The Tree Keeps the Year, Not the Voyage

Someone cut wood at L’Anse aux Meadows in the year 1021.

That is a wonderfully small historical event to know so precisely. Not a reign, a century or an approximate Viking Age. A year in which trees in Newfoundland became material for people’s work.

In a study published online in 2021, Margot Kuitems and colleagues dated three pieces of Norse-worked wood to that year. The objects were not three grand beams conveniently labeled by a builder: they included a log, a stump with cut stems and a branch. Their connection to Norse activity came from archaeological context and metal-tool marks. Their calendar came, in part, from something that happened in the atmosphere decades before the cuts. [1]

The tempting headline is that scientists discovered exactly when the Vikings arrived in America. That is not what the wood establishes.

The more interesting story is how it becomes possible to know an activity’s year without knowing whether it happened near the beginning or end of a visit.

A clock with a changing starting point

Radiocarbon dating sounds like a straightforward countdown. Living plants take in carbon, including radioactive carbon-14. Once incorporated into wood, that carbon’s isotope composition changes as carbon-14 decays. Measure what remains and infer an age.

But the starting mixture has not stayed constant.

Carbon-14 is produced through cosmic-ray interactions in the atmosphere. Production and the carbon cycle change its atmospheric abundance. An old sample cannot be assigned a calendar year simply by assuming it began with today’s proportion. Known-age tree rings help supply the missing history: their carbon can be measured and associated with the calendar years in which the wood formed. Overlapping tree-ring chronologies extend that reference beyond an individual tree’s lifetime. [2]

Even with that reference, a measurement can correspond to a range of years—or several ranges. The instruments have finite precision, and the atmospheric history has wiggles. A better measurement does not necessarily give a uniquely positioned calendar date.

The way out, in this case, was not to demand that one measurement become impossibly exact. It was to recognize a shape across successive rings.

A sudden change the trees retained

In 2013, Fusa Miyake, Kimiaki Masuda and Toshio Nakamura reported a rapid carbon-14 increase between AD 993 and 994 in Japanese cedar rings. They measured a longer sequence at two-year resolution and examined the neighborhood of the event annually. The abrupt rise, followed by a decline shaped by the carbon cycle, supplied a distinctive feature. Averaging over longer intervals blurred it into something much less remarkable. [3]

The paper discussed a solar origin as plausible. We do not need to turn that into a reconstructed spectacle: a particular flare, an ancient eyewitness or a precise description of the sky. The dating method needs a securely positioned isotope pattern, not a complete biography of its astrophysical cause.

There is a revealing wrinkle in the original discovery. Its published chronology was corrected later in 2013. A missing annual layer around AD 956 meant subsequent ring dates had to move forward one year. What had been described as a 992–993 increase became 993–994. The updated article includes the correction. [3, Change history]

A chemical signal can be real while its first calendar label is wrong. Counting the rings is part of the measurement, not the easy bit that happens after science is finished.

Getting from the marker to the edge

For the Newfoundland wood, Kuitems and colleagues first matched sequences of radiocarbon measurements against the IntCal20 calibration curve. That narrowed the terminal growth dates to AD 1019–1024. They then compared six strategically placed rings from each piece with an annually resolved reference containing the sharp event, called the “AD 993 anomaly” in their paper. The finer reference preserved a feature that the smoother curve made less distinct. [1, “Precise dating of Norse activity”]

All three pieces had their best match at an outer-edge date of 1021. Other years did pass the individual statistical test, so this was not three separate mathematical declarations that every alternative was impossible. The shared best positioning, the location of the sharp change and agreement among distinct trees supported the year together.

The counting can be made concrete without pretending to reproduce the laboratory analysis. The authors numbered the outermost growth ring zero. In their preferred positioning, ring minus twenty-nine corresponds to AD 992. Twenty-nine annual intervals later is 1021. In that positioning, ring minus twenty-eight is 993 and ring minus twenty-seven is 994—consistent with the corrected event chronology. There are thirty ring labels if you count both endpoints; there are only twenty-nine intervals between them.

The crucial object was not just a ring somewhere inside a tree. It was the preserved outer growth boundary, the waney edge. With that edge present, the sequence could be followed to the tree’s last growth. Without it, a perfectly dated interior ring would still leave an unknown amount of missing exterior wood.

A fourth wooden item was examined but lacked the event in its short sequence. It could not be dated precisely by this method. The result was three dated trees, not four. [1]

Carbon does not recognize an axe

Even a terminal growth date is not automatically the date of a human action. Someone can cut or reuse old wood.

The researchers argued against a long delay here. The preserved edges made driftwood less plausible; fresh wood was abundant locally; and three different trees converging on the same year was difficult to explain as unrelated pieces of scavenged material with matching delays. Those arguments connect the biological date to the archaeological activity. They are not an isotope measurement of a blade striking wood. [1]

Nor does carbon identify the people. The attribution depends on where the pieces were found and their characteristic metal-tool cuts, assessed within the site’s archaeology. The supplementary account describes Indigenous stone-tool-worked wood in layers below and above the Norse stratum. This is a dated episode in an inhabited landscape, not a date when the continent began to have a human history. [1, Supplementary Notes 1 and 4]

That is the distinction I want to keep: a very precise material date can leave a historical boundary unresolved.

The wood supports Norse activity at L’Anse aux Meadows in 1021. It does not name the cutter, identify a saga’s expedition, establish first landfall or tell us the total duration of the occupation. The date tightens the history without completing it.

I find the remaining question better than the inflated headline. If people were cutting these trees in 1021, where does that year sit inside their visits? Near an arrival, during a return, toward a departure?

The tree keeps the year. The voyage still needs other evidence.

Sources and reading limits

Kuitems et al., “Evidence for European presence in the Americas in AD 1021”, Nature 601, 388–391 (2022; published online 20 October 2021). Main text, Methods, main and extended captions, and Supplementary Notes 1–4 consulted. The notes belong to this same study, not independent corroboration. No figure-image inspection, source-data reanalysis or model rerun claimed.
Oxford Radiocarbon Accelerator Unit, “Radiocarbon Calibration”. Complete explanatory text consulted; an authoritative account of calibration, not a new experiment.
Miyake, Masuda & Nakamura, “Another rapid event in the carbon-14 content of tree rings”, Nature Communications 4:1748 (2013), updated with its 7 November 2013 correction. Main text, Methods, captions and correction consulted, not the supplementary material. Independent original event measurements, not a separate replication of the Norse date.

The interval arithmetic above is a check of the published indexing, not a new ring count. No fieldwork, wood sampling or isotope measurement was performed for this essay.