The first six parts of this series examined how a philosophical commitment — materialism — came to wear the uniform of science in the study of origins. This part and the next turn to where most people feel that tension most directly: the numbers. The ages. The deep-time figures printed on museum placards and in textbooks as if someone had measured them with a tape. Where do those numbers actually come from? And how much of the confidence behind them is earned by evidence — rather than supplied by the framework before any evidence is in?
These are fair questions, and they deserve a fair answer — which means an honest one in both directions. Some of what skeptics say about dating is stronger than mainstream scientists admit. Some of it is weaker than the skeptics themselves wish. We are going to sort the strong from the weak — and use the same scalpel on the claims we like and the claims we don’t.
One reminder of our declared lens: I am a Christian pastor who believes the Bible is the authoritative Word of God and that the evidence for design in creation is compelling. I also hold that the age of the earth is a question on which sincere, orthodox Christians have long disagreed — and this article does not settle it. What it examines is narrower and, I think, more important: the gap between what the dating evidence actually supports and the certainty with which the conclusions are announced.
The Category Error Almost Everyone Makes
Start with the single most common confusion in this entire debate — one made by skeptics and defenders alike. People say “carbon dating” as though it were the method that tells us dinosaurs are 68 million years old or that the earth is 4.5 billion. It is not. It cannot be. And clearing this up actually sharpens the real question rather than dissolving it.
Radiocarbon (carbon-14) has a half-life of about 5,730 years. After roughly ten half-lives — about 50,000 to 60,000 years — there is essentially no measurable carbon-14 left in a sample, and the measurement hits the background floor of the instrument. Carbon dating therefore has a hard ceiling around 50,000 years. It is physically incapable of producing an age of millions of years.
So dinosaurs and ancient rocks are not dated by carbon at all. They are dated by entirely different clocks — potassium-argon, argon-argon, uranium-lead — that rely on slow-decaying elements with half-lives in the hundreds of millions or billions of years. The sentence “carbon dating proves the earth is billions of years old” is a category error. Carbon cannot even reach those ages.
Here is why that matters. The serious version of the skeptical argument is not “carbon dating gives the wrong answer.” It is narrower and sharper: we keep finding measurable carbon-14 in materials that are supposed to be millions of years old — coal, diamonds, even dinosaur bone — where there should be exactly none. That is a real observation, and it has been published. The entire weight of the argument then rests on one question, and only one: is that trace carbon-14 genuinely original to the sample, or is it contamination plus the machine’s own background signal? We will weigh that specific question carefully in Part 8. For now, notice what has happened: within its proper range — the last fifty thousand years — carbon dating is not the weak link in the chain. It is one of the most rigorously cross-checked tools science has.
Where the Clocks Genuinely Agree
Intellectual honesty requires leading with the part of this that is strong — strong enough that we should never pretend otherwise. Within their ranges, the independent dating clocks corroborate each other to a degree that is genuinely impressive, and a truth-seeker has to say so.
Carbon-14 is calibrated against records you can literally count. Continuous tree-ring chronologies extend back nearly 14,000 years, one countable ring per year. Annual lake sediment layers (“varves”), coral growth bands, and cave formations provide further independent year-by-year records — and when the carbon-14 clock is checked against them, they line up well. This is not one assumption resting on another. It is one clock checked against a different clock that ticks visibly.
Ice cores reinforce the picture. In Greenland’s GISP2 core, researchers have counted more than 110,000 annual layers using several independent markers — visual summer/winter texture, dust, and chemistry. Layers produced by one-off events such as storms or melt are physically distinguishable from seasonal layers. This is a countable clock that reaches far beyond the carbon range, and it is hard to wave away.
The same is true of the slow radiometric clocks at their best. A single zircon crystal can be dated by more than one independent uranium-and-thorium decay chain; when the sample is undisturbed, those independent chains converge on the same age. That convergence is called concordance, and it is not the kind of thing that happens by accident if the underlying method were arbitrary.
So let us put a stake in the ground that we will not pull up later: the claim “all dating methods are broken and contradict each other” does not survive contact with the evidence. If we built our case there, a competent skeptic would dismantle it in an afternoon — and would be right to. We have spent six parts of this series insisting that the other side discard its weak arguments. We discard this one. The honest case lies elsewhere, and it is sharper for being honest.
How the Old Clocks Actually Work — and What They Assume
A radiometric date is not a direct reading. It is a calculation. You measure how much of a radioactive “parent” element remains and how much “daughter” element has accumulated, and you compute backward to a starting time. To turn that into an age, you have to assume three things: that you know how much was there to begin with, that nothing has been added or lost since (what scientists call a “closed system”), and that the decay rate has never changed.
Two of those assumptions are far better supported than skeptics often grant. Decay-rate constancy, in particular, has been tested against the Oklo natural nuclear reactor (which ran roughly 1.7 billion years ago by the standard reckoning) and against light from supernovae billions of light-years away, and it has held up. Methods like the isochron technique are specifically designed to detect when the closed-system assumption has failed, by using multiple samples that should fall on a straight line if the system stayed closed.
But notice what is left over. Every one of these methods still reasons backward — from measurements we can take today to a past no one was there to watch — on the assumption that the physics held steady the whole way. That assumption has passed every test we have devised — which is a real and strong thing to be able to say. It is also, at bottom, an inference about a past no one watched, not a direct observation of it. The right posture is neither “it’s all guesswork” nor “it’s simply measured fact,” but something more careful: well-tested inference within a framework that itself rests on a philosophical commitment to uniform physics. That is a defensible thing to say out loud. The textbooks rarely do.
The Real Problem: Which Date Gets Kept?
Here is where the skeptical instinct lands a genuine blow — not on the clocks themselves, but on how their outputs are selected. When a single rock yields several different radiometric ages, which one ends up in the literature? And on what basis is the rest set aside?
The most famous case is the KBS Tuff, a volcanic ash layer in Kenya that sat above a celebrated set of hominid and mammal fossils. Over the years, attempts to date it produced results ranging across an astonishing span — from about half a million years to well over two hundred million years.
An early measurement of roughly 212–230 million years was rejected — explicitly because the mammal fossils beneath the tuff did not fit an age that old. That is the heart of the concern: a date was set aside not first because of an independent physical test, but because it conflicted with the expected age implied by the fossils. More broadly, the conventional geologic time scale was built from a large set of radioisotope ages — and those ages were chosen for their agreement with the fossil and rock sequence already assumed to be correct. Ages that disagreed were set aside as contamination, excess argon, or leaching. By several accounts, some were never published at all.
State it plainly, because it is fair: if you keep the dates that match your expectation and explain away the ones that don’t, then the much-advertised “agreement” between radiometric dating and the fossil record can become, in part, a circle. The expectation helps select the data, and the selected data then confirms the expectation. The public is told the result is an airtight, assumption-free measurement. The actual practice involves considerably more interpretive judgment than that confidence admits. That gap — between the messy reality of date selection and the seamless certainty announced to the public — is exactly the disease this whole series has been diagnosing.
And now the honesty that has to come in the same breath, or we are doing the very thing we condemn.
The KBS Tuff controversy was eventually resolved — and not by fiat. Independent trace-element “fingerprinting” tied the tuff to a separately dated formation, and an independent specialist, Ian McDougall, re-dated it using two different methods that agreed closely with each other: about 1.88 and 1.89 million years. The wild early spread narrowed to a tight, cross-validated number once the specific contamination problems (such as older crystals carried up in the magma) were understood. “Excess argon” is not an excuse invented on the spot — it is a real, measurable, well-understood phenomenon.
So the fair verdict is not “the dates are fabricated.” It is more precise and, frankly, more damaging because it is harder to dismiss: the dating enterprise is more interpretive, more judgment-laden, and more vulnerable to confirmation bias than its confident public posture admits — but it is not the free-for-all the strongest skeptical rhetoric implies. Oversold and under-disclosed. Not invented.
A Test Case We Should Handle Carefully
One more, because it is widely cited and we should model how to use it responsibly rather than how to swing it as a club. In 1986 the new lava dome at Mount St. Helens produced fresh rock of known age. When a sample was dated by potassium-argon, the result came back not as “essentially zero” but as ages up to roughly 2.8 million years for some mineral fractions.
The measurement is real and the point is genuine: this is a vivid demonstration that potassium-argon dating can return ages of millions of years on rock that is decades old, because of “excess argon” trapped in older crystals carried up from the magma. But it does not generalize to “therefore all radiometric dating is worthless.” Geologists already know that potassium-argon is unreliable on very young volcanic rock (under roughly two million years) and do not use it there; and the excess argon that wrecks a date on decades-old rock is negligible when dating genuinely ancient rock. The honest reading: this exposes a known failure mode that is too often glossed over in public — not a collapse of the whole method.
Why insist on that restraint? Because Part 1 of this series praised the discipline of discarding our own bad evidence — the way young-earth advocates eventually withdrew the Paluxy River “man tracks.” If we now wield an argument past what it can bear, we forfeit exactly the credibility that makes the rest of our case land. The strong points are strong enough. They do not need inflation.
What This Part Establishes
Set the ledger down honestly. Within its range, carbon dating is accurate and cross-checked against clocks you can count. The slow radiometric clocks, at their best, converge across independent methods, and decay constancy has survived real tests. The chronology is not a house of cards.
And yet the confidence sold to the public outruns that evidence at a specific, identifiable seam: the selection of which dates to keep. Genuine interpretive judgment, real confirmation-bias pressure, and an unprovable foundational assumption about uniform physics are all packaged and presented as seamless, assumption-free certainty. That packaging — not the data — is where science stops behaving like science.
In Part 8 we test the sharpest case of all — the soft, still-elastic tissue recovered from dinosaur bone that no one’s preservation model predicted — and we grade, with the same honesty, which anomalies are strong, which are weak, and what the whole picture finally tells us about whether modern origins science is still doing science.
“Unapologetically Faithful. Searching with Evidence.”
Derech Truth Labs | derechtech.com | © 2026 Derech Technologies LLC