It has long been thought that the first castle ever built in England was in existence before September 1051, when the Anglo-Saxon Chronicle records that 'the Frenchmen had built aene castel in Herefordshire'. Yet until now we have been no closer to unravelling the many mysteries surrounding this historic fortress. Its physical structure long lost to us, we can only try to reconstruct it and the circumstances of its creation from obscure 11th-century sources. But these are still sufficient to enable us to discern a good deal about a building which occupies a unique place in England's history.
Several annals written by Anglo-Saxon monastic chroniclers in 1051-52 make clear that the castle was built in the territory of the Earl of Hereford by a previously unknown Norman named Osbern, who occupied it for about a year. The Domesday Book showed him holding two manors in the midlands county on the English side of the border with Wales. Who Osbern was, however, why he built this castle, where exactly it was built and what form its construction took are questions that the chroniclers do not answer.
Another mystery is how a Norman castle came to be built in England 15 years before the Norman Conquest of 1066. And, since we are saying that this was the first castle to be built here, despite the existence of sites of much greater antiquity which take the name 'castle'--such as Cadbury Hill Fort in Somerset --we also need to look at the whole question of what constitutes a 'castle'.
By the 19th century, historians had agreed that a castle was a defensive building with walls and a roof, which ruled out hill forts with earthwork defences that were essentially natural features developed for defence. But there was no agreement on who first built castles, with some scholars claiming that honour for the Anglo-Saxons. The question was largely decided by Ella S. Armitage, whose book The Early Norman Castles of the British Isles, first published in 1912, remains a standard work. Armitage made function or use of a building the key factor. Anglo-Saxons, she pointed out, built communal defences such as their many defensive towns, or burhs, fortified against Viking raids, especially in the central kingdom of Mercia. Norman castles were purely private defensive sites protecting the chosen few. English castles, she insisted, were therefore a Norman innovation, a view broadly accepted ever since.
This explains why Osbern's castle was unique, as the first recorded castle built by a Norman in England, but it does not explain what he was doing in England 15 years before the Rattle of Hastings.
That he was in England before the Conquest is actually not as unusual as it might seem, since the then king, Edward the Confessor (r. 1042-66), was half-Norman and had spent most of his life in the Duchy of Normandy before returning to England in 1041 with his part-Norman nephew Ralph and trusted Norman cleric Robert of Jumieges. The Victorian view that Edward brought a large contingent of Normans with him to England was disproved by 20th-century historians but by summer 1051 Robert of Jumieges was Archbishop of Canterbury and within a year there were likely to have been three Norman castles in England, all in the border county of Herefordshire. …