Carnac · France · 47.5928, -3.0790 · PYR record in the living atlas
At Le Plasker, in Plouharnel just west of Carnac, a rescue excavation opened a set of foundation pits averaging about 1.8 metres across, set out in at least three lines. The standing stones they once held are long gone. The organic material in and beneath them is not, and in August 2025 a team led by Audrey Blanchard, Jean-Noel Guyodo, Bettina Schulz Paulsson and Fabien Montassier published the dates in Antiquity: the alignment there was raised between 4668 and 4368 cal BC, with activity continuing to about 4249 cal BC. A whole section of the Carnac complex that nobody knew existed, with a chronology attached.
What is actually standing
The Menec rows run about 1,165 metres by 100 metres in eleven files, the tallest stones around 4 metres at the western end, dropping to about 0.6 metres partway along, with a stone enclosure of 71 blocks at the west. Kermario holds 1,029 stones over roughly 1,300 metres. Kerlescan has 555 stones in 13 lines over some 800 metres, with a circle of 39 stones at its western end. The round figure of three thousand stones is fair for the whole complex. Those three groups are not the whole of it either: the complex runs for roughly 10 km along the coast between Carnac and Erdeven, the three principal alignments run broadly north-east to south-west, and each has the partial remains of a stone enclosure, a cromlech, at one end.
What the postcards leave out is how much of this was stood back up in modern times. When the antiquarian James Miln surveyed the rows in the 1860s, fewer than 700 of the 3,000 stones were still upright. At Kermario, only 164 of 678 stones were standing before restoration began. Zacharie Le Rouzic re-erected menhirs from around 1900 onwards, and bulldozers rearranged and re-erected stones again during works in the 1930s and the 1980s. Anyone measuring an angle at Carnac today is measuring, in part, decisions taken between 1900 and 1990. That is not a reason to stop measuring. It is a reason to say so before quoting a bearing to a tenth of a degree.
The recorded extent is also an undercount. In 2026 a team led by Bruniaux published a geophysical survey in Archaeological Prospection covering about 2.5 hectares, which found buried standing stones, further segments of alignment, burial mounds and cooking pits below the modern surface. Every stone count at Carnac, this article's included, is a count of what happens to be visible.
The dates keep moving backwards
Le Plasker matters because of what sat underneath the stone sockets: a Late Mesolithic dwelling, an oval feature with a ditch around it holding granite blocks that probably steadied wooden posts, dated to 5699 to 5621 cal BC, with three small monoliths beside it. Hunter-gatherers were marking that ground roughly a thousand years before anyone raised the alignment over it. The authors' conclusion is worth quoting because it is the opposite of how these places are usually described: the creation of a cultural landscape such as Carnac "did not happen all at once, but more likely in several stages over a protracted period". UNESCO, inscribing the Megaliths of Carnac and of the shores of Morbihan on the World Heritage List in 2025, gives the whole tradition a span of about 5000 to 2300 BC. Read the late end of that span loosely: the evidence for Chalcolithic and Bronze Age use, modification and reuse is much thinner than for the early Neolithic, and some researchers argue building carried on into the second millennium BC.
Two dates for the alignments therefore circulate, and both are defensible, because they are about different things. The range of roughly 4600 to 4300 BC comes from Le Plasker, Bayesian-modelled from 49 radiocarbon dates, which makes those rows among the oldest dated megalithic monuments anywhere in Europe. The figure of around 3300 BC that older accounts give applies to Menec, Kermario and Kerlescan specifically, and those three have never been subjected to modern systematic excavation. Systematic digging in the region ran from the 1920s to the early 1940s, after which official policy shifted from excavation to preservation and the famous rows were left where they stood. So the older figure is unrefuted rather than confirmed, and the internal chronology of the complex is genuinely unresolved rather than roughly known.
There is a physical reason the dating is this loose, and it is the single most useful thing to know about the site. The soils of the Morbihan are acidic. They destroy bone and almost all other organic material, which restricts radiocarbon here to charcoal, and it is frequently impossible to demonstrate that a given piece of charcoal relates to the erection of a stone rather than to somebody's fire on the same ground centuries earlier or later. Optically stimulated luminescence has been attempted and is too imprecise to settle anything. The vagueness is in the ground, not in the archaeologists.
The builders are not in doubt: Neolithic farming and pastoralist communities of the Armorican peninsula. The Mesolithic dwelling beneath Le Plasker is a superposition, not an attribution. It shows the place already mattered before the megalithic tradition began; it does not show that hunter-gatherers raised the rows, and accounts that use it that way have turned a layer into an author. Building proceeded in stages across at least three centuries, each phase respecting and extending the layout it inherited, which is a continuity of intent across many generations who left no writing. Whether that came from centralised direction or from repeated community decisions is unknown, and so is whether the rows follow one original plan, grew piecemeal, or were modified over and over.
One much larger claim rests on this chronology. Bettina Schulz Paulsson, one of the Le Plasker authors, has argued from radiocarbon modelling across hundreds of European sites that megalithic monument-building began in the Bay of Morbihan and spread outward along Atlantic Europe from there. The 2025 dates strengthen the case and the hypothesis is gaining support, but it is not settled: other researchers still favour independent regional developments in several places at once. That is an open question with a lot riding on it, because it makes Carnac either the origin of the European megalithic tradition or one early instance of it.
The largest stone anyone raised, and what became of it
Along the same coast, at Locmariaquer, lies the Grand Menhir Brise, 20.6 metres long and the largest single block known to have been transported and erected anywhere in the Neolithic world. Published weight estimates vary, commonly between about 280 and 330 tonnes, and there is no way to weigh it now. It went up around 4700 BC as one of a row of 19 blocks, and its whole surface once carried a carved hatchet-plough motif now almost weathered away. Around 4000 BC it came down and broke into four pieces.
Why is genuinely open. One camp reads it as deliberate: the stone was toppled, broken and quarried, and pieces of it were carried off and reused, since the decorated capstone of the Table des Marchand and stone at Gavrinis carry carvings that match. The other camp favours an earthquake, with computer modelling behind it. Either way the giant was lying in pieces and being recycled thousands of years before the periods in which it is usually invoked as an instrument. The nearby Saint-Michel tumulus is the other monster of the district: an artificial hill 125 by 60 metres and 12 metres high, some 35,000 cubic metres of stone and earth, covering 15 stone chests that held jade axes, pottery and callais beads.
Thom's observatory, weighed properly
Alexander Thom and his son Archie surveyed Carnac between 1970 and 1974, and his case deserves the strong version. Thom measured more carefully than anyone before him, argued for a standard prehistoric unit he called the megalithic yard at about 2.72 feet, and read the Grand Menhir as a universal foresight for lunar observation, sighted from distant backsights around the gulf. Archaeoastronomy exists as a discipline largely because he did the work, and he was not without support: Euan MacKie found what he read as verification at Kintraw and Cultoon.
The statistics went both ways, and then mostly one way. David Kendall's analysis in 1974 is the one usually misreported: testing circle diameters against a smooth unquantised distribution, he rejected that null for Scottish circles, which is evidence for a shared unit, and found no such evidence in England. Douglas Heggie in 1981 found little evidence for a highly accurate unit. Clive Ruggles concluded in 1999 that the evidence for the megalithic yard was at best marginal and that the real uncertainties were far greater than the millimetre precision claimed, with ordinary pacing sufficient to explain the pattern. Ruggles made a second objection that applies to any precise alignment claim at this site in principle: a great many Carnac stones have been moved, re-erected or restored, and the archaeological record confirms it. Aubrey Burl called the megalithic yard a chimera and a grotesque statistical misconception, and argued separately that the precision Thom claimed could not have been achieved with the construction methods available. Add the chronology above, and the Grand Menhir was on the ground and being cut up long before the lunar geometry needed it upright. The astronomical reading of Carnac is also older than Thom: Jacques Cambry proposed a zodiac in 1805 and James Fergusson rejected it in 1872.
The line of work has not stopped. Howard Crowhurst, a researcher resident at Carnac, extends Thom's geometry and argues the alignments encode precise lunar and solar cycles. The mainstream objection to him is the same one made to Thom, and it is the objection our own engine section below describes: select alignments out of a large enough population of stones and some will land wherever you choose to look, and the results do not survive testing against a null model.
What is still unanswered
UNESCO's own justification for inscribing the site says the monuments were placed deliberately, with intervisibility between them playing a role in where they stand, and then says plainly that "it is not yet possible to explain with certainty the reasons for erecting these structures". Local tradition has its own answer, and it should be labelled as tradition rather than dismissed: the rows are a Roman legion turned to stone by Merlin, or in the Christian version pagan soldiers petrified while chasing Pope Cornelius. After three centuries of survey the honest summary is that thousands of people moved thousands of stones across many generations, it mattered enormously to them, and we do not know what it was for. Archaeology can live with that sentence, and so can we.
What our engines found
We ran the same test here as everywhere else, with the settings fixed in advance and identical at every one of the fifty sites we cover: a 200 km search radius, at least 12 km between sampled sites, 15 sites, a 2 per cent tolerance, and 2,000 Monte Carlo runs against a null model resampled from real catalogued sites in the same region. The settings are frozen because the answer moves if you tune them, and tuning them after seeing the answer is exactly the trick we object to elsewhere. Carnac returned 35 golden-ratio pairs where chance predicts 29.4, giving p = 0.2724.
That is a little more than chance would predict, and it is not significant. The control points the same way: the arbitrary ratio pi/2 scored 1.82 times its expected count against phi's 1.19, so nothing here singles out the golden ratio.
We publish the misses because a tool that only reports its hits is not measuring anything. Every figure above is reproducible from the same coordinates, the same settings and the same seed.
One limit worth stating plainly, because it is the one most often glossed over. This test compares positions between monuments. It says nothing about any building's own proportions, its height, base, slope or perimeter. That is a separate question, and at several sites a genuinely open one.
Explore it yourself
Open the Carnac record in the living atlas to see the alignments in 3D, place them among their neighbours in Morbihan, and run the alignment and ley line engines over Brittany yourself. The Carnac site-emblem tee carries the stone rows, the site geometry and the exact coordinates in the Pyramason Chronicle style.
Sources worth reading
Blanchard, Guyodo, Schulz Paulsson and Montassier, "Le Plasker in Plouharnel (fifth millennium cal BC): a newly discovered section of the megalithic complex of Carnac", Antiquity (2025). Bruniaux et al., Archaeological Prospection (2026), on the buried extent of the complex. The University of Gothenburg's summary of the Le Plasker results, for a readable version of the dating. Burl, Megalithic Brittany (Thames and Hudson, 1985). Ruggles, Astronomy in Prehistoric Britain and Ireland (Yale University Press, 1999). UNESCO's 2025 inscription documents for the Megaliths of Carnac and of the shores of Morbihan.