Sturt Manning, the lead researcher on the project, coring a centuries-old juniper tree near Petra in southern Jordan. (IMAGE: Sturt Manning, Cornell University)
In today’s era of ‘fake news’, we haven’t been entirely surprised to see recent headlines claiming new research has proven that radiocarbon dating is inaccurate or plain wrong (one even went so far as to say ‘A Crucial Archaeological Dating Tool is Wrong, and It Could Change History as We Know It’). To be fair, once you get past the headlines, the articles mostly provide a bit more of the truth and a little less clickbait. Nonetheless, we thought it pertinent to delve into the actual science of this discovery and offer a more impartial, if less sensationalist, account of the findings.
The paper, published in PNAS at the end of May (it is available for free at https://doi.org/10.1073/pnas.1719420115), was produced by a team from the Tree Ring Laboratory at Cornell University, supported by the radiocarbon laboratories at the University of Oxford and the University of Arizona, Tucson. Researchers from Cornell analysed samples of native juniper from southern Jordan, assessed their ages using dendrochronology, and then had them radiocarbon dated by both the Oxford and Arizona labs. It should be stated that this is not a new technique: tree-ring data were used to create the first radiocarbon calibration curves, and they continue to be used to refi ne it today – they are the main underpinning of the methodology.
Previously, it had been assumed that the natural carbon isotopes, including 14C, rapidly mixed in the atmosphere and that, because of this, there were unlikely to be any detectable latitudinal or regional differences in radiocarbon levels within each hemisphere. For this reason, a standard international radiocarbon calibration curve, built from known-age tree-rings from central and northern Europe and North America, has been used for dating by archaeologists and other scientists for northern hemisphere samples – the current version of this standard curve is called IntCal13.
One of the samples of juniper – from Taybet Zaman, Jordan – used in the research, with the tree rings clearly visible. (IMAGE: Sturt Manning, Cornell University)
The results of this new study, however, showed measurable differences in the radiocarbon ages from the Jordanian tree-rings versus IntCal13 for the same calendar years, upsetting the previous assumption. In fact, the radiocarbon ages obtained from the Jordanian tree-rings were on average approximately 19 years older than the radiocarbon dates in IntCal13. This was in accordance with previous radiocarbon dates for plant material with known-dates from the 18th to 19th centuries in Egypt, which also showed a discrepancy of about ±19 years. While the team only analysed tree-ring data from AD 1610-1940, they assume that a similar difference will be seen throughout the centuries in this region. When they applied a general correction to previous radiocarbon dates from the Iron Age – which they stress is just an ‘average’ correction and does not take into consideration any variations in time that are likely to exist – they found that the shifts in dates were always towards making them more recent in calendar date.
In particular, the team noted that the degree of discrepancy in 14C levels varied over time, and that it appeared to be largely driven by changes in regional climate, with warmer temperatures leading to larger differences. As Sturt Manning, the lead author on the paper, explained, ‘we see larger offsets [between] AD 1685-1782 and AD 1818-1912, and these seem to link with generally warming conditions in the southern Levant.’ These changes in climate are likely to have altered growing seasons and hence the levels of radiocarbon present in the atmosphere. For example, juniper trees in southern Jordan grow from autumn to early summer, while further north the growing period for trees (like oaks in central and northern Europe) is almost the exact opposite (late spring through late summer). In an assessment of plants in the southern Levant that are frequently recovered from archaeological contexts, the team found that approximately 77% of 14C dates from Iron Age Israel came from plants with a significantly different growing season than those in central and northern Europe, and hence are likely to be affected by a carbon offset.
While incredibly important, these new findings only affect areas in the southern Levant and mainly for projects that are trying to determine high-resolution 14C chronologies with dating precision of under 100 years. This means that these findings are unlikely to ‘change history as we know it’, but they will significantly affect many archaeologists working in the region and could alter a few timelines, especially as a lot of work has gone into developing precise dates for the region during the Late Bronze Age to Iron Age/early Biblical periods (c.1200- 600 BC). Such a problem suggests that creating a calibration curve specific to the Levant may be in order. Thus, instead of signalling the end of radiocarbon dating, these results will ultimately only make the technique stronger and more reliable, as calibration curves are able to become even more refined and regionally specific.
This article appeared in CA 341.