Subscribe now for full access and no adverts
It is often said that 1066 is one of the most famous dates in English history, when Harold Godwinson’s death at the Battle of Hastings heralded the watershed events of the Norman Conquest, and the long-established Anglo-Saxon aristocracy was replaced by a new, foreign elite. Yet, while we know where Harold met his end (see CA 286 and 318), and tradition links his grave with Waltham Abbey in Essex (CA 125), sites connected to his life, and to the influential social class from which England’s last Anglo-Saxon king had sprung, have long been less well-understood. Now, however, recently published archaeological research has shed vivid light on early medieval lordly sites, including a key power centre associated with Harold and his family.
The coastal village of Bosham sits on a peninsula within the watery landscape of Chichester Harbour in West Sussex. It is a near neighbour of Fishbourne Roman Palace, lying around 2 miles from this once-lavish residence – but Bosham can boast of its own historical claim to elite status, thanks to its long-standing association with Earl Godwin, a powerful early medieval magnate, and his sons – among them, the future Harold II. The Anglo-Saxon Chronicle includes multiple entries mentioning Godwin and his eldest son, Sweyn, keeping their ships at Bosham during the mid-11th century, while Harold is depicted visiting the site twice on the Bayeux Tapestry. In the first of these scenes, Harold is shown riding to Bosham (which is named in a sewn surtitle), feasting in a hall, and walking towards water, in order to sail to Normandy in 1064. Later, we see Harold’s return, riding towards a grand building represented by two towers.
While such sources suggest that Bosham was a key seat for the man who would be king, there had previously been little archaeological investigation focused on his residence – and, indeed, its precise position had long been lost to history, although tradition linked it to a spot now occupied by a rather more-recent house, close to the medieval church which has itself been occasionally proposed as an alternative for the king’s final resting place. Now, a two-year project examining evidence for medieval ‘lordly centres’ across England has confirmed the site of Harold’s hall ‘beyond all reasonable doubt’ – and the crucial clue came from a surprisingly humble structure.
Recently published in The Antiquaries Journal (see ‘Further reading’ box below), the work was led by Dr Duncan Wright, Senior Lecturer in Medieval Archaeology at Newcastle University, and Professor Oliver Creighton, Co-Investigator in the University of Exeter’s Department of Archaeology and History. At Bosham, their research combined analysis of standing building remains; geophysical surveys at the house and elsewhere in Bosham; a study of local historical mapping and of metal-detected finds from the parish recorded by the Portable Antiquities Scheme; and re-examining excavated evidence dating back almost 20 years.
Digging deep
This was not the first time that archaeologists had visited the house at Bosham. In 2006, George Anelay led investigations there for West Sussex Archaeology, opening 17 trenches within the property’s grounds and revealing portions of two enclosure ditches that disappeared beneath the western end of the present house. Their fills yielded a mix of pre-Conquest, Anglo-Norman, and later medieval pottery, suggesting the ditches went out of use by the 14th or 15th century. Unfortunately, the creation of a garden pond had removed most of the archaeology from the enclosure’s interior, but the footprint of one small structure had survived: an earthfast timber building incorporating a latrine indicated by the presence of a cess pit filled with fragments of late Saxon and Anglo-Norman pottery.
Back in 2006, the significance of a latrine associated with a large building could not be fully appreciated, but over the last 15 years research into similar structures (by Gabor Thomas in 2010, and John Blair in 2015) has revealed an important trend, beginning in the 10th century, whereby toilets became integrated into high-status houses – what John Blair called the emergence of ‘the Anglo-Saxon en-suite’. Duncan and Oliver argue that the Bosham latrine would have also been attached to the chamber-end of a late Saxon long-range, and that it provides unmistakable evidence of an elite-level residence pre-dating the Norman Conquest: surely Harold Godwinson’s Bosham powerbase.
A hall of this status would have sat within a larger enclosure, perhaps represented in part by the two boundary ditches excavated nearby, with a further section of its outline possibly provided by a nearby stream which today survives as a canalised mill leat. Watercourses (as we will discuss below) appear to have been an important feature of Anglo-Saxon lordly sites, and if the leat does preserve the western extent of the Bosham enclosure, we might extrapolate a rectangular compound measuring c.50m east–west and c.70m north–south.
Appended to the southern end of this complex was Holy Trinity Church, a building located within its own enclosure and linked to the elite residence via a wooden bridge. The earliest fabric of the church pre-dates the Norman Conquest, although Bosham’s religious use dates back much further. In his 8th-century Ecclesiastical History, Bede writes that, when St Wilfred travelled to Sussex to convert the South Saxons to Christianity, he found that ‘Bosanhamm’ was already home to a small Celtic monastery. This foundation may lie beneath, or close to, the present church, and its later incorporation into a lordly estate would not be surprising: from the 9th century onwards it seems to have been increasingly common for the assets of ecclesiastical communities to be divided and partly taken over by local aristocracy, a process known as secularisation.
Clues to the site’s later medieval life were revealed by the recent research, too, indicating that the timber hall shown on the Bayeux Tapestry had evolved into something more enduring. As the team examined the make-up of the modern house they found that, while it is mainly 17th-century and later in date, at its core is part of a wall from a much earlier structure that, based on its stonework, had been built c.1100-1300. Reused timbers – including soot-blackened rafters and curved windbraces speaking of an open hall design – from a second, even later medieval building, were also observed throughout the house.
Further insights emerged from the enigmatic ‘Garden Ruin’ that stands 35m west of the house. Previously interpreted as a partly demolished post-medieval barn, its surviving stonework represents the north and west walls of a rectangular building, made from coursed flint and standing to a maximum height of 2.4m. The ruin is heavily overgrown, but scrutiny by the team suggests that the footings of its west wall, being entirely of stone rather than brick, are probably medieval, and represent the earliest element of the structure. Strikingly, the stonework above these also incorporates a distinctive lancet window; it is therefore suggested that the Garden Ruin might preserve part of a 12th- or 13th-century building, constructed over even earlier footings. It is plausible that the remains of a chimney, recovered during the 2006 excavation and dated on its form by Professor David Stocker to the 12th century, also derives from this building.
Where power lies
The early medieval remains identified at Bosham offer invaluable insights into an important cultural change that took place in mid-9th-century to mid-11th-century England. This period saw the emergence of a new, secular elite who signalled their social status, wealth, and authority through the construction of high-status residences, often paired with a church or chapel. These were not just ostentatious homes, but centres of local administration, from which economies, lands, and populations were controlled. In the aftermath of the Norman Conquest, however, some of these sites and the Anglo-Saxon aristocracy who inhabited them were deliberately swept away. While a small number have been explored archaeologically, among them Raunds in Northamptonshire (CA 106) and Goltho in Lincolnshire (CA 56), historically Anglo-Saxon ‘lordly centres’ were not well understood as a nationwide phenomenon, and were often overshadowed in scholarship by the castles and manor houses that were built after (and, in many cases, over) them.
It was partly in order to redress this imbalance that Where Power Lies, a two-year project funded by the AHRC, set out to explore the origins and evolution of aristocratic centres between AD 800 and 1200. It represents the first systematic survey of surviving physical evidence for these sites on a national level, and as a first priority the team focused on incorporating all available national datasets relating to potential lordly centres (drawing on Historic Environment Records, Historic England’s Listed Buildings Register, and other research records) into a single Geographic Information System.
Having pulled together as much data as they could, the researchers then focused their efforts on two macro regions, in order to examine regional trends and compare areas inside and outside the Danelaw. Differing in the number of counties that they covered, but encompassing a similar area, these two groups comprised Devon, Dorset, Gloucestershire, Hampshire, Somerset, West Sussex, and Wiltshire; and Lincolnshire and Yorkshire respectively. Within these, the team were able to identify 870 ‘lordly’ residences and 3,528 churches with origins before AD 1200.
Examining records for these sites in more detail has revealed some interesting trends. Just over a third (34%) lay within 250m of a church, although this pattern was more pronounced in some areas than others, being clearest in Dorset (40%), Gloucestershire, Lincolnshire, and Somerset (all 42%), and West Sussex (48%), and less so in Devon (20%) and Yorkshire (29%). Proximity to water mills also seems to have been a priority for such complexes, while watercourses were themselves noteworthy features. Some 37% of the lordly residences in the two study regions were located within 250m of a stream, river, or mill leat, with many incorporated into their boundaries – as has been suggested at Bosham.
On the ground
Following their deep dive into existing records, the project team selected seven sites (including Bosham) to explore in greater detail through fieldwork; an eighth investigation at Earls Barton, Northamptonshire, is currently under way and will feature in an update in a future issue of CA. While, as we will explore below, these sites had many features in common, they also highlighted the diversity of ways in which Late Saxon lordly status could be expressed.
At Broadclyst in Devon, geophysical and landscape surveys identified a large (2.2ha) enclosure whose outline was partly preserved in curving field boundaries and the course of the River Clyst, which forms its western limit. At its centre, the team found evidence of a large, level platform overlooking the river, adjacent to the parish church, which they suggest may have been the site of a 10th- or 11th-century hall.
Another large enclosure was recorded at Little Ouseburn in North Yorkshire, where an ovoid shape can be seen in satellite and LiDAR imagery, and was also picked up by the team’s geophysical surveys. Again, it encompasses the parish church, which was built as a free-standing tower-nave – a form of structure associated with early medieval aristocratic activity – in the 10th or early 11th century. The researchers suggest that this structure may have occupied the eastern part of a lordly compound (possibly taking over an earlier minster site whose presence is indicated by later written sources and pieces of pre-Viking sculpted stone), with an elite residence in the western part of the enclosure.
A similar arrangement, and another tower-nave, was identified at Hornby, also in North Yorkshire. There, the team’s analysis of the fabric of St Mary’s Church indicated that its tower was structurally independent of the main building, possibly originating in the 11th century. To the west are a series of well-preserved earthworks forming a C-shaped enclosure with a clear gap aligned on the church. Following a photogrammetric drone survey, these were interpreted as the western edge of an oval measuring some 200m by 150m, with three possible building platforms inside: again, suggestive of an aristocratic complex with a church to the east and a residential focus to the west.
The other three sites also had churches in close proximity, though these pairings were more varied in form. At Healaugh in North Yorkshire, St John the Baptist’s Church was accompanied by the remains of an 11th- or 12th-century ringwork; while a more unusual construction was examined at Great Somerford in Wiltshire. There, a 3m-3.5m-high earthwork known as ‘The Mound’ stands 20m west of the parish church. Excavations in 1811 and 1910 recorded a stone building within it (something confirmed through GPR by the more recent research) but what was it for? The structure’s north-south orientation is wrong for a church, and its barely 2ft-thick walls belie a defensive function; might this instead have been a free-standing aristocratic tower, located close to the church, with its position later fossilised within the tall earthwork?
Perhaps the most complex story comes from Saintbury in Gloucestershire. A series of well-preserved earthworks lie just southwest of St Nicholas’ Church, in an area known from historical documents to feature an earlier minster and a manor that was later granted to Evesham Abbey. Magnetometry, GPR, and landscape survey confirmed the presence of a large, irregularly oval enclosure adjacent to the church, with a substantial building platform inside. However, OSL dating suggests that the earthworks originated in the Iron Age, were reused in the Roman period, and their embankment saw a final modification in the 9th century. If this site was associated with an early medieval high-status residence, it formed just part of a longer story.
Through exploring sites like these, Where Power Lies has added details and distinctiveness to a neglected class of site that, where it was characterised at all, has historically been stereotyped as being rather homogenous in design. While the project has highlighted a number of frequently occurring features – earthworks, enclosures, churches, sundials, mills, and watercourses – to help identify potential lordly sites in the future, it has also shown that these complexes were much more diverse than once thought, no doubt as individual as the Anglo-Saxon elites who built them.
Further reading:
D Gould, O Creighton, S Chaussée, M Shapland, and D W Wright (2024) ‘Where power lies: lordly power centres in the English landscape c.800-1200’, The Antiquaries Journal 104 (October); https://doi.org/10.1017/S0003581524000350.
