Britain in AD 500

There is a famous kind of map that shows Britain around AD 500 as a patchwork of kingdoms: Rheged and Elmet, Gwynedd and Dumnonia, Bernicia, Kent and the East Angles, each in its own colour with a neat edge. I like those maps. They are beautiful, and they make a lost world feel real. But almost nothing written in Britain survives from the century after Rome, and most of those kingdom names first appear in books written two, three or seven hundred years later.

So I tried to build a different map: one that starts from the ground. It plots what archaeology and the national records can actually place in the years around 500, from hillforts that people went back to, to cemeteries, inscribed memorial stones and early churches. Then it lays the named kingdoms on top, not as countries with borders but as faint, labelled zones, each with a note on where and when its name is first written down.

The result is patchy, and I will say where. But some patterns are so strong that even thin, uneven data shows them.

01 · The mapWhat can be placed around AD 500

Each dot is one piece of evidence; tap it for what is known and where the record comes from. Filled dots have their own dating evidence. Hollow dots are dated only by type, for example "ogham stones were mostly cut in the 5th and 6th centuries", so treat their dates as broad. The coloured haze is a smoothed density of the evidence for each cultural tradition. The italic names are kingdoms, placed by me at the area the sources point to: they are guides, not boundaries. Use Map layers for hills, modern place names and borders, Roman roads and low-lying land. Search a modern place to see what lay near it.

02 · Two habitsAn east of graves, a west of stones

Put the Anglo-Saxon cemeteries and settlements next to the Latin-inscribed memorial stones and the island splits in two. The 75 Anglo-Saxon sites in my data have a median longitude of 0.05 degrees west, almost on the Greenwich meridian; for the 143 inscribed stones and western high-status sites it is 4.09 degrees west, in the middle of Wales. 99% of the Anglo-Saxon sites lie east of 2 degrees west, and 96% of the stones lie west of 3 degrees west. Just 1 of those western sites has an Anglo-Saxon site within 50 km. Smoothed into density surfaces, the two never overlap: there is no square kilometre where both rise above a tenth of their peak.

This is a real pattern, and archaeologists have described it for a long time, but it is worth being clear about what it is. It is a map of two habits, not two peoples. In the east, people buried their dead with brooches, beads and weapons, or cremated them in urns, in the way that came across the North Sea in the 5th century. In the west, Christian Britons set up memorial stones carved with Latin names, often with Roman-style titles. Somebody living in Kent who never owned a brooch, or a family in Wales too poor for a stone, is invisible on this map. Even so, the gap in the middle is striking: a band from the Pennines to Dorset where neither habit shows in my data.

03 · Back to the hillsThe hillforts people went back to

Most hillforts were built in the Iron Age, centuries before Rome. The Atlas of Hillforts of Britain and Ireland, compiled by Gary Lock and Ian Ralston, records 4,147 of them, with a field for every period in which there is dating evidence. Just 87 (2.1%) have evidence for use between AD 400 and 800. They are not spread evenly: 48 are in Scotland, 14 in England, 6 in Wales and 14 on the Isle of Man, where that is 47% of the island's 30 hillforts.

The ones that were reused look different from the average hillfort. 39% are on the coast, against 18% of all hillforts, about 2.2 times the rate; 45% are promontory forts, against 30%. That fits what excavation has found at the famous sites: these were small, defended seats of chiefs and kings on headlands and crags by the sea, the places where wine and fine pottery from the Mediterranean and Gaul came ashore.

The dating is uneven. The Atlas rates its own confidence: for 36 of the 87 the evidence is rated high or medium, and for 28 it is rated as none, meaning the early medieval date rests on something like a place-name or a stray find. 27 of the dating notes mention radiocarbon dates.

04 · Across the Irish SeaOgham, Dyfed and the western sea

Ogham is an alphabet of strokes and notches cut along the edge of a stone, used mainly to write the early Irish language. My map has 402 ogham find-spots, merged from 1,838 records in Wikidata and Ireland's Sites and Monuments Record, and 97% of them are in Ireland. Wikipedia's article on ogham says the stones flourished in the 5th and 6th centuries around the Irish Sea, and that the largest number outside Ireland is in Pembrokeshire.

That second point is where my data falls short, and honestly so. In Wikidata most Welsh stones are recorded only as "inscribed stones", whether or not they also carry ogham, so the ogham layer shows just 2 in Wales. What the map can show is the stones themselves: 92% of the 139 inscribed stones are in Wales, and 49 of them (35%) are in the old county of Dyfed in the south-west. The written sources point the same way. Wikipedia's article on the kingdom of Dyfed says its royal line was founded by Irish settlers in the 5th century. A memorial stone from Castell Dwyran, of about 550, names a king in Latin and in ogham and gives him the late Roman title of protector. He is often identified with Vortipor, the king of the Demetae whom Gildas attacks by name, although the spellings differ and some scholars doubt it.

One stone, two alphabets, and perhaps a king who appears in Gildas: that is about as close as AD 500 comes to a named person on the ground.

The traffic went north too. Dál Riata, in Argyll, was a Gaelic kingdom that spanned the sea to Antrim, and its chief seat is thought to have been the crag of Dunadd, with its carved footprint for king-making. In Brycheiniog, the Brecon kingdom, tradition has an Irish-descended founder called Brychan, and Wikipedia's article says it was probably founded by Irish raiders in the late 5th century.

05 · Names without bordersThe kingdoms

Here are the 24 kingdoms on the map: 14 British, 8 Anglo-Saxon, one Gaelic and one Pictish. For each I give the earliest written mention as Wikipedia's article describes it, and what evidence lies in the zone I drew. Some zones are full of evidence and some are empty. An empty zone does not mean an empty kingdom: Elmet and Ergyng, for example, are known from texts and place-names, and their people left little that my layers can see.

Look at the right-hand column and a pattern appears: for the southern and eastern Anglo-Saxon kingdoms, almost all the culturally distinctive evidence in the zone is Anglo-Saxon, and for Wales almost all of it is British. In the north it is thinner. Gododdin's zone, around Edinburgh and Traprain Law, catches Pictish stones from across the Forth and hillforts with no cultural label, and nothing that says Gododdin. That is what most of this period looks like when you try to pin names to the ground.

06 · What Rome leftTowns, forts and the new settlers

Under the early medieval layers sits the Roman province. Pleiades, the gazetteer of the ancient world, places 182 Roman towns and settlements and 215 forts in my frame with good locations. It tags 81% of the towns and 41% of the forts with its Late Antique period (AD 300 to 640). Most of that reflects the 4th century, not proof that anyone lived in them in 500. A few have stronger stories: Wikipedia's article on Wroxeter says the Roman town may have served as an early post-Roman capital of Powys.

The Anglo-Saxon sites in my data sit close to Roman towns. Their median distance to the nearest Roman town is 9.3 km, against 13.0 km for random points in the same part of England, and 52% lie within 10 km of one, against 35% of the random points. It is a modest difference, and it has an obvious dull explanation as well as an exciting one: Roman towns and early Anglo-Saxon farms both favoured the same good river-valley land.

07 · Dark skiesThe 530s and 540s

Two disasters hit the middle of the century. In 536 a volcanic eruption filled the northern sky with a veil of sulphate haze and summers turned cold for years. Then the plague of Justinian spread from Egypt in 541 around the Mediterranean, and burials at Edix Hill near Cambridge show that it reached Britain. Neither can be seen in my layers, because the dating of most of the evidence is far too broad. The chart below sets the few firm dates against the evidence that is "in play" at each date.

08 · Why so thin?Writing the 5th century

Almost everything we know by name about Britain in 500 comes from texts written long afterwards. The chart shows the gap. Gildas, a British monk, is the only writer close to the time, and even his book is dated anywhere from the late 5th century to the 540s. Bede finished his history in 731. The Historia Brittonum, which gives us Arthur's battles and the British kings of the north, is of the early 9th century. The Anglo-Saxon Chronicle survives in manuscripts compiled from the 9th to the 12th centuries. The poem Y Gododdin survives in one manuscript of the later 13th century.

That is why every kingdom on my map is drawn in italics with a soft edge. The names are real: they mattered to the people who wrote them down. But they reach us through later politics, genealogies built to flatter later kings, and stories that grew in the telling. The ground, by contrast, is mute but contemporary.