Lost Kingdoms on the High Street: Early Medieval Kingdom Names in Modern British Business Names

The kingdoms of early medieval Britain disappeared more than a thousand years ago, but their names are on shop fronts, pubs, police stations and street signs. Do these names still sit where the kingdoms were? I read every UK feature in OpenStreetMap named after one of twelve early medieval kingdoms, from Wessex and Mercia to Elmet and Rheged, classified 1,087 named features as businesses and voluntary bodies, public bodies, streets or other objects, and compared their locations with approximate extents of the kingdoms around AD 800, drawn as unions of modern areas and fixed, with the analysis plan, before any feature was located. Against a baseline of census residents, kingdom names on businesses, clubs and societies are concentrated inside their kingdoms: 75 of 95 Wessex names lie inside an extent holding 10.0% of UK residents (rate ratio per resident 34, 95% CI 20 to 71), 30 of 32 Mercia names lie inside Mercia and all 33 Northumbria names inside Northumbria. Pooled across the three names with enough cases, the rate ratio is 44 (CI 19 to 100), and it stays above 14 in all 19 planned robustness checks; a placebo with the wrong kingdoms' extents shows no concentration (0.97, CI 0.51 to 1.83). Public bodies are not detectably more local than businesses, but streets travel further: only 52% of kingdom-named streets are inside the kingdom. Most names that leave travel only a few kilometres. The data are observational, cover only mapped features and cannot date the names.

Interactive version: https://maps.tristanhotham.com/lost-kingdoms/. Data and code: https://github.com/Tristan-Hotham/lost-kingdoms. Blog post: https://blog.tristanhotham.com/2026/10/31/lost-kingdoms/.

Sources: OpenStreetMap contributors via QLever (ODbL 1.0); Census 2021 (ONS, NISRA) and Scotland's Census 2022 (NRS), Open Government Licence v3.0; ONS boundaries (OGL v3.0). Text and figures CC BY 4.0; code MIT; data derived from OpenStreetMap ODbL 1.0.