How Carbon Dating Works, and Where Its Limits Are
Radiocarbon dating lets scientists put an age on ancient bones, wood and charcoal, but the method rests on assumptions that only hold within certain bounds.
The basic idea
All living things are built, in part, from carbon absorbed from the environment. Most of that carbon is the stable isotope carbon-12, but a tiny fraction is carbon-14, a radioactive isotope formed high in the atmosphere when cosmic rays strike nitrogen. Plants take up carbon-14 through photosynthesis, animals eat the plants, and so the isotope works its way into every living organism at roughly the same ratio to ordinary carbon that exists in the atmosphere at the time.
When an organism dies, it stops exchanging carbon with its surroundings. The carbon-14 already inside it keeps decaying, slowly, back into nitrogen, while the carbon-12 stays put. Because the decay happens at a fixed, well-understood rate, scientists can measure how much carbon-14 is left in a sample and work backwards to estimate how long ago the organism died. This decay rate is described by the isotope’s half-life, the time it takes for half of any given amount of carbon-14 to decay, which is a well-established physical constant of the isotope itself rather than something that changes over time or needs checking against a current source.
What actually gets measured
Modern labs rarely count individual radioactive decay events, because carbon-14 is so rare and decays so slowly that waiting for enough clicks on a detector would take too long for small samples. Instead, most dating today uses accelerator mass spectrometry, which counts the actual carbon-14 atoms in a sample directly, sorting them from the far more abundant carbon-12 and carbon-13 atoms by their mass. This lets researchers date very small samples, sometimes just a few milligrams of bone collagen, charcoal or plant fibre, which matters enormously in archaeology where destroying a large chunk of a rare object is not an option.
Why the raw number needs correcting
The date a lab first calculates assumes the amount of carbon-14 in the atmosphere has always been constant. It has not. Solar activity, changes in the Earth’s magnetic field, and the burning of fossil fuels and nuclear weapons testing in the twentieth century have all shifted atmospheric carbon-14 levels over time. To correct for this, scientists use a calibration curve built from material whose age is already known independently, such as tree rings that can be counted year by year, and layered sediments from lakes and marine cores. The internationally agreed calibration curve, maintained and periodically updated by a consortium of radiocarbon specialists, converts a raw radiocarbon measurement into a calendar-year age range. This is why radiocarbon results are usually reported as a range with a stated confidence level, not a single precise year.
Where the method runs into trouble
Carbon dating only works on material that once exchanged carbon with the atmosphere, which means organic remains: wood, charcoal, bone, shell, textile fibres, peat. It cannot date stone, metal or pottery directly, though it can sometimes date organic residue found alongside them, such as charred food remains stuck to a pot.
The method also has a practical time limit. After roughly ten half-lives, so little carbon-14 remains that it becomes very difficult to distinguish from background noise and contamination, which puts an effective ceiling on reliable dating at around fifty thousand years. Beyond that, other techniques such as potassium-argon or uranium-series dating take over.
Contamination is a persistent hazard. Even a small amount of younger carbon getting into an old sample, from soil, glue, handling, or conservation treatments, can skew a result towards a younger apparent age. Museums and excavation teams follow strict protocols for handling and storing samples destined for dating precisely because of this risk.
A further complication is the marine and freshwater reservoir effect. Organisms that get their carbon from the ocean or from certain lakes can show ages that appear older than they really are, because carbon dissolved in water bodies is not always in step with the atmosphere. Dating a coastal skeleton without accounting for a marine diet can produce a result that is off by centuries.
Reading a result sensibly
A radiocarbon date is best understood as a probability range grounded in a specific calibration curve, not a single fixed number plucked from a machine. Reputable dating labs report the isotope ratio, the statistical uncertainty, and the calibrated calendar range together, and archaeologists cross-check radiocarbon results against other evidence, such as stratigraphy, artefact style or historical records, wherever possible. The method is powerful precisely because it is well understood and its limits are well mapped, not because it produces a single unarguable figure.
Anyone wanting the technical detail behind a specific dating result, including which calibration curve version was used, should look at the documentation published by an accredited dating laboratory rather than relying on a rounded figure quoted secondhand.