Radiocarbon and Other Dating Methods

All living organisms absorb carbon-14 from the atmosphere at a ratio set by the contemporary atmospheric concentration. When the organism dies, uptake stops and 14C begins to decay at a known rate, with a half-life of 5,730 years. Measuring the residual 14C/12C ratio in a sample therefore gives a raw radiocarbon age. The problem is that atmospheric 14C concentration has not been constant through time, so a raw age in radiocarbon years does not map neatly onto calendar years. Calibration with the IntCal20 dataset, published in 2020 and built from over 14,000 tree-ring, coral, and cave-deposit measurements, converts the raw age into a probability distribution over calendar years.

The calibrated result is expressed as a calendar-year range at a stated confidence level. A 95% confidence interval means that, given the measurement and its uncertainty, there is a 95% probability the true age falls in that range. The width of the range depends on two things: the counting statistics of the AMS measurement (more material, smaller uncertainty) and the local shape of the calibration curve. Where the curve is steep, a small measurement uncertainty translates into a narrow calendar range. Where it is flat or reverses direction (a plateau or a wiggle), even a precise AMS measurement becomes a wide calendar range.

For forensic cases from the last few centuries, standard radiocarbon dating is often frustratingly imprecise. The calibration curve from roughly 1650 to 1950 CE contains a long plateau known as the 'post-medieval wiggly bit,' where dates can only be pinned to a century or more. This is where bomb-pulse dating becomes the preferred tool.

Atmospheric nuclear weapons tests, beginning with the Trinity test in 1945 and escalating rapidly through the 1950s, injected large amounts of carbon-14 into the upper atmosphere. By 1963, when the Partial Test Ban Treaty drove testing underground, the atmospheric 14C concentration had roughly doubled compared to the pre-industrial baseline. Since then it has declined steadily as the bomb carbon dispersed into the oceans and the terrestrial biosphere. This rise and fall has been measured continuously at monitoring stations including Vermunt in Austria and Schauinsland in Germany, producing a dated reference curve that is distinct for almost every year between 1955 and the present.

The method's power was demonstrated most clearly in work by Kirsty Spalding and colleagues at the Karolinska Institute, who used bomb-pulse 14C in DNA from human neurons to show that most neurons in the cerebral cortex are as old as the individual. The forensic application, refined by teams in Sweden and the UK, uses dental enamel from specific tooth types to assign a birth year, and compares that to the measured 14C in cortical bone to estimate age at death. Combined, these measurements can often establish whether a person was born before or after 1950, and for those born in the 1955-1995 window, can narrow the birth year to within two to four years.

Every tree adds one ring per growing season, and the width of each ring records the conditions of that year. Cold, dry, or stressful years produce narrow rings; warm, wet years produce wide ones. Because trees across a region experience the same climate signals, their ring sequences are correlated, and overlapping sequences from living trees, historic timbers, and archaeological wood can be chained back thousands of years. For oak in western Europe, continuous master chronologies now extend beyond 10,000 years before present.

Matching the ring pattern of an unknown timber to the master chronology assigns a felling date to the year, and in the case of heartwood the outermost rings even allow determination of the season of felling. In burial contexts, coffin planks, structural timbers lining a vault, or wooden grave markers are the primary targets. The dendrochronological date is a terminus post quem: the burial cannot have occurred before the timber was felled. If the wood shows no signs of long secondary use, the burial date is likely very close to the felling date.

The limitation is sample quality. A timber needs at least 40-50 rings for reliable matching, and the outermost sapwood rings are crucial because they record the period immediately before felling. Waterlogged or charred timbers often preserve rings well; dry coffin planks from sealed vaults can be equally cooperative. Timbers that were secondarily reused, which was common in historic carpentry, can give misleadingly old dates, so the analyst must assess whether the wood shows signs of earlier use before drawing burial-date inferences.

Quartz and feldspar grains in soil accumulate trapped electrons as they absorb radiation from naturally occurring uranium, thorium, and potassium in the surrounding sediment. When those grains are exposed to sunlight or heated, the trapped electrons are released as light, resetting the clock to zero. Measuring the accumulated electron dose in grains recovered from a burial fill, combined with a measurement of the annual radiation dose rate, gives the time elapsed since last light exposure. For an undisturbed burial fill, that event is when the fill was shovelled into the open grave.

• OSL: stimulates electrons from grains using visible or infrared light, typically measuring single aliquots or individual grains. Single-grain OSL is now routinely used in forensic contexts to detect whether individual grains were incompletely bleached before burial, which produces overdispersion in the age data.
• TL (thermoluminescence): releases trapped electrons by heating. Still used for dating ceramics and burnt flints found in burial contexts, because both materials are reset during firing rather than by light.
• Application window: both methods cover approximately 100 to 200,000 years, with typical precision of 5-10% of the age. For a 500-year-old burial the 1-sigma uncertainty might be 25-50 years; useful for historical forensic work but coarser than dendrochronology or bomb-pulse.
• Critical limitation: OSL dates the last light exposure of the sediment grains, not the body. If previously bleached grains from a disturbed older deposit were incorporated into the fill, or if the fill was never adequately bleached before burial (for example, from a deep-shade excavation at night), the date will be misleading.

Before reaching for a laboratory, every forensic archaeologist reads the stratigraphic sequence. Superposition says that in undisturbed contexts, lower deposits predate upper ones. A grave cut intrudes through earlier layers, which gives a terminus post quem from any datable material in the cut-through deposits. The fill itself is sealed by the layer that accumulated on top after the grave was closed, giving a terminus ante quem from the sealing layer. The burial event therefore falls within a stratigraphically constrained date range, often before any laboratory dating is attempted.

Artefacts add precision. Coins, ceramics, glass, and manufactured goods can often be dated by their production period. A coin has a mint date that is a hard TPQ; its manufacturing style places it within a narrow production window. Military uniform buttons, shoe soles, garment fasteners, and container glass all carry date-diagnostic characteristics documented in specialist reference works. The analyst must consider not only when an artefact was made but whether it was already old when deposited, which is why contextual analysis matters: a well-worn coin found in a primary deposit is less reliable as a TPQ than a sharp, unworn one.

Dating evidence enters a legal context that is poorly equipped to handle probabilistic ranges. Judges and juries want to know when. The analyst's job is to satisfy that need honestly without collapsing uncertainty into false precision or hiding behind impenetrable technical language.

• State the method and its assumptions: every dating result depends on assumptions (constant decay rate, complete bleaching of OSL grains, no old-wood effect). These should be stated explicitly and briefly so that opposing experts can evaluate them.
• Give the confidence interval explicitly: '95% probability that death occurred between 1705 and 1780 CE' is precise and honest. 'Death occurred in the 18th century' is impressionistic and harder to challenge.
• Explain what constrains the range: if the range could be narrowed with additional sampling or a complementary method, say so. Courts can request further analysis if the analyst flags what it would add.
• Address the forensic versus archaeological threshold: in most jurisdictions, remains older than roughly 70-100 years fall outside the scope of a criminal investigation. The analyst should state clearly whether the dating result supports or excludes the forensic threshold, and with what confidence.