creation.  They reject the smooth algebra of their predecessors, because "small populations do not tend to grow in an algebraic manner", and track individuals through marriage, child spacing, twins and death.  They conclude that "the current world population and the size of the Exodus population are easy to account for under most parameter settings."¹⁵  They are candid about the death rates they used:  "we are assuming the probability of death curve was similar then to now", using the 2009 United States actuarial tables.¹⁵  Their model of the Babel population is discussed in The Neolithic Revolution and the Flood Date.¹⁵

So the claim to be tested has three parts.  The average rate since the Flood is modest.  The rate a long human history needs is implausibly low.  And a long history should have left far more bodies.

Part 2: The arithmetic

What a young timeline needs

A population growing at a steady rate r a year multiplies by e^(rt) in t years, so the rate needed is the natural log of the growth factor divided by the years.  The table uses eight people at three Flood dates:  Ussher's 2348 BC, the Masoretic Text's 2518 BC on the Exodus-anchored reckoning, and the Septuagint's 3298 BC, the dates worked out in How Old Does the Bible Say Earth Is?.  The targets are the published estimates for AD 1, which run from 170 million (McEvedy and Jones) to 300 million (Haub), about a billion in 1800, and eight billion in 2022.^{16 10 5}

Batten's "less than 0.5%" and Morris's 0.44% are correct.^{2 6}  Carter and Hardy's 0.464% is correct.¹⁵  Morris's smaller sums check out as well.  His Babel figures reproduce to within a few people, and his 1/3% from two people reaches a billion in about 6,000 years against his 6,100.⁵

Is half a percent a lot?  No.  The French settlers of Canada are the classic case.  The 905 pioneers who married there before 1660 "produced on average 4.2 married offspring", which Massimo Livi-Bacci, an Italian demographer, calls "a doubling of the original population in a single generation".⁴  At a 30-year generation that is 2.5% a year.  The 5,000 settlers who stayed became 6 million French Canadians, about 2.4% a year sustained over three centuries.⁵  The young-earth average is a fifth of that.

What a long timeline needs

The same sum run backward from today gives the mainstream average.  From two people 200,000 years ago to eight billion now is 0.011% a year.  From 10,000 people 50,000 years ago it is 0.027%.  To reach the roughly 6 million people that demographers estimate for 10,000 BC, the rate before farming must average 0.003 to 0.04%.⁵  These are scale figures, not anyone's model:  no demographer says growth was ever steady.

Livi-Bacci's synthesis gives the estimated rates period by period.  Growth "during the 30,000 years leading up to the Neolithic era averaged less than 0.1 per 1,000 per year".  It then "increased to 0.4 per 1,000" from 10,000 BC to the time of Christ, as the population "grew from several million to about 0.25 billion".⁴  His Table 1.2 gives 0.037% a year from 10,000 BC to AD 0, 0.064% from AD 0 to 1750, and 0.594% from 1750 to 1950.⁴  Colin McEvedy and Richard Jones's estimates give 0.03 to 0.07% a millennium at a time from 5000 BC to 1000 BC, then 0.12% to AD 1.^{16 5}

The best measurement of a prehistoric rate comes from radiocarbon.  H. Jabran Zahid, Erick Robinson and Robert Kelly summed about 7,900 radiocarbon dates from Wyoming and Colorado, on the logic that more people leave more datable charcoal.  "For the population of Wyoming and Colorado 6,000–13,000 cal BP, we measure a long-term annual growth rate of 0.041 ± 0.003%."  Other continents gave the same figure, and farmers grew at the same long-run rate as foragers.⁷

So the two timelines need long-run averages about ten times apart:  roughly 0.4 to 0.7% for a young one, roughly 0.04% for an old one.  Morris was right that the old timeline needs a very low average.  The question is whether such an average is strange.

Part 3: Booms, busts, and the price of a crash

Real populations do not grow steadily

Livi-Bacci warns against reading a long-run rate as a description of history:  such rates "are, of course, an abstraction, as they imply a constant variation of demographic forces in each period, while in reality population evolves cyclically."⁴  The cycles are on record wherever there are records.

• The Black Death. Between the years before 1348 and the low point in the first half of the fifteenth century, "a loss of 30 to 40 percent is corroborated by local studies" across Europe, and Europe did not regain its 1340 level until the sixteenth century.⁴ Jean-Noël Biraben's world estimates fall from 443 million in 1340 to 374 million in 1400, a 16% loss for the whole planet.¹⁶
• The Americas after 1492. "Cook and Borah give a figure of 6.3 million Indios in 1548, which subsequently declined to 1.9 million in 1580 and 1 million in 1605."⁴ That is a fall of 3.2% a year for 57 years, after an earlier collapse from 1519 to 1548 whose size is disputed.⁵
• Prehistory. Zahid's radiocarbon curve is not smooth. "These short-term fluctuations correspond to annual rates of growth/decline (averaged over 100 y) that are substantially greater (a factor of >10) than the long-term annual growth rate of 0.04%."⁷ The Neolithic shows the same pattern from skeletons: Jean-Pierre Bocquet-Appel found the share of children among the dead jumping at the switch to farming, which he reads as about two more births per woman, a boom on a world scale.¹⁷

Zahid's team also showed why a steady low rate is unlikely in a population that is not crashing.  Hunter-gatherer survival needs 4.11 births per woman for 0.04% a year, and "a change of 0.04 in the number of births per woman (a 1% increase, i.e., going from 4.11 to 4.15), doubles the annual growth rate from 0.04% to 0.08%".⁷  A near-zero average does not come from people hovering at 4.11 children for 40,000 years.  It comes from good decades and bad ones cancelling out.

That answers Batten's "brink of extinction".²  A population of millions that averages zero growth is not on a brink.  It grows in good years and loses in bad ones, and the losses are regional.  A plague that kills a third of Europe does not touch the Americas.

How often must the crashes come?

The cost of a crash can be priced.  If a population grows at rate g in normal years and loses a fraction s in a crash, the long-run average is g minus the log loss per crash divided by the years between crashes.  Solve for the interval.⁵

Read the table both ways.  The old timeline needs a regional collapse of a third every few decades, or a halving every century or two, if good years run at 0.5 to 1%.  That is the rhythm of the historical record:  famine, war and epidemic every generation somewhere, a Black Death every few centuries.  The young timeline can absorb the same kind of history if its good years run at 1% or more.  It cannot if its good years run at the 0.5% Morris thought normal.  At that rate the slack over 4,369 years adds up to the losses of two or three Black Deaths, and nothing else may go wrong.

So the average rate does not separate the two timelines.  Each can be met by a plausible mix of booms and busts.  Which mix happened before written records cannot be read off today's population.

The young-earth path through history

The young timeline does face a real constraint, and it is the shape of the curve.  The AD 1 to 1750 rate of 0.064% is history, not theory, and the ancient world had its own wars and plagues.  A young-earth model that ran at 0.7% for 2,300 years and then dropped to a tenth of that at the birth of Christ would need a reason for the change, and none is on record.  It has to put its fast growth first, in the centuries after the Flood, when Batten's large families and empty lands apply.

Six centuries at the observed French Canadian rate, then a rate close to McEvedy's 0.12% for the last millennium BC, reproduces the AD 1 estimates from an Ussher Flood.⁵  Demographically, that is possible.  It also commits the model to definite numbers.  At 2.4% a year from 2348 BC, the whole human race passes one million only around 1860 BC and ten million around 1765 BC.  On the Septuagint's date the same milestones fall around 2810 and 2715 BC.

Those dates are where the test lies.  Egypt's Old Kingdom and the cities of Sumer are dated, by king lists and by radiocarbon, to centuries when an Ussher world would hold a few thousand people.  Whether those dates can be compressed is argued in Egypt and the Flood Date and The Neolithic Revolution and the Flood Date.  The growth rate itself decides nothing there.  It only says how many people a given date allows.

Part 4: Where each side misuses the numbers

Young-earth misuses

The steady rate.  Morris's central premise was that the long-run rate could hardly have averaged much below half a percent, because "everything that is known indicates the population has steadily increased throughout recorded history."⁶  That was not true in 1975.  The Black Death and the collapse of the Americas were standard history, and the passage Morris himself quoted put the world before farming at "between five and ten million people".  He set it aside as "based on a discredited model".⁶  Batten is more careful and concedes that growth "has not been constant".²

The 1/3% "norm" contradicts the Flood.  Morris's "most probable date of human origin", about 6,300 years ago, comes from running the 1650 to 1800 rate back to two people.^{6 6}  But on Morris's own account the human race restarted from eight people at the Flood, about 4,150 years before 1800.  At 1/3% a year, eight people in 2348 BC become about 8 million by 1800, not a billion.⁵  The 1/3% figure cannot be "the norm" of a world that began again from Noah's family.  Morris saw the problem in the same article, writing that early growth was fast and later growth "may have been much less than 0.44%".⁶  But then the model is doing exactly what he charged the evolutionist with:  using uneven rates to fit a chosen date.  He called that "arbitrary juggling".⁶  Uneven rates are not juggling for either side.  They are what real populations do.

Three hundred million at Abraham.  The Genesis Record says that eleven generations of fivefold increase would give "at least 300 million people" by Abraham's call, 367 years after the Flood.⁸  Morris hedged the claim, adding that the rate more likely "fell off as time went on".  The number still should not be printed.  A fivefold increase every 33 years is 4.9% a year, sustained for nearly four centuries, double the French Canadian rate.⁵  And 300 million in about 1900 BC is more people than most estimates give for the whole world at the birth of Christ.  If that many people were alive in Abraham's day, the young-earth model would have to explain why the next two thousand years added nothing.

Numbers nobody proposes.  Morris's figure for a million years at 0.5%, which he called impossible, and Batten's 10^43 people at 0.01% a year are correct sums of models nobody holds.^{1 2}  They show that steady growth over long times is impossible.  The mainstream agrees, and has never proposed it.

The circularity charge cuts both ways.  Thomas is right that Zahid's 0.04% is a slope against the radiocarbon timescale, so it assumes that timescale.³  But Morris's 0.44% assumes the Ussher timescale in exactly the same way.  A long-run average on any timescale is computed from that timescale.  It cannot also be evidence for it.  Thomas also quotes the paper slightly wrongly:  his "a few individuals 50,000 years ago" is "a few individuals around 50,000 y ago" in PNAS.⁷  The meaning is unchanged.

Critics' misuses

"Negative, zero or near zero."  David Milne's 1984 reply in Creation/Evolution is the standard rebuttal, and its central point is sound.  Morris committed what Milne calls the "bunny blunder":  snowshoe hares growing at 57% a year in the early 1930s would date the first hares to 1885, yet "Over the long haul, r = 0 for the bunnies".⁹  But Milne then claimed that "human population growth was probably negative, zero or near zero over much of times past", and that plagues and famines "can be considered to be the rule, rather than exceptional occurrences."^{9 9}  For hares that cycle around a fixed number, zero is right.  Humans went from a few million to a billion.  The long-run rate is small, about 0.04 to 0.06%, but it is not zero, and a model where crashes are "the rule" has to explain how that much growth happened at all.

The pyramid sum attacks the weaker version.  Milne's best-known line computes 750 people in the world in 2500 BC and about 150 working men to build the Great Pyramid.⁹  The sum is right:  it reproduces to the person.⁵  But it uses Morris's steady 1/3% from Adam in 4300 BC.  Morris's Flood version, in the same 1975 article, has fast growth after the Flood and slow growth later.⁶  Milne answered the version he could ridicule.  The real challenge, which his pyramid points at, is the young-earth path in Part 3:  how many people an Ussher world holds at a pyramid date, and whether the pyramid date can move.

One parish register.  Milne cited the London parish of St Botolph, where deaths outran births from 1558 to 1625.⁹  One city parish over 67 years is a sample of one place.  A register records baptisms and burials, not the people who moved in and out, so it cannot give even that parish's growth, let alone the world's.

"One plague refutes it."  A loss of a third or a half is a setback of decades to a century and a half, not a refutation.  At the young-earth average of 0.47%, halving the world costs about 146 years of growth.  The Flood models can absorb a few such events, as the table in Part 3 shows.  The argument fails because the average cannot discriminate, not because crashes happened.

Assuming the founders were few.  Milne listed as a flaw that the argument "assumes, without evidence, that the entire world population once consisted of two (or a few) individuals".⁹  Population genetics does not rule out a brief bottleneck of a few people.  Its problem with the biblical bottleneck is its date, not its size, as Population Bottlenecks explains.

Part 5: Where are the bodies?

How many died

Batten's arithmetic is right for his inputs:  one million people for 100,000 years at 25-year generations is four billion deaths, and ten million is forty billion.²  Morris's 3,000 billion is also right for his:  steady growth from two people to 3.5 billion over a million years gives about 3,000 to 5,500 billion deaths, depending on the generation length.^{1 5}  Neither input is the mainstream model.  Morris's million years includes Homo erectus, and his steady growth piles most of the population into the last few tens of thousands of years.

The mainstream estimates are smaller.  Livi-Bacci's table, following Jean Bourgeois-Pichat, puts 9.29 billion births before 10,000 BC, out of about 82 billion in all human history.^{4 4}  The Population Reference Bureau's estimate, starting from two people in 190,000 BC, gives 8.99 billion before 8000 BC out of 117 billion.¹⁰  Both are guesses built on guessed populations, as their authors say.  But they are the numbers a fair test must use:  about nine billion people died before farming, not trillions.

How many have been found

Batten does not say how many have been found.  Marvin Lubenow, the young-earth author of Bones of Contention, did count.  Working through Kenneth Oakley's Catalogue of Fossil Hominids, he found "A grand total of 3,998 (approximately 4,000) fossil hominid individuals" as of 1969 to 1976, counting conservatively.¹¹  His verdict was the opposite of Batten's:  "This is an immense amount of material."¹¹  Two young-earth writers looking at the same record reached opposite impressions, which is a sign that impressions are not evidence here.  Lubenow's count includes australopiths, and many more fossils have been found since 1976, so the true figure for Ice Age humans is of the same order, a few thousand individuals.

Four thousand out of nine billion is one in 2.3 million.⁵

How many should survive

Batten assumes the answer is most of them:  "we would expect the skeletons of the buried bodies to be largely still present after 100,000 years".²  His only support is that some older bones have survived, citing dinosaur blood.  Every step of that assumption fails on the evidence.

• Most of the dead were not buried. Batten says "people buried their dead". Formal burial is rare in the Ice Age record. Even at its end, Jörg Orschiedt writes, "Burials of the Late Palaeolithic (14 000– 11 600 cal years before present, henceforth BP) are a rare phenomenon in Europe."¹³
• Exposed bones vanish in years. Anna Behrensmeyer tracked carcasses of known age in Kenya's Amboseli Basin: "Most bones decompose beyond recognition in 10 to 15 yr."¹² A body that is not buried quickly, in the right sediment, is gone within a human lifetime.
• The living population was tiny and thin on the ground. Isabell Schmidt and Andreas Zimmermann estimate "a mean of 1,500 persons (upper limit: 3,300; lower limit: 800) for western and central Europe" in the Aurignacian, about 42,000 to 33,000 years ago.¹⁴ Over 9,000 years at 25-year generations, that is about 540,000 deaths spread over 1.5 million square kilometres.⁵

The recovery rate the mainstream model implies can be priced.  If loss were steady, one body in 2.3 million surviving from an average age of 50,000 years means a body's odds of surviving halve every 2,400 years.  At 20,000 years it is every 950 years;  at 100,000 years, every 4,700.⁵  For bones that mostly rot in a decade unless buried in a cave or a flood plain, a half-life of a thousand years or more is generous, not stingy.  The real loss curve is not steady:  most bodies vanish at once and a few, in caves and sediments, last indefinitely.  That makes the mainstream recovery rate easier to meet, not harder.

The young-earth alternative does not escape the same sum.  Batten's post-Babel Stone Age of "a few hundred years" still needs a population large enough to reach millions by Abraham.  Growing from eight people in 2348 BC to McEvedy's 27 million by 2000 BC means about 21 million deaths in those three and a half centuries.  Four thousand found is one in 5,200.⁵  That is still almost all of them gone.  Both models need most bodies to vanish.  The count of fossils measures the preservation rate, which nobody has measured independently, not the age of the human race.

Batten adds that "even if the bodies had disintegrated, lots of artefacts should still be found."²  They are.  Stone tools survive where bones do not, and some sites are thick with them:  a 2014 study of Kathu Townlands in South Africa reports "remarkably high artefact density" in an Earlier Stone Age deposit.¹⁸  The abundance of stone tools is a problem the young-earth Stone Age of a few centuries has to answer, not one it can pose.  This article does not price that record;  it is a fit subject for its own.

Part 6: What genetics and archaeology say about ancient numbers

Three independent lines say the human population before farming was small, and none of them depends on Morris's question.

1. Archaeology. Site counts give about 1,500 people in Aurignacian western and central Europe, about one person per 1,000 square kilometres.¹⁴ Livi-Bacci, following Biraben, puts the world before the Upper Paleolithic at no more than several hundred thousand and the world at 10,000 BC at about 6 million.^{4 4}
2. Radiocarbon. Zahid's team note that their own rate, run back from a billion people in 1800, "retrodicts a global human population of a few individuals around 50,000 y ago", and they add that the population then was larger, so the rate before the Holocene must have been lower still.⁷ That is the mainstream conceding, in its own numbers, that ice-age growth was close to zero on average.
3. Genetics. The human effective population size of about 10,000 is a long-run harmonic mean of breeding numbers, not a census. A short, deep bottleneck barely moves it, which is why genetics cannot rule out a small founding group. Its difficulty for the biblical account is the date of any such bottleneck, not its size. The details are in Population Bottlenecks and Adam, Eve, and Human Genetic Diversity.

These small numbers cut both ways.  They are why the mainstream death toll before farming is billions, not Morris's trillions, which weakens the burial argument.  They also mean the mainstream model is the one that needs the near-zero average Morris found incredible.  The record of crashes in Part 3 is the evidence that such averages are ordinary.

Objections and responses

"Eight people could never become eight billion in 4,400 years."  They could, at 0.47% a year, a fifth of the rate French Canada sustained for three centuries.  This skeptical objection fails, and skeptics should drop it.

"Plagues prove the young-earth calculation wrong."  A plague is a setback, not a refutation.  At the young-earth average, halving the world costs about 146 years.  With good years at 1% or more, a young-earth history can absorb a halving every 130 years.  What plagues refute is Morris's steady rate, not a young human race.

"The mainstream needs humanity on the brink of extinction for 200,000 years."  It needs a long-run average near zero, which is not the same thing.  A population of hundreds of thousands to millions, spread over continents, can average zero by booming and crashing regionally, as Zahid's radiocarbon swings and every documented history show.  Batten's "brink" reads a near-zero average as a near-zero population.

"Zahid's 0.04% is circular, because it assumes the radiocarbon timescale."  Yes, and Morris's 0.44% assumes the Ussher timescale.  Both are averages computed on an assumed timescale, and neither is evidence for its own.  But the ratio Zahid found, short-term swings more than ten times the long-run rate, does not depend on which timescale is right.

"Then the argument is a draw, so young-earth creationists may keep using it."  A draw is not evidence.  Morris claimed the creation model "fits the data directly" and the evolution model must be "strained".^{1 1}  Both models fit the average once crashes are allowed, and Morris's own Flood version needs uneven rates.  The argument should be used only as a defence:  a young timeline is demographically possible.

"The fossil record is too small for a long human history."  It is about one in 2.3 million of the estimated pre-farming dead, which needs a survival half-life of roughly 1,000 to 5,000 years per body.  Unburied bones last ten to fifteen years.  And a young-earth Stone Age needs most bodies to vanish too:  one in 5,200.  The count measures preservation.

"Lubenow's count is old and the fossil record is now much larger."  True, and that helps the mainstream case:  a larger count means a higher survival rate, which makes a long history easier to fit.  It does not help a short one.

"Morris's 300 million at Abraham shows how easy fast growth is."  It shows the opposite.  It needs 4.9% a year for 367 years, twice any sustained rate on record, and it overshoots the estimates for AD 1.  Young-earth writers should not repeat it.

"Ancient population estimates are all guesses."  Before written censuses, yes, and Livi-Bacci says so.⁴  That cuts against both sides equally.  The rates from AD 1 to 1750 rest on far better data, and they are about a seventh of the young-earth average.

Issues and unanswered questions

1. No measured survival rate for prehistoric bodies. The burial argument can only be settled by a taphonomic loss curve for human remains over tens of thousands of years, by environment. None was found for this article. Surovell and colleagues have published loss curves for archaeological sites from radiocarbon dates, which could be adapted; they were not obtained.
2. No modern count of Pleistocene human individuals. Lubenow's 4,000 is from catalogues closed in 1969 to 1976. A current figure, separating australopiths from Homo and Ice Age Homo sapiens from older humans, would sharpen Part 5 in either direction.
3. The young-earth path has not been modelled against history. Carter and Hardy's model uses modern United States death rates and has no famine, war or epidemic.¹⁵ No young-earth author has yet fitted a growth curve from the Flood through the ancient population estimates with realistic mortality and crashes. Part 3 sketches one; a full model is open work for either side.
4. Morris's 1966 technical paper was not read. "World population and Bible chronology," Creation Research Society Quarterly 3(3):7-10, gives the fuller formula Batten cites. Morris's 1974 to 1976 books and article state the same argument.
5. Highest sustained growth on record. Hutterite fertility is the highest recorded in a regular population, according to Livi-Bacci.⁴ Their growth rate was not obtained from a primary source for this article; French Canada stands in. If a population has sustained 4 to 5% a year for centuries, Morris's Abraham figure would need revisiting.
6. The early post-Flood world was an ice age on the young-earth model. Whether frontier growth of 2 to 4% a year is plausible across the centuries a young-earth model assigns to the Ice Age is not addressed by any source found.

Verdict

The population argument is sound arithmetic built on a false premise.  Eight people can become eight billion since any biblical Flood date at under half a percent a year, a fifth of what real frontier populations have sustained.  That answers the skeptic who calls a young human race demographically impossible.

It does not show that the human race is young.  Morris's premise, that the long-run rate could not have averaged much below half a percent, is contradicted by every period with records, including the 1,750 years after Christ.  Long-run rates are the small residue of large booms and crashes, and either timeline can be met by a plausible mix of them.  Morris's own Flood version needs exactly the uneven growth he called "arbitrary juggling" in others.

The burial argument fails for the same reason.  It compares a count of fossils with a count of the dead without a measured survival rate.  The mainstream model needs about one body in 2.3 million to survive, the young-earth Stone Age about one in 5,200, and unburied bones last a decade.

What the numbers do settle is narrower.  A young-earth history must put fast growth first.  At the French Canadian rate, that leaves the world with under a million people until about 1860 BC on Ussher's dates, or about 2800 BC on the Septuagint's.  That claim can be tested, but by the dating of Egypt, Sumer and the Neolithic, not by growth rates.

Sources