Divers off the north Norfolk coast have produced the first known images of a medieval settlement that the North Sea took in the 14th century.
Twenty-five manmade features have now been photographed and mapped on the seabed off Cromer. They are believed to belong to Shipden, a town that coastal erosion and repeated storms overwhelmed in the late medieval period, after which the settlement relocated inland to what is now Cromer.
The survey was carried out as part of the Alfred Savin Project's investigation, with volunteer divers from the East Anglian branch of the British Sub-Aqua Club. Peter Stibbons, chairman of the Friends of Cromer Museum, directed the work; Nick Schiller led the diving.
Despite the North Sea's reputation for rough and dangerous conditions, the water was unusually still during the summer dives, which is what allowed the team to photograph and map features on the seabed at all.
What they found was flint, a popular building material in medieval Norfolk, along with other pieces of cut stone and lines of flint foundations or stone-filled caissons standing eight to ten layers high, with the mortar still intact after centuries underwater. The caissons may be 14th-century attempts at sea defence, though that is not confirmed.
The documentary record is older and, in its way, more specific. The Domesday Book of 1086 records at Shipden a community of 117 inhabitants, three acres of meadow, a harbour, two churches and woodland supporting 36 swine. One of the churches was lost to the sea by about 1400, and by the 14th century the name had disappeared from the records altogether.
The survey has also unsettled a local legend. Church Rock, a formation off Cromer, has long been said to be the surviving tower of St Peter's Church, its bell still tolling under the waves. The team now thinks the rock may be medieval sea defence work, and that the name attached itself by legend, possibly helped along by the Victoria, a Great Yarmouth steam paddle tug that struck the site in 1888.
"Those of us brought up in Cromer have always known the story of church rock, so it's a big step to say the legend may be incorrect," Stibbons said. "It is now thought that it wasn't actually a church, but that the name became attached by legend."
Digital reconstructions are still being completed and the diving season has ended. "For the present, we're putting forward ideas rather than conclusions," Stibbons said. The work is due to continue in 2027.
It is worth setting the two documents side by side. The Domesday Book records at Shipden 117 people, a harbour, three acres of meadow, two churches and woodland for 36 swine. The 2026 survey records 25 manmade features, no confirmed church and, across the whole site, a depth of water.
Only one of those documents is a register. The Domesday Book was compiled to establish what was held and by whom, and it has never been amended in respect of Shipden. The survey is a set of photographs, and photographs do not discharge an entry.
Reconciled line by line, the position is as follows. The harbour survives, in the sense that it is entirely water. The three acres of meadow are present but submerged, which is a condition rather than a deletion. Two churches are outstanding. The 36 swine are unaccounted for.
The caissons, if that is what they are, are the only item on the seabed that can be read as a response. They stand eight to ten layers high and their mortar has held for seven hundred years, which makes them the most durable thing anyone ever built at Shipden and, on the available evidence, the least successful.
The Department of Unnecessary Research observes that Shipden was assessed in 1086, has never been reassessed and has never been struck off. On the only register that applies to it, the settlement is still there, two churches included.