The Middle Pleistocene Muddle: The Dates

"it would be nice if some creationist researchers would check it too." — Tas Walker, Creation Ministries International, on the 2020 dating of the Kabwe skull¹

Summary

The parent article sets out what the Middle Pleistocene Homo fossils are and why their names keep being redrawn.  The naming dispute is silent on dates.  So this article asks a different question of the same sites:  how well are they dated, how much have the dates moved, and how does the distance between mainstream disputants compare with the distance between any of them and a Flood year.  The answer is uneven, and the article prints the uneven parts.

The dates are where the young-earth side has no answer, and the naming dispute does not supply one.  At Mauer two methods on different materials agree within a few percent.²  At Sima three methods on three materials fall between 396 and 443 thousand years.³  At Kalambo Falls three luminescence clusters come out in the order of the layers.⁴  Dates have moved, at Sima, Schöningen, Kabwe and Petralona, by factors of 1.2 to 4, and once after a morphological objection.  Counting every measured age ever published, the widest spread at any site is tenfold, while the lowest measured age anywhere in the interval is 14 times a Flood date and the figures the current papers prefer are 34 to 200 times it.⁵  No creationist source read for this article names a faulty input to any of these measurements, and the one who wrote about Kabwe's date said twice that he had not read the paper.¹  The one documented case of two methods agreeing and both being wrong, at Sima in 1997, ran too young, which is the wrong direction for a young-earth reading.⁶

Parts 1 and 2 take the sites one by one and print the limits of each date with it.  Part 3 sets the spread inside the mainstream literature against the gap to a Flood date, site by site.  Part 4 gives the young-earth chronology in its authors' own words, about the fossils they actually name, before anything is priced against it.  Part 5 states what the order of the layers does and does not show, and the one assumption every method here shares.  The young-earth literature on the fossils themselves, and the places where it misstates what it cites, is examined in its own article.

Part 1: Mauer and Sima de los Huesos

Mauer: two methods, two materials, one answer

The type site of heidelbergensis was dated in 2010 by two methods that measure different events on different materials.  Infrared radiofluorescence on sand grains records when the sand was last exposed to light;  combined electron spin resonance and uranium-series on herbivore teeth records when the enamel formed.  "we show that two independent techniques, the combined electron spin resonance/U-series method used with mammal teeth and infrared radiofluorescence applied to sand grains, date the type-site of Homo heidelbergensis at Mauer to 609 ± 40 ka."²  The sand gave "Ages of 607 ± 55 ka, 603 ± 56 ka, 554 ± 33 ka, and 502 ± 27 ka were obtained for the lower sands, whereas the upper sands gave ages of 508 ± 50 ka and 420 ± 23 ka".²  The teeth gave 624 and 627 thousand years for the lower sands and 458 and 502 for the upper.²  "Different materials and events are dated (i.e., the deposition of the Mauer sands by IR-RF and the formation of mammal teeth by ESR-US).  The methods therefore deliver independent results."²  Both methods also put the upper sands younger than the lower, as the layers require.

The honest limits.  Two of the four lower-sand radiofluorescence ages, 554 and 502, are not in the final mean, which takes only the two samples bracketing the find layer.  Four of the eight teeth were set aside as minimum ages because one parameter could not be determined:  "for five samples (i.e., ≈25% of the material analyzed), determination of the p-parameter was not possible … The results obtained should be treated as minimum ages and cannot be considered in the geochronological interpretation."²  And the two methods share one assumption, a constant environmental dose rate from the same radioactive decay series.  They are independent in what they measure, not in what they assume about decay.  That point is taken up below.

Sima de los Huesos: a date that moved three times, and in which direction

This is the site in the interval where a published radiometric age was challenged because it clashed with an evolutionary model and was then given up.  The story should be told straight, with every step in it, because each side tends to tell the steps that suit it.  The published figures run:  at least 200 thousand years in 1997, more than 350 thousand in 2003, more than 530 thousand in 2007, and about 430 thousand in 2014.  That is three moves, two up and one down.

The first dating, in 1997, was by the excavators' own geochronologist.  Uranium-series on sixteen human bones gave 148 ± 34 thousand years, and on 25 bear bones 154 ± 66.  The authors did not offer either as an age, because the human bones lie below the bears:  "Because the human bones seem to be stratigraphically older than chose of the bears, the results would indicate that most of the bones have been accumulating uranium irregularly with time."⁶  Ten years later the same first author and his colleagues put it more bluntly.  The bone dates, they wrote, "are clearly affected by irregular post-depositional uranium cycling".⁷  Then came the result that matters most to this article.  Electron spin resonance on six bear bones gave 189 ± 28, their uranium-series gave 181 ± 41, and the agreement was offered as evidence that both were right:  "Such agreement is highly suggestive that uranium uptake in these bones was close to the early-uptake (EU) model, and the dates are essentially correct."⁶  Conclusion:  "a firm minimum age of about 200 ka for the human entry: and suggestive evidence of entry before 320 ka."⁶  The 320 came from three further bear bones, which the 2007 paper reports beside the six.⁷  Two things about this result need saying together.  The authors called the six bone dates "essentially correct".  And what they concluded for the humans was a minimum, not an age.

The next two moves both went up, and both came from one flowstone.  In 2003 the same group dated a sheet of cave calcite that it understood to lie over the human bones.  The upper part gave finite ages, "in stratigraphic order from 153 to 281 kyr";  the lower part was too old for the instrument to resolve, which gave only a minimum of 350 thousand years.⁷  In 2007 a more precise instrument was used on the same lower part:  "The new dates cluster around 600 kyr.  A conservative conclusion takes the lower error limit ages as the minimum age of the speleothem, or 530 kyr."⁷  The whole result hung on one phrase in that abstract, that the flowstone was "overlaying human bones".

In 2012 Stringer said in print that this was wrong, and gave two kinds of reason.  One was about the cave, and he had raised it before.  The other was that the fossils looked too Neanderthal for the date:  "More recently, I have also challenged the older age estimates obtained for the SH sample from a dated speleothem.  That challenge was based on taphonomic issues rather than the age determination itself, but dating work on the speleothem is continuing (J.-L. Arsuaga, personal communication).  Regardless of those considerations, morphological and genetic data suggest that the divergence of the neanderthalensis and sapiens lineages very likely postdates 530 ka.  Therefore, Neanderthal apomorphies would not be expected at such an antiquity."⁸  Of colleagues who accepted the old date he wrote that they argued "on the basis of an (in my view erroneous) early date".⁸  Young-earth writers followed this as it happened, and one of them reported the weak point accurately.  Elizabeth Mitchell of Answers in Genesis, in June 2012:  "The dates were derived from uranium dating of a stalagmite above the pit, which of course does not necessarily mean the stalagmite formed at the same time the people were buried, a point of concern raised by Stringer."⁹  That is the one place found in the young-earth literature on this interval where a specific input to a specific date is named as doubtful.  The point was Stringer's, and it was about which layer lay on which.  Mitchell again, in December 2013, six months before the excavators' paper with the new date appeared in Science:  "when researchers re-dated the site and declared its inhabitants much older than the original determination—600,000 instead of 350,000 years—the evolutionary story got flipped".¹⁰

In 2014 the date was revised his way.  Lee Arnold and colleagues re-read the stratigraphy, concluded the dated flowstone lay under the fossils rather than over them, and turned the 530 into a maximum:  "The previously published minimum age estimate of >530 ka for Sima de los Huesos (Bischoff et al., 2007) received considerable attention vis-à-vis its perceived incompatibility with morphological and genetic evidence for hominin evolutionary divergences during the Middle Pleistocene (e.g., Endicott et al., 2010; Stringer, 2012).  The revised stratigraphic interpretations of this U-series age presented earlier … effectively negates any direct comparisons with existing genetic divergence estimates, since these U-series results are now known to provide a non-finite maximum age estimate on the overlying fossils."¹¹  The replacement rests on new measurements by three methods on three materials.  Uranium-series on a calcite raft stuck to a skull:  "U-series dating of a cave raft speleothem deposited directly on a hominin cranium (Cranium 4) from LU-6 yielded a mean age of 434 +36/-24 ka".³  Two luminescence methods on the sediment above the bones:  "weighted mean ages of 433 ± 15 ka (thousands of years) and 416 ± 19 ka, respectively, for allochthonous sedimentary horizons overlying the hominin-bearing clay breccia.  The six replicate luminescence ages obtained for this deposit are reproducible and provide a combined minimum age estimate of 427 ± 12 ka".¹¹  "All six of the TT-OSL and pIR-IR ages (428 ± 27 to 441 ± 25 ka for K-feldspar and 396 ± 35 to 429 ± 32 ka for quartz …) and one of the ESR quartz ages (443 ± 90 ka …) for this unit are statistically indistinguishable".³  The result:  "a minimum age of ~430 ka for the SH human fossils, which is some 100 ka younger than previously reported".³

Two results sit awkwardly in the 2014 paper and belong in print.  A bear tooth gave "261 +26/-25 ka, which is interpreted as a minimum age owing to uncertainties with the geochemical and dosimetry histories of this specimen", and "two additional ESR quartz samples taken from LU-7 yielded significantly older ages of 0.9-1.5 Ma, these are interpreted as maximum ages due to potential methodological complications".³  The 430 thousand figure is the cluster of the remaining determinations, not the unanimous verdict of every sample measured.

What a young-earth writer can fairly say:  a leading worker called a published radiometric date erroneous because the fossils did not fit his model, and the date was then revised his way.  What the other side can fairly say:  Stringer's challenge was to the taphonomy, which layer lies on which, "rather than the age determination itself", and the 2014 revision was exactly a taphonomic one.  The new age rests on uranium-series and two luminescence methods whose mean ages, 434, 433 and 416 thousand years, lie within 5 percent of one another.  And the size of the change was from a floor of 530 to about 430, which is 19 percent, or 28 percent if counted from the 600 the 2007 dates clustered around.

Now the 1997 result, which both sides are tempted to overstate.  Two methods agreed on six bear bones, near 190 to 200 thousand years, and the authors read the agreement as a sign that the dates were right.  Whether those bone dates were in fact too young depends on where the six bones lay, and the 1997 paper was not read in full for this article.  The 2014 paper says the bear bones in the deposit above the humans had "previously accumulated" elsewhere in the cave before a debris flow carried them in, and it dates that flow to about 430 thousand years.³³  If the six bones were among them, their true age is more than twice what the two methods agreed on.  That is the likely reading, and it is not a proven one.

What can be said without that step is narrower.  For the humans, the 1997 paper claimed only a floor of about 200 thousand years, and a floor that lies under the true age is not a mistake.  The agreement of the two methods was not independent in any case, because both depend on when uranium entered the same bones.  That matters for Kabwe and Harbin, whose dates are also uranium-series on bone.

The direction matters too, and it should be stated exactly.  The published figure went up in 2003 and up again in 2007, and came down in 2014.  It did not rise at every step.  But every figure published since 1997 is higher than the bone dates of 1997, and the reason is understood:  bone takes up uranium after burial, so a uranium-series date on bone that has only gained uranium falls short.  Bone that has lost uranium errs the other way, as the Harbin skull shows below.  A dating history in which the bone dates fall short does not help an argument that the true age is a hundred times lower still.

Part 2: Kabwe, Petralona, Schöningen, Gran Dolina, Kalambo and Harbin

Kabwe: the fastest-aging man, and the one direct measurement

Lubenow's chapter on Kabwe is titled "Rhodesian Man:  The Fastest-Aging Man in the World", and his list runs "11,000 years of age (?).  Arthur Smith Woodward, in 1921. …  40,000 years of age.  Carleton Coon, in 1962. …  125,000+ years of age.  Richard G. Klein, in 1973. …  300,000—400,000 years of age.  Ian Tattersall, in 1999."¹²  His conclusion:  "Thus Rhodesian Man aged about 350,000 years in just 82 years, even though there was not a shred of rigorous evidence indicating that he was other than a few thousand years old.  Here we see the power of philosophy over fact."¹²  Walker's 2020 list adds Encyclopaedia Britannica's 30 to 40 thousand and a 1974 racemization figure:  "In 1974, Bada et al. in their words “tentatively assigned an age of about 110,000 years” to the skull based on aspartic acid racemisation.  Clearly, there have been lots of different dates put forward for this skull."¹

Two things in these lists need setting against their sources.  The first is the 1974 figure.  Bada's paper does not say the measurement was made on the skull.  It was made on "a 6.5-g piece (British Museum fragment E793)", it is called "a preliminary date", and the rate behind it was set at another cave, Mumbwa, about 150 km away.¹³  The paper nowhere says which bone the piece came from.

The second is a measurement that neither list has.  In 1965 a laboratory at the University of California, Los Angeles, published a radiocarbon result under the heading "Broken Hill Man".  The sample was "Animal bones dated by collagen method said to come from bone cave at Broken Hill", and "Date is composite of several bones".  The result is printed as older than 9,000 years, with no upper figure.¹⁴  That is a floor and not an age, and it was measured on animal bone, not on the skull.  Five years later Robert Whitelaw entered it in a table in the Creation Research Society Quarterly as "UCLA-630: “BROKEN HILL MAN” Rhodesia (incl. animal bones) … 9,000", with no sign in front of the number,¹⁵ and counted Broken Hill Man among human remains "datable within 30,000 years".¹⁵  Other open-ended results in the same table keep their sign, so this looks like a slip on one line.  The copy read here is the Society's own archive edition, which has been set again in modern type, so the sign may have been lost then and not in 1970.  Either way the paper's conclusion treats the bones as dated.  Of the young-earth papers read for this article it is the only one that cites a measurement at one of these sites in favour of a young age, and the figure it cites is not what the laboratory reported.

Much of this is right.  Woodward, with the bone in his hands, judged it recent on two grounds.  The fauna:  "The occupation of the cave, therefore, seems to have been at no distant date—it may not even have been so remote as the Pleistocene period."¹⁶  The preservation:  "The skull is in a remarkably fresh state of preservation, the bone having merely … animal matter and not having been in the least mineralised."¹⁶  And the later figure of about 500 thousand years rested on comparing animal bones from a site that no longer exists.  Rainer Grün and colleagues say so in 2020:  "the original site has since been completely quarried away, and-although the cranium is often estimated to be around 500 thousand years old-its unsystematic recovery impedes its accurate dating and placement in human evolution."¹⁷

What Lubenow could not know in 2004 is that the skull has since been measured.  Grün's team ran uranium-series by laser ablation on a detachable piece of the skull base and got "a best age estimate of 299 ± 25 thousand years (mean ± 2σ)".¹⁷  The details show how much the number can carry.  They are quoted here from the authors' accepted manuscript.  The printed paper rewords several of these sentences and writes "kyr" for "ka", and it gives the same figures.  It is a minimum:  "The U-series ages must be regarded as minimum age estimates as there may be an initial burial time with little or no U-uptake."¹⁷  A second tissue from the same skull disagrees.  Dentine from a tooth was dated too, and the bone result "is 83±25 ka older than the result from the dentine, which can be either explained by a delayed start of the U-uptake into the dentine, or that the dentine stayed an open system for longer".¹⁷  Electron spin resonance could not give an age, because the burial environment is unknown, but set a ceiling:  "a combined U-series-ESR calculation yields an upper 2σ age limit of 322 ka".¹⁷  The fauna was reworked and found "inconclusive".¹⁷  And the authors say what they cannot do:  "none of the materials collected from the Broken Hill site can unequivocally be connected to the depositional history of the skull, nor can we think of an additional dating analysis to independently verify our results."¹⁷  So "the 2020 direct date" is one method on one bone, stated as a minimum, bounded above by a second method, with a second tissue from the same skull 28 percent younger.  It is weaker than the Mauer or Sima dates.  It is also a measurement on the skull itself, as against a comparison of animal bones, and "not a shred of rigorous evidence" is no longer a fair description.

Kabwe is the one fossil in this article whose dating history is as loose as the young-earth literature says.  If Coon's 1962 correlation at 40 thousand years is counted, as Lubenow reports it, the skull has been published at ages 12.5-fold apart, which is more than the ninefold gap between the lowest of them and a Flood date.⁵  It is also the one fossil whose find site no longer exists.  The two facts are related.

Petralona, Schöningen, Gran Dolina, Kalambo, Harbin

Petralona.  Christophe Falguères, Stringer and colleagues in 2025:  "Previous results obtained on various types of samples yielded a large age range between about 170 and 700 ka, precluding any consensus on the age of the human fossil.  On the other hand, the original stratigraphic position of the cranium also remains enigmatic".¹⁸  The spread is wider than that summary says.  The same paper's review of the earlier work records an ESR age on a piece of the skull itself:  "The bone fragment yielded an age of ~130 ka."¹⁸  From 130 to 700 is a factor of 5.4 for one skull over sixty years.  Their new figure dates the calcite that grew on the skull, "a minimum age of 286 ± 9 ka",¹⁸ which is a floor under the death, not a date of it.

Schöningen.  The wooden spears from the open-cast lignite mine at Schöningen have been 400, 300 and 200 thousand years old in print within twenty-eight years.  Jarod Hutson and colleagues in 2025:  "Initial estimates placed the site at around 400,000 years;  this age was later revised to roughly 300,000 years.  Here, we report age estimates for the “Spear Horizon” based on amino acid geochronology of fossils obtained directly from the find-bearing deposits.  Together with a reassessment of regional Middle Pleistocene chronostratigraphy, these data place the Schöningen spears at ~200,000 years."¹⁹  A uranium-series date on the sequence was withdrawn "because of their recognized open-system behavior",¹⁹ and the earlier luminescence dates came from layers below the spears, so "can therefore only serve as maximum age estimates".¹⁹  The authors' own word for the old dates is "mismatch":  "With claimed ages of ~400 or ~300 ka for the Spear Horizon, Schöningen stood as an outlier among sites attributed to the Lower Paleolithic and to Homo heidelbergensis … Our dating evidence for the Spear Horizon corrects this mismatch and aligns the Schöningen spears within the timeframe of European Neanderthals and the Middle Paleolithic."¹⁹  That deserves to be said plainly:  the site did not fit the expected sequence, and the new date makes it fit, and the species credited with the spears changed with it.  The other side deserves saying too.  The earlier figures were correlations and maxima from lower layers;  the new one is the first measurement on the find layer;  and amino-acid racemization is not a radiometric method at all but a temperature-dependent chemical clock calibrated against other dated sites, so it is not independent of them.

Gran Dolina.  The only direct date on an antecessor fossil is unusually frank about how weak it is alone.  Mathieu Duval and colleagues ran ESR on the enamel of the same molar fragment Welker later sequenced:  "a final age estimate ranging from 624 to 949 ka, which covers all possible uranium uptake scenarios.  Last, the additional magnetostratigraphic data collected within TD6 enables to further constrain the initial ESR chronology and to propose an age of between 772 and 949 ka for Homo antecessor".²⁰  The lower end is cut off not by the ESR but by magnetism:  the layer is reversely magnetised, so it is older than the last reversal at 772 thousand years.  The paper tabulates every earlier date on the layer.  Luminescence at the base of the layer above gave "960 ± 120 ka" that "should be regarded as a maximum age estimate";²⁰ single-grain luminescence on quartz gave "846 ± 58 ka";²⁰ three herbivore teeth gave "730 ± 63 ka (Falguères et al., 1999), recently revised to 766 ± 81 ka";²⁰ and ESR on quartz gave seven ages from "591 ± 71 to 947 ± 90 ka", with the authors adding that "the significant age scatter observed (20%) makes any further chronological interpretations almost impossible".²⁰  Five numerical methods on one layer, 591 to 960 thousand years, a spread of about 40 percent.  That is a fair picture of how well a site near the limit of these methods can be dated:  to within a few hundred thousand years, not better.  No method on the layer has ever returned an age near the bottom of its own range, and the lowest of them is 136 times a Flood date.

Kalambo Falls.  Sixteen sand samples around the worked wood were dated by luminescence on potassium feldspar.  The ages for the wood-bearing levels fall in three clusters, older below and younger above:  "at least 476 ± 23 kyr ago (ka)" for the joined logs at the bottom,⁴ and "four wood tools from 390 ka to 324 ka" higher up.⁴  The order is not perfect.  Two samples in one section come out the wrong way round, and the authors say so:  "At BLB2, the central ages for samples KF10 (355 ± 18 kyr) and KF11 (422 ± 28 kyr) are reversed.  The reason for this is unknown."⁴  Nor is 476 the only figure the method gives.  Feldspar loses part of its signal over time, and with a correction for that loss the three clusters become 554, 452 and 377 thousand years, 16 percent older.  The authors print the uncorrected, younger set.  A second mineral, quartz, was dated too and agreed with the feldspar, but only in the youngest sediments, under about 60 thousand years;  at the levels with the wood the quartz signal was full and could say nothing.⁴  So for the wood itself this is one method on one mineral.

Two more facts from the same paper bear on how old the wood can be.  It is not fresh wood:  "Infrared spectroscopy shows partial mineralization of the wood (silica)".  And it was tested by radiocarbon, to check that it had not worked its way down from younger layers:  "all dates were infinite (more than 50 kyr ago (ka), T. Higham personal communication)."  The authors put its survival down to water:  "A permanently elevated water table has preserved wood and plant remains".⁴

Harbin.  Because the find spot is hearsay, Qingfeng Shao and colleagues matched the sediment stuck inside the skull to a lake bed by its chemistry and strontium isotopes, and ran uranium-series on the bone itself.  Ten drilled samples of the skull gave ten different answers:  "The corrected U-series apparent ages are highly scattered, ranging from 62 ± 3 to 296 ± 8 ka".²¹  The authors sort them into two groups of five and keep one.  The older five, 185 to 296 thousand years, are set aside as bone that lost uranium after burial, which makes a date too old:  "Leached samples do not provide any useful age information".  The younger five, 62 to 148, are kept as minimum ages, and the highest of them is taken as the floor.²¹  The two lowest, 62 and 85, are on dentine from a tooth root, and "The young ages are probably caused by a delayed U-uptake".²¹  The paper's headline is "Uranium-series disequilibrium dating directly on the cranium suggests that the cranium is older than 146 thousand years".²¹  The bed the skull is matched to is "138–309 ka",²² a range that comes from luminescence dating of the formation in earlier work, not from this study.  This is the weakest date in the article:  a floor on bone, chosen from a nearly fivefold scatter by a judgment about which samples leaked, plus a provenance argument for a skull with no recorded find layer.  Its lowest figure, 62 thousand years, is the lowest age measured on any fossil in the table below.

Part 3: How far apart the disputants are, in numbers

The arithmetic in this part is done by a short script, date-spread.ts, kept with the series' research notes in the parent article's folder at html/unpublished-ai/the-middle-pleistocene-muddle/research/date-spread.ts, with its output beside it as date-spread.out.txt.  Every input is a published age quoted in this article, and the note footnote on the table lists them with the formulas so a reader can redo the sums by hand.  Three rows below, Kabwe, Petralona and Kalambo, were worked again by hand after the sources were re-read, and their inputs differ from the script's;  the note says how.

The table sets the spread inside the mainstream literature for each site against the gap between the lowest figure in that spread and a Flood date.  The Flood date is the model's own.  Mitchell puts the Flood "about 4,350 years ago";¹⁰ Walker writes that it "ended in real time about 4,500 years ago";¹ Brian Thomas of the Institute for Creation Research has it "roughly dated to 2,500 B.C. by standard Bible chronology".²³  The longest figure read for this article is Whitelaw's, "close to 4,950 years ago", written in 1970.¹⁵  That is the longest read here, and not necessarily the longest any young-earth writer allows.  The table uses 4,350 years, because the young-earth writers quoted in Part 4 put these fossils after the Flood, and the paragraph under the table gives the same ratios against 4,950.  Whitelaw is the exception among the writers read:  on his own conversion table a radiocarbon reading of 9,000 years is a true age of a little over 5,500, which is before his Flood.¹⁵  "Lowest" and "highest" are the lowest and highest numerical ages published for the fossil or its layer that this article's research found, whatever the method and whether or not the authors call the figure a minimum or a maximum;  "accepted" is the figure the current primary paper prefers.⁵

Three things to read off it.  Where a site has a set of determinations its authors accept, they agree within a factor of 1.1 to 1.6:  Mauer, the Sima cluster, Gran Dolina, the measurements on the Kabwe skull, Bodo, Kalambo.  Counting every age ever published for a fossil or its layer, including those the authors themselves set aside as minima, maxima or open-system, the spread is 2.0 for Schöningen, 5.0 for Harbin, 5.4 for Petralona, 10 for Sima, where human bone averaged 148 thousand in 1997 and quartz gave up to 1.5 million in 2014, and 12.5 for Kabwe.  And every measured figure, the lowest included, stands 14 to 200 times above 4,350 years.  Against Whitelaw's 4,950 years the same figures stand 12.5 to 174 times above it.  In logarithms, the accepted sets disagree internally by 0.04 to 0.21 of an order of magnitude, the all-comers lists by 0.3 to 1.1 orders, and the lowest measured figure at each site stands 1.2 to 2.2 orders from 4,350 years.  "Even the scientists disagree" is true, and the disagreement is tens of percent, or 12.5-fold in the worst case, against a required factor of 14 to 200.

The one row that breaks this pattern is Kabwe with every published figure counted.  Its lowest entry is the 40 thousand years of the 1960s, which is 9.2 times 4,350 years and 8.1 times 4,950, so the published history of that skull is wider than its distance from a Flood date.  That 40 thousand was a correlation with other sites and not a measurement.  The measurements on the skull itself span a factor of 1.5.

Part 4: What the young-earth sources say about these dates, in their own words

Before any of this is priced against a young-earth chronology, that chronology has to be stated as its authors state it, and about which fossil.  Only two young-earth pieces found for this article engage a specific site's dates, and only one puts a number on a fossil.

Troy Lacey of Answers in Genesis, on Harbin, after reporting the 138 to 309 and 146 thousand-year figures correctly:  "But as mentioned many times on our website, there are many assumptions built into radiometric dating and radioisotope isochron dating methods.  Compared against known dates, the dates yielded are faulty.  This skull, in reality, belongs to a post-Babel human, roughly between 2200–1800 BC."²⁴  Mitchell, on Sima:  "We find genetic evidence that people were related because they all descended from the 8 people that disembarked from Noah’s ark after the global Flood, about 4,350 years ago.  Denisovans and Neanderthals and the people of Sima de los Huesos (whether they are called Homo heidelbergensis or something else) were simply humans who lived in the post-Flood world and left their fossilized remains in Ice Age sediment."¹⁰  Thomas, on Schöningen:  "Since they were found in coal deposits associated with the Ice Age, and since people required some time to migrate from Babylon to modern Germany, this places these artifacts relatively soon after the Flood, which is roughly dated to 2,500 B.C. by standard Bible chronology."²³  Walker, on Kabwe:  "So, we would interpret this skull as that of a human descended from the eight people that came off Noah’s Ark (Genesis 8:18–19).  Over hundreds of years, as the population grew after the Flood, some people would have migrated to the area in Zambia (Genesis 10:32)."¹  Thomas and Jeffrey Tomkins, on Sima:  "Overall, age assignments for Sima de los Huesos human bones have ranged from 148 ka to 500 ka, and seem to depend on which evidence researchers subjectively choose to include."²⁵  Walker again:  "The issue of dating is confusing to many people because they do not appreciate that researchers can essentially get any ‘date’ they like depending on what sort of date they are looking for."¹  Line, across five pieces, passes every date with "allegedly" or "supposedly" and offers no young-earth age for any site.

So the young-earth position is that every fossil in this article is post-Flood and post-Babel, within a window of centuries.  Lacey's window for Harbin is 400 years wide.  Nobody on that side says which input to which measurement is wrong, or by how much.  Walker says twice that he had not read the paper:  "I have not read the papers associated with this latest report, and so I am not familiar with the details of the dating process they used."¹  His one technical suggestion, that lead in the ore body "would have affected the isotopic composition of the coating, affecting the calculated ‘date’",¹ misdescribes the method:  the date is not on a coating and is not a lead method;  it is uranium-thorium on the bone of the skull base.  To his credit, his closing line is an invitation:  "to properly understand the reasons for these differences, it would be necessary to understand how each date was decided, including the methods and assumptions.  This would require digging out the relevant scientific papers and supporting documents for all the different estimates.  I expect that some researchers will do that, and it would be nice if some creationist researchers would check it too."¹  This article is a first attempt at that check, and the result is above.

Two scale figures follow, and they are this article's own.  No creationist source states either.  The mainstream Middle Pleistocene spans 645 thousand years.  Packed into the 4,350 years since Mitchell's Flood, that is a compression of 148-fold;  packed into Lacey's 400-year window, which he gives for one skull and not for the interval, it would be about 1,600-fold.²⁶  They are printed for scale only.  The honest young-earth claim is narrower than either:  that each fossil, individually, is a few thousand years old, which on the table above needs each measured age to be wrong by a factor of 14 to 200.

Part 5: The order of the layers, and the one assumption everything shares

Two further points, one for each side.

At Mauer and Kalambo the dates come out older below and younger above, with one pair of samples out of order at Kalambo⁴ and one pair that overlaps across the boundary between the two sands at Mauer, 508 ± 50 above against 502 ± 27 below.²  Superposition is common ground.  A sequence laid down over decades or centuries after the Flood is also in order, and any method whose signal grows with time, at whatever rate, will reproduce the order.  So stratigraphic order alone does not tell the two chronologies apart.  Only the size of the intervals does.  One young-earth writer read for this article affirms the order of the layers for these very fossils.  Mitchell:  "Homo heidelbergensis fossils are found in layers dated by evolutionists to around 600,000 years.  Neanderthals appear above that level, and only early modern humans are preserved in the layers above these."⁹  None read says anything about the order of the dates within one site.

The other point is what "independent" means, and it has to be tested against the young-earth proposal as its authors state it.  That proposal is the RATE project's:  radioactive decay ran much faster in the past, in episodes.  If so, every method that counts decay would read too old together, and their agreement would prove nothing about the rate.  So the methods in this article have to be sorted by whether they count decay.

Most of them do.  Uranium-series, at Sima, Kabwe, Harbin and Petralona, counts the decay of uranium into thorium.  Argon-argon, at Bodo, counts the decay of potassium.  Radiocarbon counts the decay of carbon-14.  Luminescence, radiofluorescence and ESR, at Mauer, Sima, Gran Dolina and Kalambo, do not count decay in the sample, but they divide an absorbed dose by a dose rate, and Wagner's team says where that dose comes from:  "The radiation field is provided by the decay of natural radioisotopes and cosmic rays."²  At Sima the tables show how much is cosmic:  0.01 out of about 2 grays per thousand years, half of one percent.¹¹  The rest is decay.  So the two methods at Mauer, which the authors call independent, are independent in what they date, sand and teeth, and in how they are measured.  They are not independent of the decay rate.  The same is true of the three at Sima.  A faster decay rate would move them all, and their agreement does not test the RATE proposal.

Three kinds of evidence in these papers do not count decay.  The first is the order of the layers, which has just been dealt with:  it gives order and no years.  The second is magnetic polarity.  The sediments at Mauer and Sima are magnetised the normal way²¹¹ and the antecessor layer at Gran Dolina the reverse way,²⁰ which puts Gran Dolina before the last reversal and the other two after it, and the dose-based ages agree with that order.  But the reversal's age in years, 772 thousand, comes from outside these papers, and how it was fixed was not read for this article.  The third is amino-acid racemization, at Schöningen and for the Kabwe fragment.  It is chemistry and no decay rate touches it.  But it gives years only after it is tied to something already dated, the ocean climate record at Schöningen and a radiocarbon-dated bone at Kabwe.  So no site in this article has a method that yields an age in years with no reliance, direct or borrowed, on a radiometric or astronomical timescale.  What the decay-free evidence adds is agreement about order.  That is the stronger kind of agreement, and it is also the kind that does not separate the two chronologies.

That leaves the size of the change RATE would need, and here RATE's own words set limits.  It places the fast decay early:  "a large amount of nuclear decay occurred during the Genesis Flood and during the early part of Creation week as well".²⁷  It leaves one door ajar.  The same chapter says some of the radiohalos it attributes to fast decay formed "during and immediately following the Genesis Flood", and it marks the youngest granites "post-Flood(?)" with a question mark of its own.²⁷

Now take the sand at Sima.  It lies on the human bones.  The method assumes the grains were emptied of their signal by daylight before they were washed into the cave, and the authors say plainly that this is an assumption:  the windblown source sediments "are likely to have received prolonged and direct sunlight exposure prior to entrainment and transportation into the cave".¹¹  If that is granted, the dose in the grains was absorbed after they last lay in the sun.  The young-earth writers quoted in Part 4 place the people of Sima after the Flood, so on their placing the sun last shone on that dust after the Flood too.  The tables give the dose, 700 to 970 grays.  To build it up in 4,350 years takes 0.16 to 0.22 grays a year, 91 to 101 times the rate measured there now, and in 4,950 years 80 to 89 times.²⁸  The gamma rays alone, the penetrating part of that dose, would come to 60 to 78 thousandths of a gray a year, against 0.62 to 0.77 today.

Two things can fairly be said for the young-earth side here.  The amount of extra decay needed is small on RATE's own scale.  RATE argues that "1.5 billion years worth of U decay, at presently measured rates of U decay, has actually occurred within the last 6000 years".²⁷  The 430 thousand years' worth needed at Sima is about one part in 3,500 of that, so a short tail of fast decay after the Flood would cover it in bulk.  And if the grains were not fully emptied before burial, the luminescence ages are too old by some unknown amount, as the authors allow.  Against that, the tail would have to fall after people had reached Spain, nobody has proposed it for this or any site in this article, and a shorter tail means a higher dose rate while it lasted.  Whether those dose rates are survivable was not worked out here, because no figure for a harmful dose was sourced.

RATE also does not propose that every kind of decay sped up alike.  Its summary says the group "began to look for a theoretical mechanism of accelerated decay which would produce variable effects dependent upon half-life or possible decay modes".²⁷  That matters at Sima.  The uranium-series age on the calcite, 434 thousand years, rests on two isotopes with half-lives of some hundreds of thousands of years.  The luminescence ages, 416 and 433, rest on the dose from uranium, thorium and potassium, whose half-lives are billions of years.  A speed-up that differed by half-life would pull those two kinds of clock apart unless the factors happened to match, and at Sima they agree within 5 percent.  This is a real constraint and a modest one.  It cannot be made exact without RATE's formula, which the chapter read here does not give.

Objections and responses

"Even the scientists disagree, so the dates are worthless."  They disagree by tens of percent, or at worst tenfold, and the tenfold case is one where the authors themselves had already rejected both ends.  A young-earth reading needs every measured age to be wrong by a factor of 14 to 200, and in the same direction.  No published disagreement among the dating specialists is of that size or that shape.

"The Sima date was changed to fit the model, which proves dates are chosen."  The 2007 date was dropped after Stringer called it erroneous on morphological grounds, and that is on the record.  But his stated challenge was to which layer lay on which, the 2014 revision was a stratigraphic re-reading, and the replacement rests on three methods on three materials that agree within 5 percent.  The change was 20 percent.  Thomas and Tomkins's "ranged from 148 ka to 500 ka" is true only by counting a figure its authors rejected in the sentence they reported it in.

"Two methods agreed at Sima in 1997 and were both wrong, so agreement proves nothing."  Half right.  Two methods on the same bones that both depend on when uranium entered them are not independent, and the article says so for Kabwe and Harbin too.  But the error ran young, because bone takes up uranium late, and every revision since has gone up.  A method that undershoots does not help a reading that needs it to overshoot a hundredfold.  Mauer is the counter-case:  sand grains and tooth enamel record different events and agree.

"Harbin's date rests on hearsay provenance, so it should be thrown out."  Its provenance is hearsay and its date is the weakest in the article.  Even so, the lowest apparent age on the skull itself is 62 thousand years, 14 times the 4,350 years used in the table.  The authors do not throw that figure out.  They explain it as tooth dentine that took up its uranium late, and they treat it as a floor.²¹  The figures they do throw out are the old ones, up to 296 thousand years.²¹

"The Kabwe skull is unmineralised and the Kalambo logs are still wood, which proves they are young."  Woodward's observation is real and the logs are real, though the logs are not unaltered:  the excavators report "partial mineralization of the wood (silica)".⁴  No measured rate of decay or mineralisation, for bone in a lead-zinc cave or for waterlogged wood, was found in the papers read for this article, from either side.  The mainstream answer, unusual chemistry and unbroken waterlogging, is stated in those papers and not measured in them.  The young-earth answer is also stated and not measured.

Radiocarbon has been tried on both, and neither result gives the young-earth reading what it would need.  The Kalambo wood gave no measurable carbon-14, "infinite (more than 50 kyr ago (ka)".⁴  Wood a few thousand years old that grew in air like today's would still hold more than half of its carbon-14 and would give a finite age at once.  A result beyond 50 thousand years means less than about one part in 400 to 500 of the modern level is left.²⁹  A young-earth reply exists in outline.  The RATE project holds that the starting level of carbon-14 cannot be taken as today's, because "the Genesis Flood, by removing so much C from the biosphere, would invalidate the crucial assumption".²⁷  Whitelaw's paper corrects radiocarbon ages on a related ground, a change in the rate at which carbon-14 was produced.¹⁵  Applied to the Kalambo logs, that reply would have to supply a starting level below about half of one percent of today's,²⁹ in wood that grew after the Flood if it is placed where that school places the people of this interval.  Nobody has made that claim about this site.  Kalambo is not named in any article that the search indexes of creation.com and the Creation Research Society Quarterly return.  At Kabwe the only radiocarbon result found is on animal bones said to be from the cave, and it is an open-ended "older than 9,000 years", which excludes nothing above that.¹⁴  No radiocarbon measurement on the skull was found.

Objections about the fossils and their names are answered in the parent article, and objections about the young-earth literature in its article.

Issues and unanswered questions

• No young-earth age for any fossil in the interval except Harbin, and no faulty input named for any measurement. Lacey gives Harbin 2200 to 1800 BC. Nobody says which input to which method is wrong or by how much. The most promising young-earth target is the shared decay-rate assumption, which rescales every method together, and its price is a factor of 34 to 198 held for centuries inside human history, with the RATE model placing accelerated decay before and during the Flood rather than after.
• The luminescence constraint has not been priced. Sand at Mauer, Sima and Kalambo was zeroed by daylight at burial, so its dose is post-Flood on a young-earth reading. The dose rate that would deliver 476 thousand years of signal in a few centuries can be computed in sieverts per year from the dose rates in Wagner's and Arnold's tables. A dune case elsewhere on this site found the implied field weak rather than decisive; it should be redone with these sites' numbers before the article leans on it either way.
• Why the dates come out in the order of the layers. Common ground, and unclaimed. No young-earth source states that stratigraphic order alone does not discriminate between the chronologies and that only the sizes of the intervals do.
• Unmineralised bone and unmineralised wood. No measured rates either way. Whether radiocarbon was ever run on the Kalambo wood or the Kabwe bone was not checked.
• Sources not read in full. Clark 1994, so which tuff was dated and how the Bodo find spot ties to it is not stated here; Bischoff 1997; the published Nature text of Grün 2020, from which every manuscript quotation above needs checking; and Thieme 1997.

Verdict

The muddle does nothing to the dates.  Where the specialists disagree, they disagree by tens of percent, or tenfold at the worst-documented site, against a factor of 14 to 200 that any reading of a few thousand years requires.  The one site where two methods agreed and were both wrong undershot, and every correction since has gone up.  The young-earth literature on this interval offers a chronology of centuries, no faulty input for any measurement, and, by its own best writer on Kabwe, an admission that the papers had not been read.  On the fossils the two readings predict the same thing.  On time, only one side has published an answer, and it is not the young-earth side.

What the naming dispute does to the arguments built on the fossils, in both directions, is the verdict of the parent article.  The young-earth literature's own account of these fossils, and where it misstates what it cites, is the verdict of the creationist-literature article.

Sources