HistoricalChristian.Faith

Are this reconstruction's witnesses securely dated, or could they be much older than assumed?

Type: supporting argument (off-timeline) · Assessment: weak as a blanket claim — but the honest answer is method-dependent, and this page reports it method by method rather than defending every date uniformly · advanced by young-earth and alternative-chronology writers

Related arguments: out-of-place artifacts (dates under skepticism) — see also whether the master calibration spine is the compressible weak point (the systemic form of this page's radiocarbon section) · Liangzhu's jades and the technology they required · the age of the Kailasa temple at Ellora · the "lathe-turned" pillars of the medieval Deccan · the vimanas and divine weapons of the Sanskrit epics · the Ta Prohm "stegosaurus" carving · Angkor's engineering and the antiquity of the Khmer capital · predynastic Egypt's hard-stone vases · the Great Sphinx's water-erosion redating · the Osireion at Abydos · the Giza radar scans and the "second Sphinx" · Sacsayhuamán's polygonal walls

This is the methodological backbone of the reconstruction. Every other argument on that timeline turns on a date, so the question of what the dates are actually worth has to be asked directly and answered without flinching in either direction. The temptation runs two ways. One is to settle the matter by authority — the field says 9600 BC, therefore 9600 BC — which is not an argument. The other is to unseat a date on bare possibility — a method has assumptions, therefore the artifact could be any age at all — which is not an argument either. What follows names each dating method the timeline leans on, states what it establishes and what it does not, and tests the "much older than assumed" reading against the specific evidence at specific artifacts. It asks whether the methods are reliable. A separate challenge grants that they are and concedes every measurement, then asks whether what they clock is elapsed time at all or an authored appearance of a past that never happened — the "mature creation" / apparent-age move, weighed on its own page. The two are mirror halves of the same worry about deep dates, and they fail in different places.

The claim

The claim comes in two families, which need to be kept apart because they fail and succeed in different places.

The first is radiometric skepticism, the young-earth case that the physical dating methods rest on unproven premises about the past. The Institute for Creation Research puts it plainly: "the calculated dates will only be accurate if the assumptions behind the method are correct," among them "that the ¹⁴C/C ratio has remained unchanged for tens of thousands of years" and "that nuclear decay rates have always been constant."¹ Answers in Genesis makes the same move about the atmosphere: the method assumes that "if carbon-14 has formed at a constant rate for a very long time and continually mixed into the biosphere, then the level of carbon-14 in the atmosphere should remain constant" — an assumption the article denies holds strictly.²

The second is monument antiquity: that particular built or carved objects are far older than the dynasty credited with them. Its best-argued instance is Robert Schoch's redating of the Great Sphinx, first "before 5000 BC," later pushed to a minimum near 9700 BC, on the ground that "the most prominent weathering pattern was caused by prolonged and extensive rainfall" rather than by the wind and sand that mark the rest of the Giza plateau.³ Applied to this timeline the two families converge on one reading: the conventional chronology is not secure, so some of these artifacts and strata could belong to a horizon the reconstruction cares about — before the Flood, or in the age of the Nephilim. That reading is what gets rated here.

What would have to hold true

The steelman is not "science is guessing." It is that dating is inference, that every method has a stated set of conditions, and that the conditions are sometimes not met — which is not a scandal but an ordinary working fact of the disciplines themselves, freely conceded in their own handbooks.⁴ Four mechanisms genuinely do move dates, and the argument is entitled to all of them.

Usurpation. A ruler claiming a predecessor's monument as his own is not a fringe hypothesis; it is a documented Egyptian habit with a scholarly literature. Peter Brand's survey lists a sphinx of Amenemhet III re-inscribed three separate times — by Ramesses II on the base, Merenptah on the shoulder, Psusennes I on the chest — and notes that under Ramesses II the surcharging of statuary and reliefs became widespread, motivated by appropriation rather than by any wish to erase the earlier king. Deep-cut and body cartouches appear in the same period as a countermeasure against the practice.⁵ A dedication is therefore not automatically a construction date.

Renovation read as construction. The same inscription that records a king "building" a temple may record his rebuilding, enlarging, or re-consecrating one already standing — and the two are frequently indistinguishable in the formula used.

Reused foundations and spolia. A later structure sitting on earlier courses dates its own walls, not the platform beneath them.

Circular stylistic dating. Where a monument is dated by its style and the style is dated by monuments, the reasoning closes on itself. This is a real hazard and it is why the sections below ask, at each witness, what the independent anchor is.

Method by method

Epigraphic and textual dating

The most direct evidence there is: contemporary writing that names the thing and the year. Where it is present the "much older" path is very hard to walk, and this is where the argument does worst.

Take the three Indian monuments most often raised in this connection. The Gommateshwara colossus at Shravanabelagola — a 17.5-metre monolith — carries inscriptions at its own base naming Chavundaraya, general and minister of the Ganga king Rachamalla, who commissioned it, with the consecration placed at 981 by the majority of the epigraphers; the hill holds hundreds of further rock inscriptions running from about the seventh century onward.⁶ Vijayanagara (Hampi) is anchored twice over: by its own imperial epigraphy and by foreign visitors who stood in it and wrote it down — Abdur Razzaq on embassy from Shah Rukh in 1443, Niccolò de' Conti earlier, Duarte Barbosa, Domingo Paes about 1520 describing the Virupaksha temple and its colonnaded bazaar and calling the city "as large as Rome," Fernão Nuniz in the 1530s, and Caesar Frederick after the 1565 battle that ended it.⁷ An eyewitness who describes a working city is not a stylistic inference; nothing in the OOPArt toolkit reaches it.

The fullest worked example of that convergence is Angkor, and it is set out on its own page because the antiquity claim there has been pressed seriously and deserves a full hearing. It is the useful control at this end of the range: dated foundation stelae from 881, Old Khmer labels cut on the gallery wall of Angkor Wat naming the king and his commanders, a reservoir dike lying on top of a temple that is itself dated, a Chinese envoy resident in the working capital in 1296–97, and two airborne laser surveys whose result was more of the same medieval city rather than an older one beneath it — with the honest concession, made there, that Angkor Wat has no foundation stele of its own and that laser scanning is a mapping method which dates nothing.

Kailasa at Ellora — the monolithic rock-cut temple, and the case the argument presses hardest — is the honest one, and it deserves its concessions; it is examined at length on its own page. The temple has no dedicatory inscription. Its attribution to the Rashtrakuta Krishna I (r. c. 756–773) rests on later copper-plate grants: the Vadodara plates of Karkaraja II, of about 812–813, which name Krishnaraja as the patron of Kailasanatha and a Shiva temple at Elapura and say the king built something so wondrous that the gods and the architect were astonished, and a Kadaba grant that credits him similarly. Those epigraphs are not physically attached to the caves, and the land grants of Krishna's successors do not mention the temple at all. On stylistic grounds Hermann Goetz argued in 1952 that work began under Krishna's predecessor Dantidurga, whose inscription is in the adjacent Cave 15.⁸ So the attribution is an inference, the construction may well span reigns, and the century is not fixed to a year.

But notice what that concession is and is not. The live uncertainty at Kailasa is a few decades within the eighth century — which king, which reign, how many reigns. The reading being rated needs thousands of years, and the mechanisms that would supply them do not transfer here. Usurpation works on a free-standing built monument: a later king carves his name over an earlier one's. Kailasa is not built; it is excavated, a temple subtracted from living basalt top-down, with no prior structure available to appropriate and no foundation courses to reuse. Reused foundations and spolia have nothing to act on. And the temple sits in a dense, dated regional context — the Pallava Kailasanathar at Kanchipuram and the Chalukya Virupaksha at Pattadakal upstream of it stylistically, Rashtrakuta epigraphy attesting Deccan temple programmes and sculptors' guilds around it, and neighbouring Ellora excavations of comparable scale.⁹ An eighth-century Kailasa has ancestors and cousins; a five-thousand-year-old Kailasa would be a monolithic rock-cut Hindu temple with no Hindu temples on either side of it for millennia. Verdict here: the epigraphic path is weak, and it is weak on the evidence, not by deference.

The same method also carries Berossus and the Sumerian King List — where, notably, the wiki has already reported the method against the reconstruction's interest: the Weld-Blundell prism is dated to a named regnal year of Sîn-māgir of Isin, and the Ur III recension that begins at Kish with no flood is dated by its own copying context. The dating there is not the weak link; the arithmetic of the reigns is.

Radiocarbon

The method's real limits are the ones its practitioners publish, and they are narrower than the argument needs. Three matter here.

It dates the sample, not the act. The old-wood effect — dating the inner rings of a long-lived tree, or seasoned or reused timber — can put a charcoal date decades or centuries before the event that burned it, and charcoal fragments often carry a time-width that cannot even be quantified. Reservoir effects shift marine and freshwater-fed samples older by several centuries. Both are systematic, both are known, and both are handled by sample selection and reported correction.¹⁰

Atmospheric ¹⁴C has not been constant. This is the one point in the young-earth brief that is simply correct — and it was established by radiocarbon scientists, which is why calibration curves exist at all. IntCal20 is built from tree rings back to about 13,900 cal BP and from floating chronologies, speleothems, corals and lake and marine sediments beyond, with an additive over-dispersion term in its Bayesian statistics and modelled ocean-circulation reservoir ages replacing fixed regional offsets. Successive revisions do move older ages by up to a few centuries. They do not move them by orders of magnitude, and the direction of travel is toward precision, not away.¹¹ The accelerated-decay premise is a different and much larger claim, and it is not conceded here: it is disputed inside creationism as well as outside it, on the internal ground that decay fast enough to do the work would leave no ¹⁴C at all.

Context can invert a sequence. The timeline's own entry for Göbekli Tepe already states this against itself: the enclosures were deliberately packed with rubble at the end of their use, so charcoal from the fill dates the burial rather than the building and can reverse the order of construction. The excavators say so in print. What actually anchors the early end there is wall plaster from Enclosure D at 9984 ± 42 ¹⁴C BP — a determination tied to a standing structure — with the late end set typologically by projectile points.¹² That is a real weakness of a real kind, and it is worth naming exactly what size it is: it moves the site by centuries inside the Pre-Pottery Neolithic. The post-Flood placement argued for the site needs it moved by seven thousand years, and no radiocarbon pathology on offer does that.

Luminescence (OSL and TL)

Luminescence measures accumulated dose since a sediment's last daylight exposure or a ceramic's last firing, divided by an estimated dose rate — so it carries two independent chains of uncertainty. Partial bleaching is the dominant accuracy problem in water-laid and colluvial deposits, and the usual workaround, isolating the lowest-dose population, lacks agreed objective criteria; feldspar adds anomalous fading, which biases ages young unless corrected. The method's own review literature states all of this, along with minimum reporting requirements.¹³ Two observations follow for this timeline. First, luminescence uncertainties are typically percentages of the age, not factors of it. Second, the direction of its best-known bias is the wrong way for the argument: fading makes things look younger than they are, which would help the antiquity claim, and it is corrected for precisely because the field knows it. Where luminescence carries weight on this wiki — the Lingjing ages behind Homo juluensis — the genuine scholarly spread is a factor of about three, and the identification would need a factor of twenty.

Stratigraphic and relative dating

Stratigraphy establishes order rather than years, and its assumptions fail where deposits are mixed, disturbed, or re-cut. But it is the method that gives the Sphinx argument its sharpest test, and the test is instructive. Schoch's case is a genuine observation — the enclosure's deep vertical fissured profile is not the horizontal wind-and-sand pattern of its neighbours — and it was not answered by consensus but by rival mechanisms and by section-drawing. Harrell attributed the profile to haloclasty, salt crystallisation driven by moisture in the burying sand, in patterns visible in comparable Old Kingdom quarries; Gauri favoured the same process driven by atmospheric moisture; Lehner's stratigraphy has the Sphinx Temple built on the pre-existing northern enclosure wall of Khafre's Valley Temple, which fixes the monument's position in a construction sequence rather than in a weathering argument; and the seismic sub-surface anomaly turned out to admit reading as a natural bedding plane, with velocities behind the Sphinx that on Schoch's own logic would make the back older than the sides.¹⁴ The Sphinx redating has its own page, where the geomorphology itself comes out genuinely contested — Schoch's rainfall reading of the enclosure was reached independently by Colin Reader and David Coxill — while these added stratigraphic and seismic props lose on the evidence, and the pre-Flood scale fails regardless, because the very geologists who accept the water erosion date it only to the Early Dynastic, a few centuries before Khafre, not to a lost world. That is what a genuinely open dating dispute looks like when the opening still will not carry the conclusion — not a challenge dismissed for being unorthodox.

Palaeographic dating

Here the argument is right, and the timeline says so before being asked. The Sumerian flood story survives on one broken tablet, CBS 10673, with no colophon; its editor Miguel Civil dates it by handwriting alone and can say only that it is "not earlier than Late Old Babylonian."¹⁵ Handwriting dating has no internal error bar, is calibrated against other undated hands, and is the closest thing on this shelf to circular stylistic reasoning. The summary already flags the placement as "softer dating, not a claim that the Sumerian tradition is younger." The concession is real — and it is worth being precise about what it buys. A tablet whose date is soft by a century or two is a text about a deep past whose own composition date the timeline never used as load-bearing evidence anyway; loosening it does not relocate a flood.

Molecular-clock and genetic dating

This is the reconstruction's most assumption-laden date, and the page that uses it should say so. Coalescent inference depends on a mutation rate, a generation time, and a demographic model, and the confidence intervals are wide. The timeline's own showpiece illustrates the point: Wangjie Hu and colleagues inferred a severe ancestral bottleneck at c. 930,000–813,000 years ago to roughly 1,280 breeding individuals — and Yuxin Deng, Rasmus Nielsen and colleagues argue in Genetics that it is likely a statistical artifact, with a second group failing to recover the signal using an independent method on the same site-frequency spectrum.¹⁶ So: genuinely open, and reported as open on the bottleneck page and on the modern scientific record entry alike. What the openness does not do is supply the argument's conclusion. The disputed date is a thousand times older than the model needs. Loosening a method that has produced no result in the required window does not produce one.

A worked example: the peopling of the Americas

The clearest illustration of how a date is actually stress-tested — and of the symmetry this page insists on — is not on the shelf at all but in the neighbouring question of when people first reached the Americas, where the disputed direction runs the other way. Young-earth readers want these Pleistocene dates younger, not older; the machinery that tests them is the same in either direction, and it is worth watching it work.

For most of the twentieth century the field held a "Clovis-first" consensus — no people in the New World before about 13,000 years ago. It was overturned not by outsiders but from within, by Tom Dillehay's excavation of Monte Verde in southern Chile, a waterlogged site whose organic preservation is exceptional, dated to roughly 14,500 years ago; in 1997 a panel of archaeologists that included prominent Clovis-first defenders inspected the site and accepted its pre-Clovis age.¹⁷ A discipline unseating its own reigning model on the evidence of a single well-preserved site is the opposite of the dogmatic consensus the blanket argument imagines.

The White Sands footprints run the test the other way. Matthew Bennett and colleagues dated human tracks in New Mexico to about 23,000–21,000 years ago — deep in the Last Glacial Maximum, far earlier than Monte Verde — by radiocarbon on seeds of the aquatic plant Ruppia cirrhosa.¹⁸ The objection came at once, and it was exactly the one the radiocarbon section above names: aquatic plants can take up old dissolved carbon — a hard-water reservoir effect — so the seed dates might read centuries or millennia too old. Mainstream researchers pressed the point in print.¹⁹ The answer was not assertion but two further methods run on the same footprint-bearing layers: radiocarbon on terrestrial conifer pollen, which fixes its carbon from the air and so escapes the reservoir problem, and optically-stimulated luminescence on quartz grains, which dates the sediment's last exposure to daylight and shares none of radiocarbon's assumptions. Both landed in the same window as the seeds.²⁰

That is the shape of this whole page in miniature. A single method with a known, published limitation is challenged on exactly that limitation; the challenge is answered by convergence rather than by authority; and the reading that needs the date moved — older at White Sands, younger on a young-earth chronology — has to explain not one method but three agreeing with each other. The dates here are increasingly multiply-anchored, which is why the "much younger" reading is weak on the same evidential terms the rest of this page applies to the "much older" one. The symmetry is the point: the method resists the pull in both directions.

The biblical yardstick, and where a soft date actually bears

One question this shelf invites has been left implicit, and it is the one on which the argument might seem to gain a real footing, so it is worth making explicit. Every artifact here is being judged against a biblical Flood horizon, and this wiki's working yardstick for that horizon is not Ussher's 2348 BC but the Septuagint's long chronology — the Flood near 3300 BC, Creation near 5500 BC — which it rates the stronger of the two biblical reckonings. So the fair form of the "much older" case is not the blanket one already weighed but a narrower and better one: never mind the deep sites; is the dating soft at the Flood horizon itself, right where the yardstick is laid down? Here the honest answer concedes more than anywhere else on the page — before it turns out not to bear.

The concession. There is a genuine flat stretch in the radiocarbon calibration curve at almost exactly the Septuagint Flood window. Around 3350–3100 BC past atmospheric ¹⁴C behaved such that a spread of real calendar years all return nearly the same radiocarbon age, so a single determination there smears across more than two centuries — the late-fourth-millennium plateau, sitting squarely on the horizon the yardstick cares about.²¹ That is the strongest thing the "not securely dated" case can truthfully say on this whole shelf: not that a deep site is secretly young, but that dates at the Flood horizon are, for one real and specific reason, genuinely fuzzy — and fuzzy in the very place a Flood is being sought.

Why it still does not bear. Three things, and they are the discipline the rest of the page already applies. First, a plateau is imprecision, not bias: it widens the error bar, it does not slide the curve. An isolated date on it is uncertain across its own ~250-year span and no further — and stratigraphic sequence pulls even that back to decades — so it cannot relocate anything by the millennium a pre-Flood or antediluvian reading needs; the fuzz is real and the wrong size. Second, the direction of the marquee dating disputes runs against the argument, not for it. Where radiocarbon and a documentary chronology genuinely collide — the textbook case is the Thera eruption, radiocarbon dating the buried olive tree to 1627–1600 BC against an Egyptian-synchronism date near 1550–1480 BC — the isotope runs about a century older than the history, deepening antiquity rather than compressing it.²² A slack that mostly pushes dates older is the wrong tool for pulling a horizon nearer. And third, where the horizon actually has to be tested it is not soft at all: the continuous third-millennium sequences of Egypt and Mesopotamia are radiocarbon-anchored to within a few decades, independently of the king-lists, so even at the Flood window the record runs unbroken across it on a physical clock the plateau does not touch.

Two regimes, and the honest split. This is why the yardstick has to be laid down with care. For the deep artifacts this page mostly weighs — Göbekli Tepe near 9600 BC, the Sphinx enclosure on any reading — the choice of biblical reckoning is moot: they sit before both Creation dates, and the absolutely-dated tree-ring spine that underwrites the whole curve runs older than even the Septuagint's Creation, as the calibration page sets out. The yardstick becomes load-bearing only for artifacts at the Flood horizon itself — and there, as just shown, the dating is tight. So the Septuagint does real work on this timeline, but not here: where it is genuinely load-bearing is the opposite move to this page's — the longer post-Flood window that lets an antediluvian craft-tradition survive through the Flood into a post-Flood heir, weighed on the Liangzhu-jade page. That is a claim about chronology, not about dating security. The slack this page concedes cannot be converted into support for the long chronology; to treat "the dates are soft" as though it showed "the Septuagint horizon is right" would launder a fit the evidence does not give — the very move the yardstick exists to prevent.

Adversarial assessment

What holds. Three things, and they are not small. Atmospheric ¹⁴C really has varied, and the calibration curve is the admission. Radiocarbon really does date the sample rather than the act, and at Göbekli Tepe the backfill really does mean that fill charcoal can invert a sequence — the excavators publish it. And palaeography and molecular clocks really are single assumption-laden methods carrying weight alone, at the Eridu Genesis tablet and at the Hu bottleneck respectively. A reconstruction that defended every date on this shelf uniformly would be doing apologetics. These three are not concessions extracted; they are the honest result of asking the question.

What is overstated. The argument moves from localized uncertainties to a blanket conclusion, and the move does not survive inspection. Every real pocket of softness listed above is either the wrong size or points the wrong way. Palaeography is soft by a century or two, not by millennia. Luminescence's best-known bias runs young. The disputed genetic date is three orders of magnitude too old to help. The Göbekli Tepe backfill moves the site inside its own Neolithic, and the argument needs it moved out of the Neolithic entirely — into centuries in which a literate, bronze-using Ur III state stands five hundred miles downriver from a site with no pottery, metal, or writing.

The deeper failure is that the mechanisms which genuinely unseat dates are artifact-specific, and the argument uses them as though they were general. Usurpation is a strong tool against a free-standing built monument and has no purchase on a monolithic rock-cut excavation with nothing prior to appropriate, or on portable grave goods in a stratified sequence — which is exactly why the antiquity sub-claim also fails at Liangzhu. Renovation-as-construction bites where the formula is ambiguous; it does not touch a Persian ambassador describing a city he walked through. And where converging independent methods anchor a date — speleothem uranium–thorium alongside radiocarbon at Liangzhu, radiocarbon alongside typology and stratigraphy at Göbekli Tepe, seed radiocarbon met by pollen radiocarbon and quartz luminescence at White Sands — the argument would have to explain not one method's failure but several failures agreeing with each other.

Symmetry, stated. None of this is settled by counting authorities. The Sphinx case is not refuted because Egyptology dislikes it; it is refuted by a rival weathering mechanism, a construction sequence in section, and a seismic reading that cuts against its own proponent. The Kailasa attribution is not secure because the field asserts it; it is an inference from grants issued forty years later, and the honest report is that its century is well anchored and its reign is not. Where the wiki's own evidence is thin, this page says thin.

And not by miracle. The wiki grants that God can act in history, and this timeline says so at the top. That grant does not relocate a stratum, reset an isotope ratio, or unwrite an inscription. Dating is weighed on dating evidence; a supernatural premise may be argued elsewhere on its own terms, and importing it here would be exactly the special pleading this page exists to refuse.

Assessment

Weak as the blanket claim — with the differentiation itself as the finding, and it is the finding worth carrying away. Dating security on this shelf is method-dependent, and it is high precisely where the argument most needs it to be low. Contemporary writing (Shravanabelagola's base inscriptions, Vijayanagara's epigraphy and its eyewitness chroniclers, Angkor's dated stelae and its resident Chinese envoy) and converging independent methods (Liangzhu's speleothems with its radiocarbon; Göbekli Tepe's plaster date with its typology and stratigraphy) resist the reading outright. The genuine pockets of uncertainty are conceded and named — the Kailasa attribution's reliance on later grants, the Göbekli Tepe backfill, the Eridu Genesis tablet's handwriting-only date, the Hu bottleneck's model dependence — and none of them is of the size or direction required, because the mechanisms that move dates are specific to artifact types and do not transfer to monolithic excavations, portable grave goods, or eyewitness testimony. The correct conclusion is not that conventional dates are unreliable, and not that they are beyond question; it is that they are unreliable in named, bounded, publishable ways which have been named here and which leave the pre-Flood horizon exactly where it was. The softest spot of all — the radiocarbon plateau lying right at the Septuagint Flood horizon — is the exception that proves the rule: real imprecision, but of the wrong size and the wrong direction to move that horizon anywhere, so it supports the long chronology no more than the short, and the dating record supplies the biblical reckoning no help it could not earn on its own terms.

Sources