Cross-Calibration

Any single dating method can be questioned on its assumptions. The argument that carries the most weight is that methods resting on unrelated physics agree with one another.

An error large enough to compress the timescale to a few thousand years would have to affect counting, nuclear decay, electron trapping, light travel, and plate motion — by the same factor, in the same direction, while leaving them consistent with each other.

The methods and what they share#

The first four involve no radioactive decay at all. If every decay-based method were wrong, they would be unaffected.

Documented agreements#

Tree rings and radiocarbon. Carbon-14 measured in individually dated rings tracks the calibration curve smoothly across more than 12,000 years. A counted record and a decay record, agreeing year by year.

Ice cores and tree rings. The beryllium-10 spike of 774–775 CE, produced by a cosmic ray event, appears in Japanese and European tree rings and in Greenland and Antarctic ice at the same counted year. Two archives, two continents, two materials, one year.

Volcanic layers across archives. Historically dated eruptions — Tambora 1815, Laki 1783, Vesuvius 79 CE — appear as sulfate spikes at the counted depth in ice, as ash in varved lake sediment, and as frost-damage rings in trees.

Varves and radiocarbon. Lake Suigetsu's counted layers contain terrestrial plant fragments whose radiocarbon ages match the counted position to about 52,800 years.

Radiometric methods against each other. Uranium-lead, potassium-argon, rubidium-strontium, and samarium-neodymium involve different elements, different chemistry, and half-lives spanning two orders of magnitude. On well-behaved samples they agree within one or two percent.

Seafloor spreading and magnetic reversals. Stripe widths on the seafloor are proportional to the durations of magnetic reversals dated independently from lava flows on land. Distance divided by measured GPS spreading rate reproduces the radiometric ages.

Hotspot chains and plate motion. Age progression along the Hawaiian-Emperor chain gives a Pacific plate velocity matching the GPS measurement.

Astronomy and nuclear physics. The light curve of supernova SN 1987A, 168,000 light years away, is powered by cobalt-56 decay at the laboratory half-life. The light left before any proposed acceleration and records the same decay constant.

Tidal rhythmites and celestial mechanics. Ancient tidal deposits record the number of days per lunar month, and the values decrease going back in time at the rate predicted by tidal friction slowing Earth's rotation — calculated from orbital dynamics.

Why the agreement is hard to explain away#

Each pair above involves methods with no shared assumption. A systematic error in one has no reason to produce a matching error in the other.

The comparison is also directional. Counted records — rings, varves, layers, bands — are the ones with the fewest assumptions, and they are used to calibrate the decay-based ones rather than the reverse. If the decay methods were systematically wrong, the counted records would expose it, because they do not depend on decay.

The counted records alone already exceed a 6,000-year chronology by an order of magnitude, before any decay method is involved.

Common objections#

"Methods that disagree are discarded, so agreement is manufactured"

Samples are rejected, and this is a real practice. Discordant results are excluded when there is evidence of contamination, weathering, or an open system.

The rejection criteria are stated in advance and are independent of the answer. A concordia plot identifies lead loss from the geometry of the data; a failed isochron shows scatter rather than a line; a luminescence sample shows a partially bleached dose distribution.

These diagnostics identify which samples are disturbed without reference to the expected age, and the raw data is published.

The agreement across independent methods is also not something sample selection could produce. Rejecting bad zircons does not cause tree rings in Germany to match ice layers in Greenland.

"They all assume the present is the key to the past"

Correct, and this is the uniformity of process assumption discussed in Catastrophe and Gradualism.

It is tested rather than merely assumed. Decay rates have been measured under varied temperature, pressure, and chemical conditions; SN 1987A confirms the cobalt-56 rate at 168,000 years ago; and counted records return correct ages for historically documented events.

Rates that have varied are measured rather than assumed constant. Atmospheric carbon-14 varied, and the calibration curve records the variation directly. Seafloor spreading rates varied, and stripe widths record it.

"Accelerated decay would affect all radiometric methods equally, preserving agreement"

The most technically serious response, from the RATE project. If all decay constants were multiplied by the same factor, the methods would remain concordant with each other while giving inflated ages.

The proposal is internally coherent for the decay-based methods, and this is worth acknowledging.

It fails for the methods that do not use decay. Counting tree rings, varves, and ice layers is unaffected by decay rates, and those records reach beyond 50,000 years independently.

It also makes the wrong prediction for luminescence: a vastly higher dose rate would saturate every sample, and samples are routinely unsaturated.

Different half-lives would also need different acceleration factors to remain concordant, since the systems span two orders of magnitude in half-life — and the RATE team's own work found the required heat release unresolvable.

"Circular reasoning — each method is calibrated against the others"

Calibration does occur, and radiocarbon is explicitly calibrated against tree rings and varves.

The direction is the point. Counted records are primary: a tree ring is counted, not calculated, and requires no input from any other method. Radiocarbon is adjusted to match them.

Radiometric dating is not calibrated against counted records at all. Its decay constants are measured in laboratories by direct counting of decay events, and the agreement with counted archives where they overlap is a test rather than a calibration.

What the evidence shows#

Methods resting on counting, nuclear decay, electron trapping, magnetic recording, plate motion, and light travel time give consistent ages where their ranges overlap.

Several of these share no physics. Tree rings and ice layers are counted; luminescence measures accumulated dose; astronomical distances measure light travel.

Single-year events — the 774–775 CE radionuclide spike, historically dated eruptions — appear at the same year in archives on different continents in different materials.

Compressing the timescale requires every one of these to be wrong by the same factor while remaining consistent with each other, and requires the counted records, which involve no decay, to be wrong for reasons unrelated to the decay-based ones.