Thousands of stones along Europe's Atlantic edge carry the same pecked sign: a cup, a set of rings and a channel running out. Here is what the records say about when, where and why.
Walk up onto Ilkley Moor and you will find the Badger Stone, a gritstone boulder covered in cups and rings. Fly to Marín on the Galician coast and you will find the Mogor stones, granite outcrops carved with rings and labyrinths. The two places are 1,200 kilometres apart, separated by the Bay of Biscay, and they share a visual language.
This post pulls together the open records that now exist for these carvings: Scotland's Canmore register, Ireland's national rock-art layer, Wikidata and OpenStreetMap for England, Galicia, Portugal and Brittany, and the published counts and dates for each. Then it asks the three questions everyone asks: how old are they, why are they where they are, and what were they for. At the end I set out my own reading of the evidence.
The basic mark is a cup: a shallow hollow, usually 2 to 6 centimetres across, pecked into rock with a harder hammer stone. Around some cups the carver added one or more rings, sometimes broken by a straight groove, the gutter, that runs from the centre out past the rings and often downhill. Scotland's Rock Art Project counts nine rings around one cup at Castleton. Most carvings are simpler: in Scotland's register two cup-only rocks are recorded for every one with rings.
The figure below is a working model of the sign. Change it, and watch how fast the effort grows.
Anatomy of a carving
Each groove is drawn as a line of peck marks. The readouts show the total length of groove a carver would have had to peck.
Groove length is the sum of every ring's circumference, less the gap where the gutter breaks it, plus the gutter. A ring's length grows with its radius, so each new ring costs more than the last: with nine rings at 2 cm spacing the outermost ring alone is longer than the first four together.
Rock carvings are hard to date. You cannot radiocarbon-date a pecked groove, so archaeologists date what touches it: a cairn built over a carved stone, a broken slab reused in a tomb, charcoal in the soil around an outcrop, or a carved weapon whose metal type is known. Each gives a "no later than" or "around the same time as" date, not a birth certificate.
Put the national evidence side by side and a sequence appears. Brittany's carved megalithic tombs and standing stones are among the oldest, from the fifth and fourth millennia BC. Britain and Ireland follow from about 4000 to 3700 BC, with the open-air rocks mostly Late Neolithic and Early Bronze Age. Galicia and northern Portugal are the late bloomers: most of their open-air art is placed in the Bronze Age, and some researchers push part of it into the Iron Age.
When each Atlantic region carved, and who was living there
Bars show the main period each national literature gives. Diamonds mark dated evidence; Ilkley and Mogor are labelled. The lower strip shows the ancient-DNA turnovers. Hover or tap a mark for detail.
Sources: Scotland's Rock Art Project; Kilmartin Museum; Historic England list entry 1012012; National Monuments Service; Galician studies in Adoranten; Brace and others 2019; Cassidy and others 2016; Olalde and others 2018 and 2019. Full links in the notes.
That lower strip matters. The first farmers reached Britain around 4000 BC with ancestry that traces back, through the Mediterranean and the Atlantic coast, to Anatolia and the Aegean. Around 2450 BC, with the Beaker culture, about nine-tenths of Britain's gene pool was replaced within a few centuries by people carrying steppe-related ancestry. In Iberia the same arrival around 2500 BC replaced about 40% of the ancestry and nearly all the male lines by 2000 BC.
The carvings did not stop. In Britain people went on carving, and reusing carved stones in burial cairns, well into the Bronze Age. In Galicia the main period of open-air carving comes after the steppe arrival. Whatever the rings meant, they were not tied to one bloodline. They were tied to places, and new people picked them up.
The map folds the seaboard so that both ends fit on one screen: Britain and Ireland on the left, Brittany, Galicia and northern Portugal on the right, at the same scale. Every dot is a carved rock from an open register. Scroll or pinch to zoom, drag to pan, and tap a dot for its record.
4,584 carved rocks on the Atlantic edge
Coloured by what the record says is on the rock. Ireland and Galicia's open records do not separate cups from rings, so they show as "carved rock".
Scotland: Historic Environment Scotland, Canmore (2,845 cup and cup-and-ring records). Ireland: National Monuments Service rock-art layer (826 in place, after dropping 68 stones moved to museums and gardens and 56 with a placeholder location). England and Wales: Wikidata, mostly scheduled monuments. Galicia, Portugal, Brittany: Wikidata and OpenStreetMap, with duplicates within 150 m merged.
Two things jump out. The first is the clustering: Argyll and Perthshire in Scotland, the Iveragh peninsula in Kerry and Inishowen in Donegal, the Pennine moors around Ilkley, and the Pontevedra coast in Galicia. The second is the gaps, and some of those are gaps in recording, not in carving. The chart below compares what the open records hold with what specialists say exists.
Open records against published counts
Dark bar: rocks in this post's map. Outline: the count specialists publish for the same area.
Scotland's Rock Art Project counts 3,210 known panels; Rodríguez Rellán, Vázquez Martínez and Fábregas Valcarce (2018) count 3,374 Galician stations; England's Rock Art project recorded more than 1,500 panels in Northumberland and Durham alone.
Scotland's register is nearly complete; Galicia's open data shows only about one rock in seven; England's open data covers mainly scheduled monuments. So the map is honest about Scotland and Ireland and only a sketch elsewhere. Treat the densities as minimums.
Cups are everywhere; rings are the special case. In Scotland a third of the records have rings. In Galicia cups appear at 62% of the 3,374 stations and circular figures at 44%, with animals, weapons and labyrinths far behind. Mogor's labyrinths belong to a motif found at just 24 Galician stations, less than 1%.
Scotland: rings belong on bedrock
Share of Canmore records with rings, for carved outcrops ("rock") and for loose boulders and slabs ("stone").
Galicia: what each station carries
Share of 3,374 rock-art stations with each motif. A station can carry several.
The Scottish split is worth a second look. On fixed outcrops, 44% of records have rings. On loose boulders and slabs it is 24%. Rings were more often carved on rock that could not be moved. That is a small, checkable clue that the ringed sign was about a place, not a portable object.
If the motif travelled along the Atlantic by boat, as many archaeologists think farming did, you would expect it to cling to the coast. I tested that directly. For each mapped rock I measured the distance to the nearest coastline and compared it with 4,000 random points on land in the same region.
Carved rocks sit closer to the sea than the land around them
Share within 5 km of the coast. Dot: carved rocks. Ring: random points on land in the same region.
Coastline from Natural Earth 1:10m, measured in an equal-area projection. Scotland excludes 143 records flagged "possible". Survey effort is not even, so treat this as a strong pattern, not a precise rate.
The western heartlands are strongly coastal. In Ireland 55% of rock art lies within 5 km of the sea, against 16% of the island's land. Galicia shows the same pull, and so does Scotland. England is the exception: its carvings are on the Pennine and Northumbrian moors, far from salt water. That fits a picture where the sign arrived by sea in the west and then moved inland along upland routes.
Why concentric circles, of all shapes? There are three good mathematical answers, and they point in different directions.
A circle is the only curve you can make by keeping one distance fixed. Pin a cord at the cup, hold a pointed stone at the other end, and the circle draws itself. Add another knot and you get the next ring. That is why concentric rings turn up wherever people peck rock: the shape needs no plan, just a fixed centre and a cord. It also explains why a ring broken by a gutter still comes out round.
The commonest folk explanation is ripples on water. The physics says no. Waves on water spread out at different speeds depending on their length, so the rings from a dropped pebble are never evenly spaced: the pattern stretches as it travels and the gap between crests changes from front to back. Carved rings are evenly spaced. Run the model below and compare.
A real ripple against a carving
Left: a small splash, 3 mm across, on still water, computed from the equations above. Right: rings at the even spacing typical of carvings. Ticks under each show where the crests are.
The deepest answer may be inside the head. In 1926 Heinrich Klüver noticed that people under the influence of mescaline see the same few geometric forms: lattices, cobwebs, tunnels and spirals. Later work showed why. The map from the retina to the primary visual cortex is close to a complex logarithm: a point at distance r and angle θ from the centre of gaze lands at position (log r, θ) on the cortex. Under that map a set of concentric rings becomes a set of straight parallel stripes, and stripes are the simplest pattern a sheet of nerve cells can fall into when it is overexcited.
From eye to cortex
Pick a pattern. Left is what you see; right is the same pattern as it falls on the visual cortex.
David Lewis-Williams and Thomas Dowson built this into a theory of rock art in 1988: geometric signs record trance visions. You do not need drugs or trance for the maths to matter, though. Rings are a shape the human visual system is built to find striking. That helps explain why they appear on every inhabited continent.
One more number, because it has a famous history. In the 1960s and 70s the engineer Alexander Thom argued that the carvers used a standard unit, the "megalithic inch" of 2.07 cm. The readout in the anatomy model shows radii in that unit. A later test on English carvings by Alan Davis found the evidence weak and not statistically significant. Even spacing does not need a standard unit; a hand or a finger width will do.
The Atlantic carvings are one regional chapter of something much older and wider. Cupules, the simplest pecked hollows, are found on every continent except Antarctica, and some may be among the oldest marks humans ever made. Concentric circles appear in the Alps, Sardinia, the Levant, India, Hawaii and the Americas.
Cups and rings around the world
A hand-picked set of well-documented places, coloured by whether the main mark is the cup alone or the ring. Tap a point for its story.
Deep time, on a log scale
Years before today. Each step to the left is ten times older.
The striking thing about the wider record is that wherever we actually know what cups meant, because someone wrote it down, the meanings are local and practical. Hawaiian families placed a newborn's umbilical cord in a hole at Pu'uloa for long life. Pomo women in California pecked "baby rocks" when they wanted to conceive. Shasta people made "rain rocks". In Estonia and Finland almost all cup stones sit in farmland. The shape is shared; the meaning is not.
Archaeologists have argued about this for two centuries; one review lists 71 suggested meanings for cupules. These are the main theories for the Atlantic carvings, with the evidence for and against each.
Putting the data together, I think the cup-and-ring is best understood as three layers stacked on top of each other.
The cup is the simplest mark a hammer stone makes, and the ring is the simplest shape a cord makes. The visual system finds both compelling. That is why they appear across the world, independently, from the Palaeolithic onwards.
The particular combination of cup, rings and gutter, carved on open bedrock, belongs to the Atlantic seaboard from about 4000 BC. It sits close to the sea in Ireland, Scotland and Galicia, which fits a motif spread by sea along the routes the first farmers used.
The meaning was rewritten by each community that inherited it. The carvings carried on through the Beaker turnover in Britain and flourished in Galicia after the steppe arrival. People kept the places and the sign while their ancestry changed.
This reading makes some predictions that could be tested with better data. If the grammar travelled by sea, the earliest securely dated ringed rocks in each region should sit nearest the coast, and inland groups like Ilkley should be younger. If rings were about places, rings should favour fixed bedrock over portable stones everywhere, not just in Scotland. And if meanings were local, the details of the grammar (gutter direction, number of rings, what sits next to the carvings) should vary more between regions than within them.
What it does not support is the idea of a single "Atlantic people" with a single message. The DNA shows the people changed. The rings stayed.
Everything mapped here comes from open sources you can query yourself: Canmore for Scotland, the National Monuments Service for Ireland, and Wikidata and OpenStreetMap elsewhere. The biggest gain would come from opening Galicia's heritage inventory and England's Rock Art database as point data.