The Neolithic Revolution and the Flood Date

"This tells us that the difference, in calendar years, between the two timbers is at least 132 years, and suggests that the life span of level VII and part of VIB is likewise at least 132 years." — Peter Kuniholm and Maryanne Newton, counting tree rings at Çatalhöyük, 1997¹

"It is impossible at this point in time to give an accurate assessment of the time taken for the build-up of these strata." — A. J. M. Osgood, the only creationist to attempt one, 1986²

"it was the archaeologists themselves who delivered and subsequently cultivated the now popular opinion that Göbekli Tepe is the site of the »World's First Temples«." — Lee Clare, head of the Göbekli Tepe project, 2020³

Summary

Genesis puts Noah's landing in the mountains of Ararat and Babel in Shinar:  eastern Anatolia, the northern Levant and Mesopotamia.  That is also where archaeology finds the earliest monumental architecture, the earliest villages, the earliest bread, and the first domesticated sheep, goats, cattle and wheat, dated from about 12,500 BC to 4000 BC.  Every Flood date Genesis allows falls between 3300 and 2350 BC.  The gap is six to ten thousand years.

1. The gap is six to ten thousand years, and no manuscript choice touches it. Wall plaster between two pillars of Enclosure D at Göbekli Tepe dates to 9745 to 9314 cal BC,⁴ 6,800 to 7,200 years before a Masoretic Flood at 2518 BC. The Septuagint's 780 extra years close about 11% of that. Bread at Shubayqa 1 in Jordan is 14,400 years old on seven short-lived seeds from inside two hearths.⁵
2. Occupation runs through every proposed Flood horizon without a break. Jericho "records over 10,000 years of human history … without any apparent significant breaks in occupation".⁶ On the excavator's own period table a Masoretic Flood lands inside Early Bronze IIIB and Ussher's inside Early Bronze IVA, with about 8,000 years of stratified tell beneath.⁷
3. The tell-depth objection fails, and this article concedes it. Osgood's mechanism, mud-brick houses collapsing every 7.5 to 15 years, builds a 21 m tell in 200 to 800 years on the computation below. Anyone arguing from depth alone is wrong.
4. What does not fail is the count of building levels, and it uses no radiocarbon. Çatalhöyük houses "were lived in for 50-100 years",⁸ so seventeen superimposed levels is 850 to 1,700 years. Tree rings agree: two timbers from levels VII and VIB are 132 rings apart, so those levels alone span "at least 132 years".¹ Osgood allows 330 years for the whole world's Stone Age.²
5. Superposition forbids the contemporaneity escape. Neighbouring cultures overlap by a few centuries, as Osgood says. Periods stacked vertically in one column at Jericho cannot be contemporary with each other.
6. Domestication is a rate, not only a date. Goats managed at Ganj Dareh, directly dated to 9,900 BP, "show no evidence of size reduction or any other domestication-induced morphological change".⁹ Domestic form appears 500 to 1,500 years later.
7. The creationist side has four dates for Göbekli Tepe and no method for any. Habermehl 2015 gives "around 2500 BC (LXX) or 2100 BC (MT)";¹⁰ Habermehl 2018 puts the whole site into "as little as 25 years";¹¹ Answers in Genesis gives "about 2150–1900 BC";¹² Cox gives "up to and possibly after Babel".¹³
8. The one creationist who read the radiocarbon paper quoted its caveats and left out its answers. The Dietrich sentence Cox quotes about data that "fail to provide absolute chronological points of reference" is about pedogenic-carbonate dates the authors say ran too young, and the wall-plaster date of 9984 ± 42 follows on the same page.⁴
9. The excavators have retracted the two things creationists cite most, and both retractions cut the wrong way. Intentional backfilling at Göbekli Tepe is withdrawn, and the buildings' final phases are "in excess of five centuries after their initial foundation".³ Habermehl allows 25 years.
10. Some of the creationist critique is simply right, and one mainstream citation fails here too. The "first temple" reading was wrong, Kenyon's fortress is read by Bar-Yosef as flood defence, and Halaf, Samarra and Ubaid II do overlap. Hodder's much-quoted claim that Çatalhöyük's plaster coats are annual cites a chapter that says nothing about plaster.¹ Interpretation is unstable. The dates are not, and every mainstream disagreement here is worth a few centuries against a gap of six to ten thousand years.

Verdict.  The Near Eastern Neolithic cannot be fitted between any Genesis Flood date and Abraham.  The failure is not radiocarbon.  It is counted building levels, measured house lives, counted tree rings, stacked periods in single columns, and a domestication lag that is a biological rate.  The one creationist who tried to give a number said in print that he could not.

Part 1: Five sites and how each is dated

The Flood dates used throughout are fixed at a common Abraham anchor:  2518 BC for the plain Masoretic sum, 2348 BC for Ussher, and 3298 BC for the Septuagint.  Where a creationist source uses its own figure, that figure is used for that source and said so.

Göbekli Tepe

Göbekli Tepe is a hill in southeast Turkey, about 15 km from Şanlıurfa, with rings of carved T-shaped limestone pillars up to 5.5 m tall.  The project's own summary gives its span as about 9500 to 8000 cal BC.³  The date that matters most is structural.  In 2013 Dietrich and colleagues dated organic material in the wall plaster between pillars 41 and 42 of Enclosure D:  9984 ± 42 ¹⁴C BP, calibrating to 9745 to 9314 cal BC at 95.4%.⁴  Plaster on a standing wall dates the building, not whatever slid into it later.  Three charcoal samples from Enclosure D at bedrock cluster at 9664 to 9311 cal BC.⁴

Against a Masoretic Flood at 2518 BC the wall plaster is 6,800 to 7,200 years too early.  Against the Septuagint's 3298 BC it is 6,000 to 6,450 years too early.  The Septuagint's extra 780 years buy about 11% of the gap.

The interpretation of the site has changed a great deal since 2013.  The dates have not.  Part 5 sets out what changed.

Shubayqa 1

Shubayqa 1 is a Natufian hunter-gatherer camp in the Black Desert of northeast Jordan.  In 2018 Arranz-Otaegui and colleagues identified 24 charred food remains from two stone fireplaces as bread-like products made from wild einkorn and club-rush tubers.  "Seven radiocarbon dates of short-lived charred plant remains from within the fireplaces indicate their use around 14.4–14.2 ka cal BP, which corresponds with the early Natufian period."⁵  The site as a whole runs from 14.6 to 11.6 ka cal BP.⁵

This is the extreme case, and it is dated on the material every laboratory prefers.  Short-lived seeds cannot carry an old-wood offset, and seven of them from inside two hearths agree.  14,400 cal BP is 12,450 BC, about 9,930 years before a Masoretic Flood and 9,150 before a Septuagint one.  The same paper says bread "were probably not routinely consumed foodstuffs or dietary staples during the Natufian.  The exploitation of cereals increased gradually between 11.5 and 9 ka cal BP".⁵  A 2,500-year gradual change is itself hard to compress.

Jericho

Tell es-Sultan at Jericho is a 21 m mound in the Jordan Valley, 250 m below sea level.  Bar-Yosef describes the Pre-Pottery Neolithic A tower from Kenyon's sections:  "The tower, placed inside the perimeter of the settlement, was 8.2 m high (ca. 9 m in diameter at the base and 7 m at the top) and built of undressed stones.  It had a staircase leading to the top with 22 steps built of dressed slabs."¹⁴  It weighed about 1,000 tons and had to be repaired after subsiding into the marl.¹⁴

Bar-Yosef's Table 1 lists fifteen radiocarbon determinations tied to numbered construction stages in Kenyon's Trench I.¹⁴  His summary:  "The available ¹⁴C dates indicate that the construction occurred between 8300 B.C. and 7800 B.C., which may mean around 8000 B.C. as estimated by Kenyon."¹⁴  Those are uncalibrated years in the old "bc" convention, where bc is the radiocarbon age minus 1950, and calibrating them makes Jericho older, not younger:  8300 ± 200 bc is 10,250 ± 200 ¹⁴C BP, which on IntCal20 spans about 10,665 to 9,340 cal BC at 95.4%.  That was computed here against the IntCal20 curve file, not taken from a source.  The current excavator's calibrated table gives the Late Natufian at 10,500 to 8800 BC, PPNA at 8800 to 7500 BC and PPNB at 7500 to 6300 BC.⁷

That table is the clearest exhibit in this article.  It runs period by period from the Late Natufian to the Iron Age.  A Masoretic Flood at 2518 BC falls inside Sultan IIIc2, Early Bronze IIIB.  Ussher's 2348 BC falls inside Sultan IIId1, Early Bronze IVA.  The Septuagint's 3298 BC falls inside Sultan IIIa2, Early Bronze IB.  In each case the tell already carries seven to eight thousand years of stratified deposit beneath the Flood horizon, and a year-long global Flood would have to leave a sterile layer and a population break somewhere in that column.  A 2023 isotope study by a Max Planck and LMU team, with no stake in this argument, says the site was occupied "without any apparent significant breaks in occupation".⁶

One honest complication belongs here.  Bar-Yosef reports a gap at the very bottom:  "following a gap of nearly a millennium (ca. 9200–8350 B.C.), the site was settled by one of the early farming communities."¹⁴  Wang's "without any apparent significant breaks" is a summary sentence citing Kenyon's volumes.  The two conflict at the Natufian-to-PPNA transition, about 6,000 years before any Flood date.  Both are reported here because citing only the one that helps would be dishonest.

For Late Bronze Jericho and Joshua, see Jericho and the Conquest.  This article owns the Neolithic tell 7,000 years below.

Çatalhöyük

Çatalhöyük on the Konya plain in central Turkey is two mounds.  The East Mound is the famous one.  Orton and colleagues summarise its chronology:  "Dating has concentrated on the higher southern eminence, where a 21m sequence divided into 'levels', XII (earliest) to 0 (latest), by Mellaart (1964), with the later addition of four 'pre-XII' levels (Cessford 2001), probably represents almost the entire duration of occupation. … Mellaart (1978: 13) estimated an occupation of c. 7100–6300 BC.  While his start date has proved remarkably accurate (Bayliss et al. 2015), subsequent research has pushed the undated final levels steadily towards the end of the seventh millennium".¹⁵  Reporting Marciniak's 2015 Bayesian model, Orton gives the end:  abandonment of the last known house "is now securely dated to 6015–5905 BC at 95% probability… while cessation of occupation is estimated as 5975–5865 BC at 95% probability".¹⁵  The West Mound then overlaps the East by up to two centuries and runs on to about 5600 BC.¹⁵

So the East Mound is 21 m of stratified houses in about 17 numbered levels, spanning roughly 7100 to 5900 cal BC.  Its last occupation ends 3,350 to 3,460 years before a Masoretic Flood;  its first level is 4,580 years before one.  Bayliss 2015 and Marciniak 2015, the papers behind those dates, are closed, so the start and end are cited as Orton reports them.

Abu Hureyra, and the domestication lag

Abu Hureyra on the Euphrates in Syria runs straight through the transition from hunting to herding.  Smith and colleagues' 2022 dung-spherulite study gives the phase table on IntCal20:  Epipalaeolithic 1A at 13,300 to 12,800 cal BP, through the Intermediate Early Neolithic at 11,400 to 10,600, to Pre-Pottery Neolithic B phases 2A and 2B and a Late Neolithic 2C ending at 7,800 cal BP.¹⁶  Spherulite concentrations "dropped abruptly during Neolithic 2B (9,300–8,000 calBP) and 2C (8,000–7,800 calBP), when sheep/goat herding surpassed gazelle hunting".¹⁶  That is 5,500 years of occupation ending about 5850 BC, with burned dung as an independent proxy for the herding chronology.

The domestication evidence is a rate as well as a date.  Zeder's synthesis reads herd management from the age and sex of slaughtered animals.  "Application of this approach … has identified the clear signature of a managed herd of goats (harvesting of young males and prolonged survivorship of females) at the site of Ganj Dareh in highland Iran.  Directly dated to 9,900 B.P., the goats from this site show no evidence of size reduction or any other domestication-induced morphological change."⁹  Management of all four livestock species begins about 11,000 to 10,000 BP, but "Clear-cut morphological responses to domestication … are not evident in these four livestock species until ca. 9,500–9,000 B.P."⁹  Plants show the same shape:  intensified management "at least as far back as ca. 12,000 B.P.", with nonshattering seed heads "not well established until ca. 10,500 B.P."⁹  Goats carry four to six distinct domestic mitochondrial lineages, cattle five.⁹

The caveat a creationist would raise is fair.  The management dates rest on harvest profiles, an inference from age-at-death curves, not a physical clock.  What survives is the lag itself, a thousand years or more between keeping an animal and seeing it change, which a model compressing the Neolithic into a few centuries has to compress too.  The farm-fox experiment is taken up under objections.

Mesopotamia before 2000 BC

Osgood's Part 2 places the whole Mesopotamian Chalcolithic, Hassuna through Jemdet Nasr, between Babel and Abraham.  The standard reference on the Ubaid states:  "Radiometric data suggest that the whole southern Mesopotamian Ubaid period, including Ubaid 0 and 5, is of immense duration, spanning nearly three millennia from about 6500 to 3800 b.c."¹⁷  The southern Ubaid therefore ends 500 years before a Septuagint Flood and 1,300 before a Masoretic one, and begins 2,700 years earlier.  The eclipse-anchored first millennium is a separate article, Mesopotamian Absolute Chronology.

Part 2: What the creationist literature says

Osgood 1986: the only model, and its own number

A. J. M. Osgood's "A Better Model for the Stone Age" is the one sustained creationist attempt to fit the whole Stone Age into post-Flood time.  He dates Babel to 2200 BC and Abraham's entry into Canaan to 1870 BC.  The consequence is stated as a number:  "Consequently, on the biblical model, any other civilization outside of Mesopotamia, with the possible exception of small areas along the route between Ararat and Sumer, would have to be dated from 2,200 B.C. until 1,870 B.C. for a stone age period, that is, for all that is embraced within the Paleolithic to the end of the Chalcolithic, wherever such a relative chronology can be applied satisfactorily."²  For Palestine he repeats it:  "from 2,200 B.C. to 1870 B.C. or 330 years."²

He does not re-date radiocarbon.  He rejects it:  "Carbon-14 (or radiocarbon) dating in particular assumes that the influx and outflow of carbon-14 atoms into and out of the biosphere is in equilibrium.  This simply is not so, and that alone invalidates the method."²  Dendrochronology "is so shot through with assumptions that it is surprising that anyone dared to present it as a scientific test."²  In place of dates he offers a model in which "The Paleolithic, Neolithic and Chalcolithic could well be contemporary, and might simply be an indication of the different conditions and the different environment and distance from the centre point available to each of the different cultures."²  And a deposit-rate argument, the one quantitative move in the paper:

"The wet conditions may also explain the very large number of stone-age, particularly Neolithic strata, in such places as Mersin, Catal Huyuk and Jericho, where the main building materials were sun-dried mud bricks.  In north-eastern Iraq the Jarmo expedition found that the average expectation for a 'casually built house with some dried mud bricks and mud finished roof' was only 15 years.  In much wetter conditions of earlier days the life of a building may well have been considerably shorter, even half that time, making rapid build up of strata with rebuilding of levels in tells a very highly likely proposition."²

And then, in the next paragraph, the admission:  "It is impossible at this point in time to give an accurate assessment of the time taken for the build-up of these strata.  Long periods of time that have artificially been assigned to them simply cannot be sustained on any present evidence."²

Osgood wrote in 1986, nine years before Schmidt began digging at Göbekli Tepe.  His contemporaneity evidence is real and correctly reported:  Moore's four-stage Levantine Neolithic has Halaf replacing Neolithic IV, and Jasim's Tell Abada shows Halaf, Samarra and Ubaid II together.  That shows neighbouring regional cultures overlapping by a few centuries.  It cannot show that a culture 6,000 radiocarbon years older is contemporary with one 6,000 years younger;  Part 4 explains why.

Part 2 of his paper adds a population figure, "approximately 12,500 male people alive at the time of Babel (excluding deaths from wars and disease) or a total of around 25,000 human beings",¹⁸ and the rule "All cultures outside of lower Mesopotamia should be POST Babel, that is, after 2,200 B.C."¹⁸  Osgood puts Babel about 100 years after the Flood.²  At exactly that point Carter and Hardy's 2015 creationist population model reaches more than 1,000 people only under favourable settings, and 40,000 only under what they call "not likely scenarios".¹⁹

CMI 2011 and ICR 2012: assertion without a number

Creation Ministries International's first answer to Göbekli Tepe was a 2011 feedback reply by Lita Sanders and Robert Carter.  Its dating argument is one sentence:  "Our response to that is the same as the dates that put the earliest Egyptian pyramids before the biblical date for the Flood and those that claim dinosaurs died out 65 million years ago— the dates are wrong."²⁰  The positive claim is "a tentative dating would put it soon after the Flood",²⁰ and the argument is plausibility:  "hunter-gatherers don't build great big monuments, because they're too busy, well, hunting and gathering."²⁰

They deserve credit on that last point.  Clare's 2020 report describes round-oval domestic structures with hearths and bead production on bedrock, and "a dense domestic occupation at the site in the EPPNB".³  The site was not built by people who only visited.  But the correction went the other way:  Göbekli Tepe is now read as a settlement occupied for well over a millennium, which is worse for a short compression, not better.

Brian Thomas's 2012 ICR piece makes no chronological claim and does not mention radiocarbon.  It says "it makes sense that Gobekli Tepe was one of the first building sites of post-Flood peoples."²¹  Nothing by Clarey or Snelling on Göbekli Tepe, PPNA Jericho or the Neolithic radiocarbon was found on icr.org.

Cox 2021: the one who read the paper

Gavin Cox's 2021 CMI article is the only creationist piece found that engages the Göbekli Tepe radiocarbon series directly, and he quotes four of Dietrich's caveats verbatim.  His conclusion:  "When all these dating issues for Göbekli Tepe are assessed without the baggage of uniformitarian thinking, this Neolithic site fits comfortably within the historical constraints of Genesis as a post-Flood settlement."¹³

An audit of his article against the Dietrich PDF finds three things, each checkable.

A units error.  Cox writes that the samples "gave a range of dates of 9,990±30 years to 8,880±60 years BC".¹³  Dietrich's table gives those figures as radiocarbon years before present, not calendar years BC.  The paper prints "9984 ± 42 ¹⁴C-BP (9745-9314 calBC at the 95.4% confidence level)" on p. 36 and "8880 ± 60 ¹⁴C-BP, 8241-7795 calBC" on p. 38.⁴  This is a slip, not an argument, and it does not help him:  reading 9990 BP as 9990 BC makes the site look older, not younger.  His count of eleven samples and his choice of the highest and lowest values both match Table 1 exactly.

Selective quotation at the decisive point.  Cox quotes Dietrich:  "the data fail to provide absolute chronological points of reference for architecture and strata".¹³  On p. 36 of Dietrich, "the data" in that sentence are the pedogenic-carbonate dates, not the radiocarbon on organic samples.  The sentence before it says "absolute ages are clearly too young".  Two sentences later comes "A far better source of organic remains for the direct dating of architectural structures is the wall plaster used in the enclosures", and then the wall-plaster date of 9984 ± 42.⁴  Cox introduces the quotation as being about "carbon dates for layer II", which is the wrong data, and neither the correction nor the wall-plaster date appears anywhere in his article.  In fairness he treats the pedogenic-carbonate method in its own later section and quotes its variable-rate caveat there, so this reads as a misplacement rather than a concealment.

The backfill caveat has since been withdrawn by the project.  Cox's inverse-stratigraphy argument rests on Dietrich's statement that the fill "poses severe problems for the dating of this layer using the radiocarbon method, as organic remains from the fill-sediments could be older or younger than the enclosures".¹³  Clare 2020 replaces intentional backfilling with slope-slide inundation and reports that mud-mortar dates from the standing fabric "have proven far more homogeneous than those made previously on samples from their fill".³  Cox could not have known this in 2021.  Any article written now has to say it.

Cox also places Babel at "c. 140–200 years after the Flood" and, elsewhere in the same article, "around 2,500 BC".¹³  Together those put his Flood at about 2650 to 2700 BC, roughly 150 years before the plain Masoretic sum used here.  He gives no Flood date directly.  His calibration-plateau point is real and belongs to the radiocarbon article.  A plateau widens a calibrated range by decades to a couple of centuries.  The gap here is six to seven thousand years.

Habermehl 2018: telescoping, and the number it produces

Anne Habermehl's 2018 ICC paper is the most serious creationist treatment of Göbekli Tepe in print.  She writes that "Scholars consider this ancient date of Gӧbekli Tepe to be solid, with carbon dating backing up their other dating".¹¹  She accepts the conventional ordering and maps it onto biblical time non-linearly, with conventional years running faster the further back one goes.  Her result:  "It is proposed that the Gӧbekli Tepe site was most likely founded somewhat more than 100 years before Abraham's visit to Egypt (Masoretic timeline) or, alternatively, about 250 years before Abraham's visit to Egypt (Septuagintal timeline)."¹¹  On the Masoretic figure it "was built about 210 years after Babel".¹¹

The claim to test is this one:  "The two thousand conventional years claimed for the time that Gӧbekli Tepe was in process of being built, from its founding to its abandonment, becomes a very small number on our biblical timeline.  From Fig. 3 (MT), we see that there is as little as 25 years allowed for the building of the entire Gӧbekli Tepe site before its abandonment.  Figure 4 (LXX) shows about 50 years."¹¹  The same map places "The Babel dispersion … about 1.76 million years ago on the conventional timeline."¹¹  She is candid about the method:  "placement of these data points on the conventional timelines is approximate, and therefore the conclusions that are drawn from these figures are not precise.  This is because we do not have the necessary data for precision."¹¹

This is the strongest creationist move available, and it is falsifiable.  It does not need radiocarbon to be broken.  It needs 25 to 50 years to be enough to build Göbekli Tepe.  Two years after she wrote, the project reported from mud-mortar samples in the standing fabric that the enclosures' "final building phases are consistent with a construction date around the mid-ninth millennium calBC, which is in excess of five centuries after their initial foundation."³  Her model allows 25 years;  the project measures more than 500.  The price of her map is every hominin date at once:  if Babel is 1.76 million conventional years ago, Acheulean handaxes, Neanderthals and Göbekli Tepe all fall inside about two centuries of real time.

Her own 2015 site report, written after a visit with Lee Clare, gave a different answer:  "the beginning time of building Gӧbekli Tepe falls somewhere around 2500 BC (LXX) or 2100 BC (MT)."¹⁰

Answers in Genesis 2023: a ratio without a method

The only place Answers in Genesis prints a date for Göbekli Tepe is a 2023 review of Graham Hancock's Netflix series.  "Conventionally dated between 9500 and 8000 BC, which Hancock seems to agree with, they can be dated to about 2150–1900 BC in a biblical timeline (early post-Babel)."¹²  Karahan Tepe, 46 km east, "is dated to roughly the same time, 9500–8,000 BC (and the same biblical timeframe of 2150–1900 BC)."¹²  No reasoning is given.  What the numbers supply is a compression ratio:  1,500 conventional years onto 250 biblical ones, a factor of six.  Habermehl's is a factor of 30 to 60.  No primary excavation report with a radiocarbon series was found for Karahan Tepe.

Four dates from three authors on two ministry websites, none derived from a sample, none agreeing with another.  That is the state of the creationist literature on the hardest Neolithic site.

Part 3: The depth test, computed here

Can a 21 m tell form in a few centuries?  Yes.

Osgood's mechanism is that mud-brick houses fall down and are rebuilt every 7.5 to 15 years.  Bar-Yosef, no creationist, says the same about how tells form:  the deep stratigraphies of Mureybet, Abu Hureyra, Jericho, Ain Ghazal and the rest were "mainly formed by the natural collapse and subsequent rebuilding of adobe or mud-brick houses."¹⁴  He also grants Osgood's wet climate:  "it seems that both PPNA and PPNB periods had more favourable climatic conditions than today… Geomorphological evidence supports this general picture and indicates that the 7th millennium B.C. was wetter than the previous or following ones".¹⁴  What Bar-Yosef supplies that Osgood never did is a thickness:  "while the height of PPNA deposits in other parts of the mound is 3.5–4.5 m, in the area of the tower the total thickness amounts to 8.5 m (including the so-called Proto-Neolithic)."¹⁴

Computed here

The script research/deposit-rate-test.ts beside this article takes every input from a published measurement.  Its output, condensed to the lines the text uses:

1. MEASURED ACCUMULATION
   Catalhoyuk East: 21 m over 1180 yr (7100-5920 cal BC)
     = 17.8 mm/yr;  17 levels = 1.24 m and 69.4 yr per level
   Jericho PPNA 3.5-8.5 m over 1300 yr = 2.69-6.54 mm/yr

2. INSIDE OSGOOD'S 330-YEAR WINDOW (Catalhoyuk East alone)
     21 m / 330 yr = 63.6 mm/yr (3.6x measured); 19.4 yr per level

3. CAN OSGOOD'S MECHANISM DELIVER THE VOLUME?  (Yes.)
   7.5 yr house life, 0.4 m debris -> 53.3 mm/yr; 21 m in 394 yr
   7.5 yr, 0.8 m -> 106.7 mm/yr; 21 m in 197 yr
   15 yr, 0.4 m -> 26.7 mm/yr; 21 m in 788 yr
   15 yr, 0.8 m -> 53.3 mm/yr; 21 m in 394 yr
   VERDICT: raw deposit thickness is NOT the binding constraint.

4. THE BINDING CONSTRAINT: COUNTED LAYERS, NOT THICKNESS
   Measured house use-life at Catalhoyuk: 50-100 yr, not 7-15.
     17 levels x 50 yr = 850 yr;  x 100 yr = 1700 yr
   Tree rings: CTL-16 and CTL-6 are 132 rings apart, so level
     VII plus part of VIB alone span >= 132 yr.  Two of 17 levels.
   One 3 cm wall fragment carries 80 plaster couplets, but their
     periodicity is not established, so no year count from them.

5. SIMULTANEITY
   Gobekli Tepe 1500 + Jericho 6200 + Catalhoyuk East 1180
   + Catalhoyuk West 500 = 9380 yr of occupation.

If each collapse of a 2.5 m mud-brick wall spreads 0.4 to 0.8 m of debris over its own floor, and houses really do fall every 7.5 to 15 years, the tell reaches 21 m in 200 to 800 years.  Osgood's 330-year window sits inside that range.  Anyone who argues that Neolithic tells are simply too deep to be young has not done the arithmetic and is wrong.

One caution on the inputs.  The 15-year house life is the Jarmo excavators' figure, which Osgood halves for wetter conditions;  the 50 to 100 year figure in step 4 is Hodder's, measured at Çatalhöyük.  Both may be right about their own sites.  What the computation shows is that at the deepest sequence the measured house life is four to thirteen times Osgood's assumption, and Osgood applied a Jarmo figure to Çatalhöyük and Jericho without checking either.

Part 4: The clock with no radiocarbon in it

If radiocarbon is the objection, the test that gets around it counts things inside the tell.

Tree rings.  Kuniholm and Newton measured 500-odd charcoal fragments left from Mellaart's excavation and built "a 570 year tree-ring chronology for the site", internally replicated across nine trees from levels IV to VII.¹  That 570 years is the span of the ring series, not of the occupation, because one old tree contributes centuries of rings by itself.  One result in the chapter is a duration.  A post from level VII kept its terminal ring, so its felling year is fixed:  "The terminal year for CTL-16 at relative year 1437 is 132 years earlier than the last preserved ring (relative year 1569) of CTL-6 from level VIB.  This tells us that the difference, in calendar years, between the two timbers is at least 132 years, and suggests that the life span of level VII and part of VIB is likewise at least 132 years."¹  One level and part of a second, 132 years at the least, counted in rings.  There are seventeen levels.

Two cautions, both from the same chapter.  The chronology is floating:  "we cannot yet convert our relative dating system to calendar years", which is why decadal samples went to Arizona for radiocarbon wiggle-matching.¹  And most timbers lack bark, so "the end dates for the different levels are not in fall-back order as one might expect".¹  The 132 years survives both, because a ring count between two samples needs no calendar date.

House lives.  Hodder's 2012 summary:  "Many houses were lived in for 50-100 years.  The people of Çatalhöyük had great problems keeping the buildings standing for this long as the mud brick walls made of smectitic clay tended to lean and bend."⁸  Buildings were "replaced in the same location, using the same layout of walls over 4-6 rebuilds."⁸  Osgood needs them to fall every 7.5 to 15 years.

Plaster layers, and why this article does not count years with them.  The wall plasters at Çatalhöyük really are stacked in the dozens and hundreds.  Wendy Matthews and colleagues put one 3 cm fragment of collapsed wall plaster under a microscope:  it "includes 160 layers of plaster", read as "80 couplets" of a preparation layer plus a finishing coat.²²  Hodder's 2012 summary reports over 450 coats on the main room of Building 5 against only 3 or 4 in adjacent rooms.⁸

None of that converts to years, and a citation has to be corrected here.  Hodder wrote in 1999 that "Up to 40 layers of replastering were found on the walls and floors of Building 1.  We believe, on the basis of correlations with dendrochronological sequences, that these replasterings occurred annually (Kuniholm & Newton 1996)."²³  That chapter was obtained for this article.  It reports timber chronologies.  It says nothing about plaster, and neither does anything else in the volume.¹  Hodder's own 2012 text calls the plastering "frequent (annual, seasonal and monthly – Matthews 2005)".⁸  So the annual reading is an assertion whose stated authority does not carry it, and no year count is drawn from plaster here.

Now the arithmetic, using only the house lives.  Seventeen superimposed building levels at 50 to 100 years each is 850 to 1,700 years.  Osgood's window for the entire world's Stone Age is 330 years.  The Çatalhöyük East Mound alone is between two and a half and five times that, before Jericho, Göbekli Tepe or Abu Hureyra get any of it.  Nothing in that paragraph is a radiocarbon date.  Rejecting radiocarbon does not touch it.

Superposition

Osgood's contemporaneity model fails at the same point.  The four site sequences in step 5 sum to about 9,400 years, and Osgood's answer is that cultures which look sequential were really parallel.  That works for Halaf beside Samarra.  It does not work for Jericho, whose single column holds the Natufian, PPNA, PPNB, Pottery Neolithic, Chalcolithic and Early Bronze stacked vertically in one place.⁷  A layer cannot be contemporary with the layer beneath it.

The Flood horizon nobody has named

If the Flood was in 2518, 2348 or 3298 BC, some stratum in some Near Eastern tell is the Flood deposit.  No creationist source found for this article names one.  Osgood's framework puts the whole Neolithic above the Flood, so the deposit at Jericho would lie beneath the Natufian, more than 21 m down.  Habermehl's puts it below 1.76 million conventional years of record.  A young-earth model of the Neolithic would need to say which layer, at which site, and why the sequence above it is 330 years long and not 8,000.

Part 5: Where the interpretation really is unstable

The creationist critique of Neolithic archaeology is partly right.  Four headline readings have been reversed or are contested, three of them by the excavators themselves.

The world's first temple.  Clare, head of the Göbekli Tepe project, writes that "it was the archaeologists themselves who delivered and subsequently cultivated the now popular opinion that Göbekli Tepe is the site of the »World's First Temples«."³  Deep soundings for the new shelter anchors "have exposed the weaknesses of the temple-narrative, meaning that a revision of the popular scientific view is now unavoidable".³  Thomas at ICR attacked the "religion before agriculture" story in 2012, and on its instability he was right.

Schmidt's three layers.  The mortar dates show "a clear chronological overlap of the round-oval special buildings (Layer III after Schmidt; PPNA; ~9.500–8.700 calBC) and the rectangular structures (Layer II after Schmidt; EPPNB; ~8.700–8.200 calBC)."³  The tidy tripartite scheme of the 2013 paper is gone.

The ritual backfilling.  "While previous conclusions postulated an intentional (ritually-motivated) backfilling of these structures at the close of their active use-lives as buildings, a review of the available evidence now suggests an inundation by displaced archaeological deposits from adjacent higher-lying knolls and slopes of the mound."³  This is the reading Cox's inverse-stratigraphy argument depended on, withdrawn by the project.

Kenyon's fortress.  Bar-Yosef reads the Jericho wall and tower not as fortifications but as flood defence:  "a plausible alternative interpretation for the Neolithic walls of Jericho is that they were built in stages as a defense system against floods and mudflows."¹⁴  That is closer to a creationist's instincts than Kenyon's reading.

The Ubaid.  The editors of the standard volume write that "in the case of absolute chronology the term 'Ubaid period' is not consistently defined", and that in some regions "the Ubaid period runs only from about 5300 to 4300 b.c."¹⁷  The same ceramic label covers different absolute spans in different places.  Osgood's instinct that the neat serial sequence is partly parallel was not stupid.

Every one of these is a change of interpretation, not of date.  Göbekli Tepe was 9500 to 8000 cal BC before Clare's revision and is 9500 to 8000 cal BC after it, and the revision made the buildings longer-lived, not shorter.  Bar-Yosef's flood-defence tower is dated by the same fifteen determinations as Kenyon's fortress.  The Ubaid's regional looseness does not move the southern sequence, which is where Osgood puts Babel.  The live mainstream disagreements here, the Natufian gap at Jericho and the East and West Mound overlap, are worth a few centuries at most against a gap of six to ten thousand years.  "Even the scientists disagree" is true and is worth about three percent of the distance.

Part 6: Babel, the King List and the flood layers

Three further creationist moves bear on the Neolithic, and each is treated briefly because a sibling article owns the detail.

Moving Babel.  Habermehl's 2011 paper argues that Shinar is the Khabur triangle of northeast Syria, with Akkad at Tell Brak and Calneh at Tell Fakhariya, because southern Iraq's alluvium "is a delta that was formed later on, well after the Babel dispersion".²⁴  That does not get the Neolithic out of the way.  It moves Babel into the middle of it:  Tell Brak has stratified Ubaid and Late Chalcolithic levels beneath its historical ones, Tell Fakhariya is 2 km from Tell Halaf, the type site of the Halaf culture, and Göbekli Tepe is about 200 km west.

The same paper puts a timing problem to its own side.  Under Ussher the 106 years between Flood and Babel must hold an 80 m fall in sea level, the drying of Babylon and a 600 km migration:  "This seems like a very unlikely scenario to have taken place within 106 years."²⁴  The Septuagint's 531 years would help, but "no Ice Age model for this timeline has been produced."²⁴  Thirteen years later Eric J. Tweedy, in the same journal, reached the opposite result:  the Oard Ice Age model "most likely is not compatible with the Masoretic Text and secular settlement traditions for northern Europe without substantial changes", while "Using the Septuagint timeline, Oard's model can be aligned both with the text of Genesis and with the cited settlement traditions without alteration."²⁵  Both say the model is the thing that has to move.  It is adjudicated in the Ice Age article, not here.

The Sumerian King List.  The list says of the kings before the flood:  "5 cities were they;  8 kings reigned their 241,200 years.  The Flood swept thereover."²⁶  It is often laid beside Genesis as an independent witness.  Its own editor judged otherwise.  Jacobsen's note to that line reads "Here ends the secondary section on the antediluvian period which was added to the King List."²⁶  Three manuscripts carry the antediluvian section and four do not,²⁶ and "we cannot doubt that its original setting should be sought in the lore of Eridu(g) and not in the King List."²⁶  The flood line is an import from another Mesopotamian text.  Where the list bears on this article is geography:  its antediluvian cities include Eridu and Shuruppak, tells with stratified Ubaid levels beneath their historical ones.  Raúl López's creationist argument that the 241,200 years and the Genesis 5 ages are the same numbers in different bases is about number-writing, not the Neolithic.²⁷

The flood layers.  Woolley's deposit at Ur is the most cited physical evidence for the Flood in Mesopotamia, and Woolley's own words are the answer:  "so far from the disaster being universal some at least of the local centres of civilisation survived it."²⁸  Lambert and Millard, editing the Babylonian flood epic:  "Several ancient sites have revealed flood layers separating strata of different civilizations.  Presumably at various sites and on several occasions floods did wipe out the existing culture… (The present writers do not belong to that school which relates flood stories all over the world to one prehistoric cosmic disaster.)"²⁹  Lambert and Millard's "at various sites and on several occasions" is the point:  one year-long global Flood cannot be several deposits that leave the neighbouring towns occupied.  Woolley's own stratigraphy puts his deposit between hand-made al 'Ubaid pottery below and wheel-made pottery above, so it sits inside the Mesopotamian sequence rather than beneath it.  Mallowan's 1964 "Noah's Flood Reconsidered", the standard treatment of the Kish and Shuruppak layers, was not read, so no claim is made here about the height or date of those two.

Objections and responses

"The dates are all radiocarbon, and radiocarbon assumes a steady-state atmosphere that the Flood would have broken."  This is the creationist position at its strongest, and it is a real argument about calibration.  It belongs to the radiocarbon article and to Egypt and the Flood Date, which shows that no published creationist recalibration curve exists.  Here the answer is narrower.  The counted building levels, the measured house lives, the 132 tree rings between two Çatalhöyük timbers and the superposition at Jericho use no radiocarbon.  A model that breaks radiocarbon still has to fit 850 to 1,700 years of building levels into 330.

"Tells are mud-brick rubble.  Bar-Yosef says so.  They pile up fast."  Correct, and conceded in Part 3 with the arithmetic.  The problem is that at Çatalhöyük houses measurably did not fall every decade.  They stood 50 to 100 years and had to be braced to manage it, and two timbers from adjoining levels are 132 tree rings apart.  Depth was never the argument.

"The plaster counts are soft, and you admit it."  They are, and this article counts no years from them.  Hodder's 2012 text calls the replastering "annual, seasonal and monthly", and the 1999 claim that Building 1's forty coats were annual cites a chapter that says nothing about plaster.  That is why Part 4 rests on the house lives and the tree rings instead.  Dropping the plaster counts makes the answer longer, not shorter:  the floor rises from 680 years to 850.

"The excavators got Göbekli Tepe wrong for twenty years.  Why trust them on anything?"  They did, and they said so in print.  But every correction Clare reports lengthens the site's life.  A single ritual burial would have been the short reading;  the revision replaces it with five centuries of rebuilding and a village on top.  A critic who cites the reversal has to accept its direction.  And this article applies the same rule to itself in Part 4, where a mainstream citation of twenty-seven years' standing turns out not to support what it is cited for.

"Osgood showed that Halaf, Samarra and Ubaid II are contemporary.  Sequence is an illusion."  He showed that neighbouring cultures overlap by a few centuries.  Jericho's column has Natufian under PPNA under PPNB under Pottery Neolithic under Chalcolithic under Early Bronze.  No regional overlap makes the bottom of that column the same age as the top.

"The Septuagint fixes this."  It buys 780 years.  That is 23% of the gap at Çatalhöyük, 11% at Göbekli Tepe and 8% at Shubayqa.  A creationist who adopts the Septuagint gains Egypt's Old Kingdom and does not gain the Neolithic.

"Domestication can be fast.  The Soviet fox-farm experiment changed morphology in about 40 generations."  This is the best creationist answer to the domestication lag.  Belyaev's foxes were tame and visibly altered within about 40 generations of deliberate selection.³⁰  Zeder's lag of 1,000 to 1,500 years is about 300 to 500 goat generations, an order of magnitude more, which fits weaker and less directed selection.  What the fox result cannot do is shorten the archaeological lag, which is measured, or supply four to six goat mitochondrial lineages from one ark pair.

Objections to this article's own verdict.  Every absolute date in Part 1 is radiocarbon, and radiocarbon needs a calibration curve that is itself a construction.  The Çatalhöyük start and end dates are cited at one remove.  The 50-to-100-year house life is Hodder's summary figure and he gives no method for it in the chapter cited.  The Mesopotamian flood paragraph rests on Woolley and Lambert and Millard, not on Mallowan.  The PPNA tower's dating comes from Bar-Yosef's reading of Kenyon's list.  None of these gaps is small enough to ignore and none is large enough to move a date by more than a few centuries.

Issues and unanswered questions

• Hodder 1999 cites Kuniholm & Newton for a claim they do not make. The chapter was read for this article. It reports timber chronologies and says nothing about plaster. No later Çatalhöyük publication found here repeats the annual reading, and nobody appears to have checked the citation.
• Matthews 2005, the chapter behind the 450-coat figure, is on JSTOR and was not read. Matthews et al. 1997, same author and same material, supplies the 80-couplet count instead.
• The 50-to-100-year house life has no published method attached. Hodder states it flatly in a four-page summary. It is the number Part 4 leans on hardest and it deserves a primary.
• Bayliss 2015 and Marciniak 2015, the Bayesian start and end models for Çatalhöyük East, are closed and cited through Orton 2018.
• Mallowan 1964 on the Kish and Shuruppak flood layers is unread. It is on JSTOR.
• Karahan Tepe has no open excavation report with a radiocarbon series. Kenyon's Jericho volumes were not read; the fifteen dates come from Bar-Yosef's Table 1. Zeder 2011 is closed; Zeder 2008 stands in.
• No creationist has counted the layers. Not one source found here engages the superimposed levels at any tell, the plaster counts, or the measured house lives. A creationist model would need to show that 17 levels fit into under 330 years, or that the 50 to 100 year house lives are wrong, or that two timbers 132 rings apart can come from a settlement that lasted a few decades.
• No creationist has answered Clare 2020, six years on.
• The population problem is unresolved inside the creationist literature. Osgood needs about 25,000 people at Babel, which he puts about 100 years after the Flood. Carter and Hardy's model passes 1,000 at that point only under favourable settings.
• The Flood horizon is unlocated at every tell. See Part 4.

Verdict

The Near Eastern Neolithic does not fit between any Genesis Flood date and Abraham, and the reason is not radiocarbon.  Göbekli Tepe's wall plaster is 6,000 to 7,200 years before the Flood on any text.  Jericho's column carries 8,000 years of stratified occupation beneath the Flood horizon and no break at it.  Çatalhöyük's seventeen building levels at a measured 50 to 100 years each give 850 to 1,700 years for one mound against 330 for the world, and two of its timbers are 132 tree rings apart, with no radiocarbon in either count.  Goats took a thousand years to look domestic after people began keeping them.  The Septuagint closes about a tenth of the distance.

What the young-earth side gets right is not nothing.  Tell depth proves nothing on its own, and the arithmetic here says so before anyone else does.  The first-temple reading, the ritual backfilling, Schmidt's three layers and Kenyon's fortress were all wrong or are contested, and the excavators said so themselves.  Cox read the radiocarbon paper, Habermehl's model is falsifiable, and Osgood saw in 1986 that the question was the rate of tell formation.

What an honest young-earth position looks like.  Say that the objection is to the radiocarbon calibration and that no creationist curve yet exists.  Say that the Çatalhöyük level counts and Clare 2020 are unanswered.  Name the stratum that is the Flood at Jericho, or say that nobody has.  Give one date for Göbekli Tepe with a method, rather than four without one.  And repeat Osgood's own sentence, still true forty years on:  "It is impossible at this point in time to give an accurate assessment of the time taken for the build-up of these strata."

Companion pages

• The literature — every source consulted and which way it cuts.
• Quarantine — the claims held back, and why.
• Egypt and the Flood Date — the Old Kingdom, and why no creationist recalibration curve exists.
• Radiocarbon — the calibration curve, the plateaus, and the post-Flood depletion argument.
• The Ice Age — the model Habermehl and Tweedy disagree about.
• Jericho and the Conquest — Late Bronze Jericho, 7,000 years above the tower.
• Mesopotamian Absolute Chronology — the eclipse-anchored first millennium.

Sources