Radiocarbon Dating measures radioactive carbon in once-living material, but a laboratory age is not automatically the date of a human event. To understand a hearth, burial, building or landscape, students need to ask **which material was dated, when its carbon stopped exchanging, how the result was calibrated and whether it belongs to the event being studied**.
Our original exercise compares a fictional carbonised cereal grain and oak charcoal from a hearth at Stone Orchard. Their example ages are 1,040 ± 30 BP and 1,350 ± 35 BP. These are made-up conventional radiocarbon ages, not calibrated dates, and neither can be converted into an exact AD year by subtracting from 1950. The real featured image, credited to the US National Park Service, shows ancient seeds at White Sands, an independent archaeological illustration.
The Historic England Radiocarbon Dating and Chronological Modelling guidelines provide an authoritative source for selection, laboratory reporting, calibration and Bayesian methods. The eduKate Gold Standard requires immediately useful explanations, original worked cases, transparent uncertainty and tests of independent understanding.
.jpg){kind=link}
Find the radiocarbon question
Follow the fictional grain and charcoal through the entire dating argument, or move to calibration, reservoir effects and independently checked chronology.
Chapter index
Carbon, specimens and archaeological context · Chapters 1–6
- Begin with the event the archaeologist wants to date
- Explain carbon-14 formation and biological uptake
- Teach radioactive half-life without treating it as a calendar shortcut
- Choose short-lived remains when the research question requires them
- Demonstrate the old-wood effect with tree-ring reasoning
- Separate primary burning, redeposition and later intrusion
Measurements, BP and calibration · Chapters 7–11
Multiple sources and chronology models · Chapters 12–15
Independent transfer · Chapters 16–16
1. Begin with the event the archaeologist wants to date
A fictional excavation at Stone Orchard uncovers a hearth filled with ash, old-looking charcoal and two carbonised cereal grains. Which material should be sampled to determine when people last used the hearth? A novice may pick the largest piece of charcoal because it offers the most material. Yet the oldest carbon in a long-lived tree can predate its final burning by centuries. Radiocarbon dating measures the carbon history of the sampled organism or material, and the archaeological interpretation must establish how that history relates to the event being investigated.
The teaching question should be written before selecting the specimen: ‘When did the hearth’s last well-documented burning episode occur?’ A short-lived cereal grain securely deposited in the activity may be more closely associated with that event than an inner-ring fragment from an old tree. The Historic England Radiocarbon Dating and Chronological Modelling guidelines put problem definition and sample selection at the centre of good practice. Stone Orchard, its grain and all laboratory codes below are invented; they are not real dates for any settlement.
Back to chapters · Next chapter
2. Explain carbon-14 formation and biological uptake
Radiocarbon, written carbon-14 or ¹⁴C, is a radioactive isotope of carbon. It is formed through processes involving cosmic rays and atmospheric nitrogen, enters the wider carbon cycle and is incorporated into living organisms. Once an organism stops exchanging carbon with the relevant reservoir, its ¹⁴C inventory changes through radioactive decay. Measurement can therefore inform the time since that biological carbon ceased exchange under appropriate assumptions. The method does not measure the literal age of clay, rock, metal or an archaeological layer by examining its colour.
Give students four imagined objects: a charred cereal seed, a stone axe, a wooden beam and an iron nail. Ask which contain potentially suitable datable carbon and what event might be recorded. The iron nail and stone axe do not directly provide ordinary carbon-14 ages; an associated organic sample may date a different event. A dated seed may document the age of its growth, not necessarily the age of nearby masonry. The first scientific skill is knowing what was measured, not merely knowing the name of the dating technique.
Back to chapters · Next chapter
3. Teach radioactive half-life without treating it as a calendar shortcut
Carbon-14 has a physical half-life of approximately 5,730 years. In a simplified model, after one half-life half the initial ¹⁴C atoms remain; after two, one quarter remain. The fraction at time t can be represented as 2 raised to the power −t/5730, when t is in years. This is a helpful physical model of decay, but raw radiocarbon ages also depend on measurement conventions, atmospheric history, isotopic fractionation and calibration.
Use a fictional material containing 100 arbitrary units of radiocarbon at the beginning of the teaching model. After one physical half-life, 50 remain; after two, 25 remain. Students should understand that an archaeological laboratory does not simply see an exact ‘25%’ and print a flawless calendar year. Calibration curves are essential because historic atmospheric ¹⁴C concentration was not constant. Explain the distinction between the physical half-life and the conventional radiocarbon-age calculation, which uses its own internationally established reporting convention.
Back to chapters · Next chapter
4. Choose short-lived remains when the research question requires them
Short-lived plant parts such as individual cereal grains, seeds or small twigs often reduce the risk that sampled carbon substantially predates the archaeological activity. This does not make them infallible. A grain can be residual from an earlier layer, intrusive from a later event, charred at a different time or misidentified. Material choice must therefore consider the object’s life history and its stratigraphic association, not only its biological lifespan.
Have learners compare a carbonised grain found within an intact firing deposit and a similar grain from a nearby disturbed ditch. The first is a stronger candidate for dating the hearth event if the context and identification are secure; the second might date a different event. Students should not automatically award the shortest-lived sample the highest confidence without checking context. Historic England’s sampling guidance explicitly couples the archaeological association with suitable material and carbon-reservoir behaviour.
Back to chapters · Next chapter
5. Demonstrate the old-wood effect with tree-ring reasoning
An oak may live for centuries. Carbon laid down in an inner ring stops being exchanged long before the outer growth rings form. If old inner heartwood is burned in a hearth, a radiocarbon measurement on that ring can be older than the burning event by a substantial and potentially uncertain interval. The method is not malfunctioning: it dates the carbon in the sampled ring, while the archaeologist wants a different event.
In the fictional Stone Orchard case, sample CH-02 is charcoal provisionally identified as old oak heartwood and reports a larger radiocarbon age than a cereal grain from the same context. Pupils should not average the two numbers to ‘fix’ the disagreement. They need to examine wood anatomy, carbon age and site formation. Dendrochronology can identify a particular growth-ring year in suitable wood and is important in building calibration references. The key transfer is that a scientifically sound determination can still be chronologically unsuitable for the archaeological question.
Back to chapters · Next chapter
6. Separate primary burning, redeposition and later intrusion
A hearth fill may have more than one formation event. Primary burning residues can be mixed with swept material, older rubbish, later soil or intrusive roots. An old-looking charcoal piece could have been deposited during the last fire or moved there later. The radiocarbon age of a sample will not automatically distinguish those possibilities. Sediment context, feature cuts, microstratigraphy, associated objects and documented collection methods help establish what the sample can date.
Give students three invented context cards: an intact burning surface, a rake-out deposit and a pit cut through the upper hearth fill. Ask which datable materials are associated with the targeted last fire and where uncertainty remains. A cereal grain recovered from the later pit could yield a plausible laboratory number and still be inappropriate for dating the original hearth. The exercise teaches event association rather than a rote preference for any one biological material.
Back to chapters · Next chapter
7. Document contamination and pretreatment as scientific necessities
Modern carbon contamination can bias an older sample toward an apparently more recent radiocarbon age. Ancient carbon contamination can sometimes produce an older apparent age. Conservation chemicals, soil carbon, handling, consolidants and poor source storage may complicate analysis. Laboratories use specialist pretreatment and quality control appropriate to the material, but the existence of a pretreatment protocol does not guarantee that every archaeological association is valid.
Students prepare a mock laboratory submission with context number, specimen type, taxonomic identification, storage history, visible consolidant and the question the age is meant to answer. A professionally treated specimen is not equivalent to a securely contextualised one; both conditions matter. Do not ask schoolchildren to perform radiocarbon chemistry or damage heritage material. They can learn from published laboratory reports and source forms, while specialist facilities handle the physical sample.
Back to chapters · Next chapter
8. Understand the measurement and laboratory identifier
Modern radiocarbon laboratories may use accelerator mass spectrometry to measure carbon isotope ratios, alongside preparation and corrections under their established protocols. A published conventional radiocarbon age has a laboratory code and an analytical uncertainty. The code enables readers to distinguish individual determinations and trace later recalibrations or modelling decisions. A bare number such as ‘a thousand years old’ lacks the information needed for archaeological interpretation.
Use the explicitly fictional code EDU-RC-01 for the cereal grain and EDU-RC-02 for the oak charcoal, and mark them as classroom labels, never actual accredited laboratory identifiers. Ask what should accompany each: precise specimen identification, context, pretreatment, radiocarbon age, one-sigma uncertainty, calibration method and relevant reservoir details. The Historic England dating section explains standard reporting. A result’s credibility comes from its documented source and methods, not an impressive number of significant figures.
Back to chapters · Next chapter
9. Teach BP as a reporting convention anchored to AD 1950
Uncalibrated conventional radiocarbon ages are conventionally expressed in radiocarbon years BP, meaning ‘before present’ with AD 1950 as the reference. The convention is not a moving reference to today’s date. A determination such as 1,040 ± 30 BP is not automatically a calendar date of AD 910 obtained by subtracting 1,040 from 1,950. Historical atmospheric ¹⁴C variations make that simple arithmetic misleading for archaeological dating.
Ask pupils to contrast three fields: conventional radiocarbon age BP, calibrated calendar-year interval and the archaeological event interpretation. These cannot be collapsed into one year column. A helpful class task is to reject a fabricated report reading ‘1040 BP = AD 910 exactly’ and rewrite it with a required calibration stage. The guide should remain optimistic: the radiocarbon determination provides useful quantitative evidence, but appropriate reporting explains why it needs an independently built conversion curve.
Back to chapters · Next chapter
10. Explain calibration with IntCal20 and other curves
Radiocarbon calibration converts a measurement’s radiocarbon age into calendar-date probabilities using independently dated material. Historic England identifies IntCal20 as the agreed relevant terrestrial northern-hemisphere curve for suitable pre-1950 material, with Marine20 and other choices for different reservoirs and periods. The choice must be matched to sample origin and the version reported. A curve itself has structure and uncertainty, so it can produce wider or separated calendar windows rather than a simple one-to-one conversion.
Give students a fictitious curve sketch with a plateau where several calendar years correspond to similar radiocarbon values. Ask why a very precise measured radiocarbon age might still calibrate to a broad range. If the lab report is recalibrated using an updated accepted curve, its calendar ranges can change while the measured conventional age remains the same. That is why publications preserve both the original lab code and BP determination. Students should not invent an IntCal20 result from a toy picture.
Back to chapters · Next chapter
11. Teach calibrated probability ranges without false precision
A calibrated radiocarbon result typically has a probability distribution over possible calendar dates rather than one certain calendar year. Historic England discusses common reporting at 95% probability and sometimes 68%, with the selected curve, method and relevant corrections stated. Separate calendar intervals can appear because the calibration curve is not monotonic at all times. Selecting the midpoint of the most visually prominent interval does not automatically produce the true event date.
Use two simulated reports: one announces ‘the hearth was used in AD 910’; another describes a hypothetical 95% calendar probability range and explicitly says that the numerical range is illustrative, not calculated. Pupils identify why the second is structurally more honest but still requires a real calibration. To teach accurate practice, show a published dated sample and its actual software and curve description rather than invent calendar endpoints. A narrow confidence interval for carbon measurement does not guarantee a narrow archaeological chronology.
Back to chapters · Next chapter
12. Recognise marine and freshwater reservoir effects
Organisms can incorporate carbon from reservoirs that are not in equilibrium with the contemporary atmosphere. Marine shell, fish and some freshwater food webs can therefore produce apparent radiocarbon ages older than contemporary terrestrial material. The size of an offset varies with location, time and carbon sources. A person whose diet contained marine resources may also have bone collagen influenced by a mixed reservoir, complicating direct comparison with terrestrial samples.
In Stone Orchard’s expanded fictional packet, a freshwater shell is proposed for dating the same hearth. The class must ask whether a local freshwater reservoir offset is known and whether the shell actually represents the targeted event. Historic England’s reservoir-effect guidance distinguishes terrestrial, marine and freshwater calibration and warns against a universal correction. Pupils should not subtract a convenient fixed number from every shell date or use a terrestrial curve uncritically.
Back to chapters · Next chapter
13. Compare repeated dates without hiding outliers
Several determinations on related short-lived samples can help investigate a context’s chronology, but they are not necessarily measurements of one identical event. A sample can be residual or intrusive, and laboratory determinations have uncertainty. A statistically unusual result should not be discarded simply because it conflicts with a preferred artefact typology. It needs material identification, context review and transparent modelling or rejection criteria.
Give three fictional grain sample cards from different microcontexts, two apparently compatible in BP and a third noticeably different. Ask pupils to examine whether all three were genuinely deposited in the same burning episode before averaging them. A weighted mean may be appropriate under supported shared-event assumptions, not just because the spreadsheet can calculate one. The Bayesian guidance discusses replication, outliers, offsets and reporting. Honest discrepancy analysis is better archaeology than silently deleting the inconvenient result.
Back to chapters · Next chapter
14. Introduce Bayesian chronological modelling as explicit reasoning
Bayesian chronological models can combine radiocarbon likelihoods with independently justified archaeological information, such as stratigraphic order, to obtain date estimates for events or phase boundaries. The prior information is not an intuitive guess added invisibly; it should be documented and tested. Models can improve precision when independent evidence is genuinely informative, but poor stratigraphic relationships or hidden sample biases can produce confidently wrong estimates.
Build a harmless paper exercise in which layer A precedes layer B and both have fictional radiocarbon distributions. Ask what the order constraint contributes, and how the result would change if a later excavation showed that layer B was disturbed. Historic England recommends sensitivity analysis and explicit documentation of the model. Students should recognise that statistical sophistication is not permission to force data into a favourite historical period. Where chronology is not sufficiently constrained, the model should retain that uncertainty.
Back to chapters · Next chapter
15. Use the real seed image without inventing a date from appearance
The featured image is a US National Park Service photograph of ancient Ruppia cirrhosa seed deposits relevant to research at White Sands National Park. It is a public-domain image of real archaeological dating material, not a direct readout of a radiocarbon age. Such aquatic remains also remind students why carbon source and reservoir effects warrant investigation. The photograph cannot, on its own, date any footprint, sediment horizon or specific human activity.
Ask learners to compare an image caption with a laboratory report. The picture may identify material and source context; only reliable research supplies measured values, pretreatment, reservoir assessment and archaeological interpretation. It would be a mistake to transfer a published White Sands date to the fictional Stone Orchard hearth or to imply that all aquatic seeds carry one universal offset. The central skill is keeping illustration, real research and classroom calculation distinct.
Back to chapters · Next chapter
16. Prove transfer through a new site with conflicting samples
The final invented packet contains a timber fragment from old heartwood, two grains from a possible hearth, a freshwater shell and a stratigraphic note that one deposit may have been disturbed. Students must select suitable samples, state what each carbon source could date, identify possible offsets, and explain why uncalibrated radiocarbon ages cannot be turned directly into calendar years. They then prepare a source-aware dating plan rather than a single unsupported date.
A strong answer distinguishes measured isotope history, calibrated calendar probability and archaeological event association. It can identify a preferred grain sample while still flagging uncertainty about its deposition. The student should ask for genuine calibration and specialist review rather than manufacture a precise historical year. The gold standard of teaching Radiocarbon Dating is a defensible chronology whose material, laboratory methods, reference curves and historical interpretation can all be checked independently.
Worked sampling packet: laboratory ages are not automatically event dates
All values, source codes and interpretations in this table are **fictitious**. No calendar calibration has been undertaken on these invented numbers. The appropriate learning outcome is to distinguish sample carbon history from its proposed relationship to the hearth.
| Fictional specimen | Invented lab age | Likely interpretive issue | Best next action |
|---|---|---|---|
| EDU-RC-01, carbonised cereal grain, intact hearth context | 1,040 ± 30 ¹⁴C BP | Short-lived carbon, but verify association with final fire | Confirm context and calibrate with appropriate terrestrial curve |
| EDU-RC-02, inner oak charcoal from same fill | 1,350 ± 35 ¹⁴C BP | Old-wood effect possible; ring carbon older than burning | Identify wood and ring position; do not average blindly |
| EDU-RC-03, freshwater shell from disturbed ditch | Not measured | Reservoir and event-association problems | Consider local carbon source and context before submission |
| Modern metal object from hearth fill | Not suitable for ordinary direct ¹⁴C dating | Not a conventional organic sample | Use independently associated dated materials if relevant |
Why the numbers matter: EDU-RC-02 is 310 conventional radiocarbon years older than EDU-RC-01 as reported, but this difference is not automatically a 310-calendar-year gap between two human activities. The carbon may have originated in an older tree ring, samples may be mixed, and calibration is not a single subtraction. An independently justified sample selection and calibration plan are required before historical interpretation.
Case A — the exact calendar year generated by the wrong subtraction
A fabricated school report states ‘EDU-RC-01 equals AD 910 because 1950 minus 1040 is 910’. Although its arithmetic is correct, its archaeological conclusion is not a valid calibration. Radiocarbon BP is a conventional laboratory measure; atmospheric radiocarbon varied in the past and a calibration curve translates the measured age into a probability distribution over possible calendar times. The relevant curve must fit the carbon reservoir and era. A calendar date must be reported with the calibration method, curve version and probability convention.
Ask pupils to prepare a three-column correction: raw age and lab identifier, calibrated calendar probability and archaeological event association. Leave the second column unfilled until a real calibration has been performed. Then compare with the Historic England section on calibration. The strongest response does not invent a plausible AD range merely to make the report look complete.
Case B — the old oak charcoal is scientifically dated but historically misleading
A tree’s inner rings can be much older than its outer rings when the wood is cut. If old heartwood is burned shortly after felling, the dated carbon from the inner ring may predate that burning by centuries. A perfectly measured age can therefore be inappropriate for dating the fire that occurred when the charcoal entered its final context. Students should distinguish measurement validity from event association, rather than call the laboratory result an error solely because it disagrees with the grain.
Provide two hypothetical ring positions from one old oak tree. Ask which could have grown closer to the tree’s felling and why ring position matters. In a professionally sampled timber, dendrochronology can identify specific years of growth, while radiocarbon can be used for calibration and chronological modelling. The student should understand that the technique dates sampled carbon, not all events involving the object.
Case C — reservoir offsets in fish, shells and human diets
A freshwater shell may incorporate carbon with a different radiocarbon history from atmospheric carbon, and marine organisms can also be affected by reservoir effects. The magnitude and uncertainty depend on ecological and regional circumstances. A conventional age on shell material cannot be calibrated responsibly by subtracting one universal number of years. Some human bone samples may contain carbon from mixed dietary reservoirs and require specific modelling.
Ask learners which material would more directly date a hearth using terrestrial cereal grains versus freshwater shell fragments, assuming both were securely associated. The preferred choice depends on source suitability, and a strong report explains the carbon pathway for each rather than make a rigid ‘shell never works’ claim. Historic England discusses appropriate marine and freshwater corrections and why specialists are needed when mixed reservoirs complicate the sample.
Case D — a Bayesian model should not conceal disputed contexts
A fictitious site has an older deposit below a younger one, with two suitable dates but a suspected intrusive pit. A Bayesian chronological model can incorporate reliable sequence constraints, yet the intrusive cut might make one apparent association incorrect. Adding more confident priors cannot turn a false stratigraphic interpretation into a secure age. Model assumptions should be declared and tested with alternative plausible site histories.
Students draw the two layers and pit cut before proposing an age model. They should write one scenario in which the samples genuinely reflect the sequence and another in which one is intrusive. Explain what would change and what new stratigraphic evidence would discriminate the cases. The Bayesian process guidance emphasises documented sampling, outlier treatment and sensitivity analyses.
Case E — a real seed photograph does not provide a dating result
The real White Sands photograph shows ancient seeds associated with radiocarbon research and can help readers visualise organic dating material. It does not display the measured ¹⁴C ratio, specimen’s provenance details or the actual calendar calibration. Aquatic material can raise reservoir-effect questions, making it especially important to distinguish a scientific investigation and its later interpretations from an illustrative image.
Ask students to write a museum caption stating what the photo shows and a separate statement of what a laboratory report would need to show before an age could be quoted. They should not attribute the fabricated Stone Orchard ages to the National Park Service. This transfer skill also applies to viral images of ancient artefacts: an authentic image does not authenticate every date circulating beside it.
Four lessons from isotope to responsible date
- Lesson 1 — Material and event: distinguish organic sample age from the historic activity, with short-lived grain and old-wood examples.
- Lesson 2 — Laboratory literacy: explain radioactive decay, the conventional BP scale, sample codes and contamination checks.
- Lesson 3 — Calendar modelling: investigate calibration curves, probability intervals, reservoir effects and independent stratigraphy.
- Lesson 4 — Transfer: use a new fictitious burial or hearth packet to choose samples and explain why an exact calendar year remains unjustified.
Frequently asked radiocarbon dating questions
Does 1,000 BP mean exactly AD 950?
No. BP uses AD 1950 as a reference for conventional radiocarbon ages, which require calendar calibration and an uncertainty statement.
Does radiocarbon dating measure when a stone axe was made?
Ordinary radiocarbon dating concerns carbon in suitable organic material, not stone manufacture itself. Association must be established.
Why can charcoal in a hearth seem older than a seed in it?
Old heartwood carbon may predate burning; residual material and mixing are other explanations needing investigation.
Can a marine shell be calibrated using the ordinary terrestrial curve?
It may require the appropriate marine reference and local reservoir correction. Source carbon history determines the method.
Why are calibrated date ranges sometimes discontinuous?
Atmospheric radiocarbon history and calibration-curve shape can make multiple calendar windows consistent with one result.
Do more precise lab measurements always produce an exact historical date?
No. Sample association, calibration plateaus, reservoir effects and other uncertainties limit event dating.
Official radiocarbon science, calibration and sampling references
- Historic England: Radiocarbon Dating and Chronological Modelling — Guidelines and Best Practice
- Historic England: Radiocarbon Dating, Calibration and Reservoir Effects
- Historic England: The Bayesian Process and Sample Selection
- Historic England: Bayesian Chronological Modelling
- Historic England: Scientific Dating in Archaeology
- US National Park Service: White Sands Ancient Seed Photograph (Public Domain)
Related eduKate archaeology and chronology teaching
- Teaching Chronology | Dating Sources and Ordering Events
- Teaching Dendrochronology | Tree Rings and Crossdating
- Teaching Archaeomagnetic Dating | Ancient Fire and Field Curves
- Teaching Archaeobotany | Carbonised Plant Remains
- Teaching Geoarchaeology | Sediments and Site Formation
- Teaching Statistics | Valid Inference and Uncertainty
- The Gold Standard master hub
The gold standard of teaching Radiocarbon Dating is a learner who can identify suitable biological material, distinguish measured carbon age from a human event, use calibration and uncertainty appropriately, and explain which independently documented observations would strengthen or overturn the proposed chronology.