Creationist criticism of radiometric dating usually begins by making the science small enough to defeat.
Scientists, we are told, take a rock, measure a daughter product and calculate an age only by assuming three things: the initial amount of daughter material, a perfectly closed system and an unchanging decay rate. Since nobody was present when the rock formed, none of these assumptions can be known. The date is therefore circular, the deep past speculative and Genesis restored to office after a brief administrative misunderstanding.
The description has the advantage of being simple. Its remaining advantage is that almost none of it describes modern geochronology.
Radiometric dating is not one method. It is a family of methods applied to different isotopes, minerals, materials and timescales. Scientists select a system appropriate to the event being dated, measure isotope ratios with uncertainty, test whether the mineral behaved as required, compare minerals within a sample and compare the result with other isotopic and non-isotopic evidence.
Bad samples exist. Disturbed systems exist. Contamination exists. Scientists know this because the methods expose them. A clock can be misread. It does not follow that clocks are imaginary.
What the clock actually measures
Some atomic nuclei are unstable. They transform into other nuclei through radioactive decay. A parent isotope decreases while a daughter product accumulates. For a large population of atoms, decay follows a statistically regular law characterised by a decay constant or, more familiarly, a half-life - the time in which half the parent nuclei are expected to decay.
If the initial conditions and later history of a mineral system can be established or constrained, measured parent and daughter isotopes provide the time elapsed since a relevant event. That event is usually not "when the rock began to exist" in some vague metaphysical sense. It may be crystallisation, cooling below a closure temperature, metamorphism, mineral growth or the death of an organism.
This precision matters. A zircon crystal may preserve the time it crystallised from magma. Argon in a volcanic mineral may record when cooling allowed radiogenic argon to remain trapped. Carbon-14 dates the end of carbon exchange in once-living material. Different clocks start under different conditions. Asking why a carbon date cannot give the age of a dinosaur is rather like complaining that a kitchen timer has failed to record the Norman Conquest.
The mathematics is not mystical. In a simple parent-daughter system, the remaining parent population declines exponentially. An age equation relates the measured parent, the accumulated radiogenic daughter and the decay constant. The difficult work lies in mineral behaviour, sample preparation, calibration, analytical precision and geological interpretation. Creationist presentations often display the equation and quietly remove the laboratory.
There is no single radiometric clock
Different isotope systems cover different materials and ranges.
Radiocarbon
Cosmic-ray interactions in the atmosphere produce carbon-14, which enters carbon dioxide and the biological carbon cycle. While an organism lives, it exchanges carbon with its environment. After death, exchange largely stops and carbon-14 declines with a half-life of about 5,730 years.
Radiocarbon is useful for once-living material over tens of thousands of years, not for dating the Earth. Measurements must be calibrated because atmospheric carbon-14 has varied. Modern calibration curves are built from independently dated tree rings, corals, speleothems, marine records and other archives. IntCal20 extends Northern Hemisphere calibration to about 55,000 calibrated years before present.
This is not an embarrassing repair to a failed method. Calibration is the method becoming more exact by comparison with independent evidence. The raw radiocarbon age is translated into a calendar range with explicit uncertainty. Reservoir effects, contamination and isotopic fractionation are investigated rather than prayed away.
Potassium-argon and argon-argon
Potassium-40 decays partly to argon-40, with a half-life of about 1.25 billion years. Argon is a gas and ordinarily escapes from molten material. As a mineral cools, it begins to retain argon. The clock therefore relates to cooling and closure.
The older potassium-argon method measures potassium and argon separately. The argon-argon method irradiates a sample to convert a stable potassium isotope into argon-39, allowing isotope ratios to be measured in the same analytical system. Step-heating can reveal whether argon components were lost, inherited or trapped. A disturbed age spectrum looks disturbed. The machine does not print "six thousand years" out of politeness.
Uranium-lead
Uranium-lead dating is especially powerful in zircon. Zircon incorporates uranium into its crystal lattice but strongly excludes ordinary lead when it forms. Two independent decay chains operate: uranium-238 to lead-206, with a half-life of about 4.47 billion years, and uranium-235 to lead-207, with a half-life of about 704 million years.
Agreement between the two clocks produces a concordant age. Lead loss or later disturbance moves measurements away from concordance in structured ways that can reveal both the original crystallisation and later geological events. The system contains an internal audit. Young-Earth criticism often speaks as though scientists had forgotten to ask where the lead came from while using a mineral selected partly because its chemistry answers that question.
Rubidium-strontium and isochrons
Rubidium-87 decays to strontium-87 over a very long timescale. Minerals from the same rock can begin with different rubidium-to-strontium ratios while sharing a common initial strontium isotope composition. Plotting appropriate isotope ratios for several minerals produces an isochron. Its slope yields the age and its intercept estimates the initial daughter ratio.
This directly addresses the standard claim that initial daughter material must simply be assumed. The initial ratio is one of the quantities inferred from the data. Samples that do not share the required history may fail to form a meaningful line or reveal mixing. An isochron is not immune to every geological complication, but neither is it the naive one-sample calculation creationist diagrams continue to prosecute.
Fission tracks and other thermochronometers
Spontaneous fission of uranium-238 leaves damage tracks in minerals and natural glass. Track density, uranium content and the thermal history of the material can provide an age or cooling history. Heating can anneal tracks, which makes the method sensitive to temperature as well as elapsed time.
Other systems include samarium-neodymium, lutetium-hafnium, uranium-thorium and cosmogenic nuclides. Modern geochronology is not a single vulnerable pillar. It is a building full of clocks with different mechanisms, start conditions and susceptibilities. The important fact is not that every clock always agrees. It is that scientists can often explain why a particular clock recorded a different geological event.
The assumptions are questions
Every measurement depends upon assumptions. So does reading a thermometer, identifying a supernova or accepting that the author of a social-media post is not three raccoons operating a subscription service. The relevant question is whether an assumption is unsupported, testable, independently constrained or contradicted by the data.
Initial daughter material
Some systems minimise initial daughter incorporation through mineral chemistry, as zircon does with lead. Some use non-radiogenic isotopes to distinguish initial from radiogenic components. Isochron methods solve for an initial ratio. In other cases, geological context, mineral inclusions or known reservoir compositions supply constraints.
The initial state is therefore not handled by writing "zero" in a box and hoping God was not watching. Different methods contain different controls.
Closed systems
No geochronologist imagines that every rock has remained chemically sealed since formation. Weathering, metamorphism, fluids, diffusion and recrystallisation move elements. The task is to identify minerals and domains that preserve the event of interest, and to detect or model later disturbance.
Microscopic imaging can reveal growth zones and altered regions. Replicate analyses can expose outliers. Minerals with different closure temperatures can reconstruct cooling histories. Discordant uranium-lead data can diagnose lead loss. Argon step-heating can separate components. A sample that has not behaved as a closed system is evidence about its history, not evidence that isotope physics has resigned.
Constant decay rates
Decay constants are measured experimentally. Nuclear physics explains why ordinary changes in temperature, pressure and chemical state do not produce the enormous variations required by young-Earth chronology. Small environmental effects can occur in particular electron-capture decays, but these are measured, isotope-specific and nowhere near the factors of hundreds of thousands or millions needed to compress geological history into a few millennia.
Decay rates also leave consequences outside the dating calculation. Radioactive decay produces heat and daughter isotopes. Dramatically accelerating billions of years of decay into a creationist timescale would release catastrophic energy. It would not simply make the clocks run faster. It would melt or vaporise large portions of the crust and then require another miracle to remove the heat while preserving the evidence that the heat was never there.
At that point the proposal has ceased to be an alternative scientific chronology. It has become continuous supernatural intervention designed to manufacture the appearance of a history that did not occur. A deity may be imagined capable of this. The rocks are not evidence for it.
Calibration and convergence
The most powerful answer to hidden assumptions is independent agreement.
Radiocarbon dates overlap with annually counted tree rings. Tree-ring sequences extend beyond the life of any single tree by matching patterns in living and dead wood. Lake varves, ice layers, corals and speleothems supply additional comparisons. The methods are not identical and their uncertainties differ. Their broad agreement cannot be produced by assuming the same age in the same equation because they do not use the same equation.
Volcanic ash layers can be dated radiometrically and traced across regions. They appear in stratigraphic order among sediments and fossils. Reversals of Earth's magnetic field recorded in rock sequences correlate across oceans and continents and can be tied to radiometric dates. Astronomically driven climate cycles leave patterns in sediments that can be counted and compared with orbital calculations. Plate motions inferred from magnetic stripes and measured directly by geodesy belong to the same physical history.
Meteorites dated by several isotope systems cluster around 4.56 billion years. The oldest terrestrial minerals are younger because Earth has been geologically active, recycling much of its earliest crust. Lunar samples, meteorites and terrestrial lead-isotope relationships converge upon an Earth and Solar System approximately 4.54 billion years old.
This age was not selected because it offended Genesis most efficiently. It emerged from measurements made with different materials and physical relationships. A conspiracy would require not only geologists but nuclear physicists, chemists, astronomers, palaeontologists, oceanographers, instrument builders and the minerals themselves to maintain a remarkably disciplined correspondence.
When dates disagree
Creationist literature delights in anomalous dates. A recent lava flow yields an age of millions of years. Living shellfish appear old. Diamonds contain traces of carbon-14. Different methods disagree. Therefore, all dating is unreliable.
An anomaly is useful only after one asks what was dated, whether the method was appropriate, how the sample was prepared, what correction was required and whether the laboratory itself warned against the interpretation.
Marine organisms can appear older than contemporaneous terrestrial organisms because dissolved ocean carbon includes carbon isolated from the atmosphere for long periods. This reservoir effect is measured and corrected regionally. Freshwater systems can have their own reservoir problems. Scientists did not discover these effects from a tract. They discovered them by comparing dates and investigating the discrepancy.
Very young volcanic rocks can contain inherited or excess argon and lie below the useful range of a conventional potassium-argon analysis. Submitting such material to an inappropriate method does not test whether the method can date young rocks. It tests whether the instrument is capable of detecting that argon exists. A bathroom scale that reads 0.1 kilograms while weighing a postage stamp has not disproved mass.
Trace carbon-14 measurements approach backgrounds where contamination, instrument blanks and in-situ production matter. Detecting a tiny signal is not identical to showing that all of it is surviving primordial carbon-14. Analytical laboratories devote unromantic effort to blanks because machines, unlike apologetic diagrams, have histories.
Some dates are genuinely wrong. Samples are altered. Context is misunderstood. Instruments drift. Results are retracted or revised. Science is not protected by inerrancy. It is protected, imperfectly, by exposure: methods, standards, uncertainties, replication and the possibility that another laboratory will find the embarrassing thing one preferred not to find.
The Mount St Helens theatre
A frequently repeated creationist claim concerns dacite from the 1986 lava dome at Mount St Helens, submitted for potassium-argon analysis and reported with component ages ranging into hundreds of thousands or millions of years. The argument is that a method producing old dates for a known-young rock cannot be trusted on unknown-old rocks.
The samples were far younger than the effective range of the conventional whole-rock method used and contained minerals with older histories as well as potential excess argon. The laboratory reportedly indicated limits on measuring such young material. A whole-rock sample is not necessarily a clock reset to zero at eruption, particularly when it includes inherited crystal components.
A scientifically useful test would use a method capable of resolving the expected age, characterise the mineral phases, assess inherited components and compare results with known eruptive history. Instead, an unsuitable measurement was obtained and toured through creationist literature as though geochronologists had never encountered xenocrysts.
The performance survives because it illustrates a desired moral: experts are arrogant, assumptions are hidden and a simple man with a Bible has exposed the laboratory. The geology is incidental. The parable was commissioned in advance.
The RATE escape and its heat
The Institute for Creation Research's RATE project acknowledged that conventional amounts of radioactive decay imply vast ages, then proposed that decay rates accelerated enormously during Creation or the Flood. This concedes more than its public rhetoric usually admits. The isotope evidence is real enough to require new nuclear history.
Accelerated decay creates several problems. The heat alone is devastating. Nuclear decay across the crust accelerated by the required factors would release energy capable of destroying the geological structures the model seeks to explain. Radiation damage, daughter-product distributions, helium behaviour and agreement among isotope systems also require coordinated intervention.
Creationist authors have acknowledged a heat problem and appealed to unknown cooling mechanisms. An unknown mechanism may be a research question when it is constrained by physics and evidence. Here it functions as a miracle-shaped drain into which an unwanted planet's worth of energy is poured.
Once miracles may alter decay, remove heat, control diffusion, reset minerals and preserve apparent concordance, no observation can threaten the model. Any measured history can be declared the product of unmeasured intervention. The account becomes compatible with every possible rock and explanatory of none.
What a scientific objection would require
A serious alternative to radiometric chronology would need more than a list of possible errors.
It would need to identify a systematic physical mechanism producing false ages across several isotope systems. It would need to predict which samples should be affected, in what direction and by how much. It would need to explain concordant uranium-lead ages, isochrons, calibrated radiocarbon, meteorite results, stratigraphic order, magnetic reversals and non-radiometric chronologies. It would need to survive laboratory testing and produce better quantitative predictions than existing models.
Most creationist objections do none of this. They point out that contamination is possible, although laboratories measure it. They point out that open systems exist, although geochronologists diagnose them. They point out that initial daughter material matters, although several methods solve for it. They point out that decay constancy is an assumption, then propose variation so enormous that the crust requires miraculous refrigeration.
This is not a rival theory. It is an objection generator protected from the obligation to become one.
The antiquity that does not care
Radiometric dating does not prove atheism. An old Earth is compatible with many forms of Christianity, Judaism and other religion. Most Christians worldwide do not require the planet to be younger than agriculture. Geochronology measures physical history; it does not issue metaphysical citizenship papers.
The conflict arises only where a particular interpretation of Scripture is granted authority over every independent record. Then the age of a zircon becomes a moral threat. The laboratory is accused of naturalism, the result of circularity and the scientist of rebellion. A mineral has been dragged into a doctrinal dispute it spent four billion years avoiding.
The cosmicist finds no consolation in antiquity. Deep time does not make humanity important. It reduces our species to the final film on a long geological surface. Civilisations occupy a fraction of the last fraction. Every scripture was written yesterday by planetary standards.
That does not make the evidence hostile. It makes it indifferent.
The atoms decay at rates no creed selected. Zircons retain events no priest witnessed. Meteorites carry a common history from before Earth had oceans, life or anyone available to misunderstand Genesis. The clocks are not trying to humiliate us. Humiliation enters when a species expects nature to keep its chronology in agreement with a local text.
The complaint has been heard.
The rocks have supplied their documentation.
The appeal is denied.
References
U.S. Geological Survey, "Geologic Age: Using Radioactive Decay to Determine Geologic Age".
U.S. Geological Survey, "A Beginner's Guide to Dating Rocks", 2024.
U.S. Geological Survey, "Going, Going, Argon!", 2024.
P. J. Reimer et al., "The IntCal20 Northern Hemisphere Radiocarbon Age Calibration Curve (0-55 cal kBP)", Radiocarbon 62, 2020.
G. Brent Dalrymple, The Age of the Earth, Stanford University Press, 1991.
Alan P. Dickin, Radiogenic Isotope Geology, third edition, Cambridge University Press, 2018.
Ian McDougall and T. Mark Harrison, Geochronology and Thermochronology by the 40Ar/39Ar Method, second edition, Oxford University Press, 1999.
F. Oberli et al., "A Systematic Re-evaluation of the Uranium-Lead Decay Constants", Geochimica et Cosmochimica Acta 68, 2004.
Clair C. Patterson, "Age of Meteorites and the Earth", Geochimica et Cosmochimica Acta 10, 1956.
Steven A. Austin, editor, Radioisotopes and the Age of the Earth, volumes 1 and 2, Institute for Creation Research and Creation Research Society, 2000 and 2005. Cited as the principal young-Earth accelerated-decay proposal.
Loading this essay's discussion...