The art of time travel: exploring radiocarbon dating in cultural heritage

Abstract

In the world of cultural heritage, time is everything. Whether a curator is authenticating a medieval manuscript or a conservator is tracing the history of an ancient wooden panel, knowing when something was made is often just as important as knowing what it is. Radiocarbon dating – often termed ¹⁴C dating – has long been a trusted tool for establishing timelines in archaeology, but its role in the arts is less widely understood. How can a technique that measures the age of prehistoric bones or ancient campfires help verify a canvas thought to be from the Renaissance? The answer lies in the organic and inorganic materials containing carbon and embedded in art objects – canvas, wood, stucco, parchment and even lacquer – all of which may hold radiocarbon clues.

---

1 Radiocarbon dating

Radiocarbon dating (¹⁴C) is based on a simple idea: living beings absorb radioactive carbon (¹⁴C) from the atmosphere, and when they die, the ¹⁴C begins to decay at a predictable rate (half-life = 5700 ± 30 years) making it an excellent clock for dates up to about 55 000 years ago.^1–3 By measuring how much remains, we can estimate how long ago the organism stopped exchanging carbon with the environment. Careful physical cleaning, close inspection and tailored chemical pre-treatment are essential to isolate the original components of the sample, taking into account the material type, state of preservation and possible contaminants.

This method, developed in the late 1940s, revolutionised archaeology and, with the advent of accelerator mass spectrometry (AMS), became an essential tool in heritage science due to reduced sample size requirements; where several grams, or even kilograms, of sample were necessary before, with AMS, only a few milligrams are needed (even less using direct measurement on a gas but the uncertainty of the measurement is higher). AMS allows rare isotopes (such as ¹⁴C) to be measured with high sensitivity by separating them from the far more abundant ones. However, environmental and anthropogenic factors have introduced complications.

For instance, the Little Ice Age (especially 1650–1950 CE) altered atmospheric ¹⁴C levels, creating a plateau in the calibration curve and making it difficult to distinguish between objects made in the 17th, 18th or 19th centuries. The Suess effect, caused by fossil fuel emissions, diluted atmospheric ¹⁴C. Later, atmospheric nuclear testing produced the bomb peak, a sharp spike in ¹⁴C levels during the 1950s and 1960s, which is sometimes useful for dating modern objects but can also complicate interpretation. In heritage science, these quirks mean that interpreting ¹⁴C dates – especially for artworks from the last few centuries – requires more than just lab results. It takes context, calibration and sometimes, a little detective work.

Calibration curves, built from extensive tree-ring records and other precisely dated datasets, are constantly being improved, and ensure that radiocarbon dating remains a robust tool for estimating age. The calibration and error of the measurement are key elements on how narrow the time range can be, it may vary from less than 50 to 200 years⁴ according to the shape of the calibration curve at a specific moment in time. Fig. 1 shows examples on how to read radiocarbon results from a radiocarbon concentration to a calibrated (cal) date. The radiocarbon determination, expressed in BP (before present – before the reference year for ¹⁴C, which is 1950 CE), needs to be entered in a calibration program (such as OxCal, CALIB or ChronoModel, available free of charge online) to obtain a range of dates with a probability of 95.4% (2 standard deviations). In the example shown in Fig. 1a, 350 ± 25 BP, we can state with 95.4% confidence that the sample dates to between 1465 calCE and 1635 calCE. More precisely, the sample dates to between 1465 and 1530 calCE, with 39.5% confidence, and to between 1540 and 1635 calCE, with 55.9% confidence. For the second example (Fig. 1b), the radiocarbon determination of 500 ± 25 BP indicates, with 95.4% confidence, that the sample dates between 1405 calCE and 1445 calCE. The interval is reduced because the slope of the calibration curve is much higher in Fig. 1b compared with Fig. 1a where the curve forms a plateau. A higher uncertainty of the radiocarbon determination will also result in a larger possible time frame.

Fig. 1 How to read the calibration of a radiocarbon measurement and the effect of the position on the calibration curve. Examples of a radiocarbon concentration: (a) 350 ± 25 BP and (b) 500 ± 25 BP. The results were obtained with the program OxCal v4.4.4² and the calibration curve IntCal20.⁵ The IntCal20 raw curve often contains data points every year, OxCal frequently defaults to a 5 year resolution to strike a balance between accuracy and calculation speed during calibration. This was the case for these two graphs. The red curve represents the concentration of radiocarbon in the sample. The blue curves show radiocarbon measurements from tree rings, with the upper and lower bounds corresponding to ±1 standard deviation. The calibrated result is expressed as a probability distribution, with 95.4% of the probability falling within the range equivalent to two standard deviations (often referred to as the 95% confidence interval).

Because radiocarbon dating is a destructive analytical technique, its application must be carefully considered. Numerous academic and private laboratories possess the expertise and instrumentation required to perform such analyses (a list is available on https://radiocarbon.org (ref. 6)), with costs typically amounting to several hundred pounds, depending on the sample type and the number of samples submitted. To protect cultural heritage, about 50 radiocarbon laboratories have agreed to follow a specific procedure when analysing antique objects requested by non-research clients.⁷

This brief explores how ¹⁴C dating is applied to cultural objects, especially artworks. We will look at which materials can be analysed, with a case study, and the challenges that must be navigated with care.

2 What can be dated?

¹⁴C dating works on materials of biological origin or on inorganic materials that incorporate atmospheric CO₂ during production. This includes a surprising variety of artistic and heritage materials:

• Wood – found in panel paintings, sculptures, furniture and architecture.

• Textiles – such as linen, cotton, silk and wool used in tapestries, garments or painting supports.

• Ivory – from sculptures, jewellery or inlays.

• Stucco – if it contains organic inclusions such as straw or animal hair, or the carbonated lime binder.

• Lacquer – derived from tree resin, commonly found in East Asian art and decorative objects.

• Paintings – especially those on canvas, but also binders such as drying oils or even pigments such as lead white containing carbonates.

• Paper and ink – paper fibres, such as flax, hemp, kozo, mitsumata, and organic pigments and binders in ink.

All these materials contain carbon that can, in principle, be dated – but each poses its own interpretive challenges. It is generally preferable to date a clearly identified single specimen rather than a mixture of unidentified materials. Another important aspect to bear in mind is that radiocarbon dating will tell how old a material is, but not necessarily when it was used. Long-term storage, reuse or restoration materials can mislead. Caution should be exercised when restoring objects if radiocarbon dating is planned. Sampling should be performed prior to restoration, or alternatively, only materials that can be easily removed should be used, and the restoration procedure should be thoroughly documented. For example, petroleum-based materials such as acrylic resins (e.g., Paraloid B-67) are radiocarbon-free. If these materials are not fully removed before analysis, the resulting date can be significantly too old – sometimes by centuries.

2.1 Wood

Wood is frequently used in heritage objects and remains one of the most commonly dated materials. It can indicate when a tree was felled, giving a terminus post quem for the object's creation. However, the ‘old wood effect’ – where heartwood is significantly older than the actual use date – can introduce uncertainty. For sculptures, where sapwood and bark are removed, the effect is often limited.

2.2 Textiles

Textiles such as linen, cotton, wool and silk all originate from biological sources, that is, plants or animals. Linen and cotton are cellulose-based, while wool and silk are protein-based materials. In heritage contexts, textiles are often found as supports for paintings, components of costumes, or as decorative arts.

2.3 Ivory

Ivory, a form of dentine, is often dated through the extraction of collagen. It is most obviously present in elephant tusks, but is also found in narwhal, walrus and hippopotamus teeth. Conservation or trade laws (e.g., CITES, which regulates trade in endangered species) sometimes require age estimates with high certainty. Thus, ¹⁴C dating must be interpreted with care, especially around recent centuries where bomb-peak ¹⁴C levels can sometimes complicate the results. Dating sections from different periods of growth can help reveal whether the sample comes from the rising or falling side of the bomb peak.

2.4 Stucco

Stucco or renders are decorative materials which are applied wet and harden into a durable finish used for wall surfaces, ceilings and sculptural ornamentation. It is made from an aggregate (often sand), and a binder, such as lime, gypsum, earth or Portland cement in more modern variants. Although primarily inorganic, stucco may contain organic inclusions, such as straw or hair, which can be dated. In lime-based stuccoes, the carbonated binder also captures atmospheric CO₂, enabling radiocarbon analysis of the setting date.

2.5 Lacquer

Lacquer, made from tree sap, contains datable carbon. It is especially important in the analysis of East Asian decorative arts. Multilayered lacquers or restoration materials can introduce mixed ages; sampling must therefore be precise.

2.6 Paintings

Paintings are complex multilayered objects. While the canvas can be dated, so too can binders such as linseed oil or synthetic pigments, for example, lead white. The latter incorporates atmospheric CO₂ during its manufacture. Binders may degrade or mix with later additions. Some pigments or treatments can introduce additional sources of carbon that can distort the results.

2.7 Paper and ink

Paper is typically made from plant fibres (e.g., linen rags, wood pulp, cotton). The major challenge lies in sampling: modern contamination from handling, conservation treatments or pollution can introduce new carbon. The 19th-century shift from rag-based to wood-pulp paper is another factor that must be considered when interpreting results.

Ink poses a greater challenge. Many historical inks, especially iron gall ink, are made from a mix of plant-derived tannins, iron salts and binders such as gum Arabic. While the organic portion can, in theory, be dated, it is often present in tiny amounts and mixed with carbon from other sources, including soot or charcoal (carbon black), which may already be centuries old when used because of the old wood effect.

3 Case study – dating a poly-lobed Asian lacquer box

A carved poly-lobed box (from a private collection) offered a rare opportunity to date multiple components of a single object:⁸ a wooden core, a textile interlayer and both red (exterior) and black (interior) lacquer layers (Fig. 2). All components returned consistent radiocarbon dates, suggesting a coherent manufacturing phase. The most straightforward analysis that can be applied to a series of samples is the combination of multiple dates into a single estimate, thereby reducing the associated uncertainty. The average ¹⁴C age calculated (combined radiocarbon dates using the OxCal 4.4 program) is: 260 ± 12 BP so a 95.4% probability that the object was made between 1636 and 1663 calCE (93.3%) or 1788 and 1792 calCE (2.2%). The results reinforce the value of sampling multiple layers, especially when restoration is suspected, and demonstrate that even thin lacquer coatings, when pretreated correctly, can yield reliable dates.

Fig. 2 Poly-lobed Asian lacquer box. Radiocarbon analysis of the wood, textile and lacquer layers yielded consistent dates, confirming a 17th-century origin with a 93.3% level of confidence.

4 Conclusion

Radiocarbon dating is an essential tool for detecting anachronisms, but it cannot confirm authenticity. An object may contain genuinely old materials but have been assembled recently. Conversely, authentic pieces may use long-stored or recycled materials.

¹⁴C analysis is most powerful when combined with stylistic evaluation, provenance research and technical study. In short: radiocarbon dating can tell you when something is too new to be true – but never when it is certainly genuine.

This Technical Brief was prepared by Marine Wojcieszak, Gaia Ligovich and Mathieu Boudin (Royal Institute for Cultural Heritage, Belgium) on behalf of the Heritage Science Expert Working Group of the Analytical Methods Committee and approved by the AMC on 21 January 2026.

References

1. I. Hajdas, L. Calcagnile, M. Molnár, T. Varga and G. Quarta, Forensic Sci. Int., 2022, 335, 111292.
2. ORAU - Dating Services - Radiocarbon Dating, https://c14.arch.ox.ac.uk/dating.html, accessed 1 December, 2025.
3. WebsiteNI, Radiocarbon Dating | the Chrono Centre, QUB, Belfast, https://14chrono.org/radiocarbon-dating/radiocarbon-dating-background/, accessed 1 December, 2025.
4. M. Jędrzejowski, C. Hatté, M.-J. Nadeau, J. Pawlyta, A. Z. Rakowski, D. Sieczkowska-Jacyna and A. Ustrzycka, Radiocarbon, 2025, 1–8.
5. P. J. Reimer, W. E. Austin, E. Bard, A. Bayliss, P. G. Blackwell, C. B. Ramsey, M. Butzin, H. Cheng, R. L. Edwards and M. Friedrich, Radiocarbon, 2020, 62, 725–757.
6. Laboratories | Radiocarbon, https://radiocarbon.org/laboratories, accessed 5 February 5, 2026.
7. I. Hajdas, A. T. Jull, E. Huysecom, A. Mayor, M.-A. Renold, H.-A. Synal, C. Hatté, W. Hong, D. Chivall and L. Beck, Radiocarbon, 2019, 61, 1133–1134.
8. M. Wojcieszak, J. Veenhoven, T. Van den Brande, S. Saverwyns, F. Lynen and M. Boudin, Radiocarbon, 2023, 65, 1193–1212.

---