Subscribe now for full access and no adverts
Our story begins on the north-eastern coast of Anglesey, in the parish of Llanbedrgoch, where a farm called Y Glyn lies about 1km (0.6 miles) from the sea. Its Grade II*-listed farmhouse has attracted much interest from antiquaries because of its richly decorated interior, which boasts exceptional painted plaster relief friezes and panels. Dating from 1613, their imagery features martial emblems (cannon, guns, swords, pikes, and a powder barrel), a coat of arms flanked by Classical figures in loincloths, and a biblical scene. From 1994, however, it was the area outside this historic structure that was the focus of attention, as Mark Redknap and his team began to uncover unusual archaeological features in one of the farm’s fields (see CA 184).
It had long been known that this landscape had witnessed Viking Age activity: between 1887 and 1894, limestone-quarrying at a coastal site about 2km (1 mile) north-east of the farmhouse revealed a hoard of silver arm-rings. Described as Hiberno-Scandinavian in style, and featuring a variety of punched decorative schemes, these artefacts are notable for being complete – unlike most silver of the period, they had not been cut into smaller pieces (known as hacksilver) for use as a form of currency to pay for goods by weight. They post-date the first seaborne Viking raids on Anglesey in the mid-9th century, and the hoard has been associated with the Hiberno-Norse warlord Ingimund and his followers, who were expelled from Dublin in AD 902. According to Welsh and Irish annals, they first made an unsuccessful attempt to establish a base on Anglesey, but subsequently moved on to the Wirral Peninsula. By then, the Wirral already possessed an informal beach market at Meols, and finds from this location point to the presence of traders of Scandinavian origin in the 9th century. Perhaps they chose to draw up their boats and operate from the sands of Meols in preference to the more regulated and taxed trading environment of Chester.
What Mark and his team were to discover at Llanbedrgoch was something comparable to the market at Meols but unique in form. Known as Site 2 (the team excavated four areas in all), it was home to at least two successive enclosures. The first, smaller example was surrounded by a (probably late Roman) ditch. This was filled in by the early 7th century, when a new and much larger ditch was dug, accompanied by an imposing rampart measuring 2-2.2m (6.5-7ft) in width and faced with clay- bonded stone walls. The new defences incorporated the rock-cut parts of the earlier enclosure to the north, but they also extended much further south, encircling some 0.8 hectares (2 acres) of well-drained land, and forming an enclosure three times larger than its predecessor. The ditches secured a natural spring, too, which had served the site’s prehistoric inhabitants and continues to provide a source of fresh water today – it was modified by the construction of a large stone-lined cistern, measuring 4m (13ft) square, with steps on one side to provide access to the water.
Piecing together the past
Large parts of the site had been truncated by later ploughing, so Mark and his team did a remarkable job of recovering what little evidence survived of the buildings that had once stood within the enclosure. They found scattered post-holes – never enough to completely reconstruct a building’s footprint – and stretches of walling, hearths, and floors in the narrow strips that had escaped the plough because they were close to rocky outcrops or had slumped into pits, gullies, and middens.
From such scant remains, they identified two main building styles. An earlier (5th- to 9th-century) phase of rectangular structures was characterised by ground-fast posts: typical of Wales in this period, and relatively short-lived because the structural timbers were prone to rot. Some of the posts showed evidence of occasional repairs or replacement. Meanwhile, the buildings of the later phase (from the 9th century) were rectilinear in design, with sill-wall construction, using either timber sill-beams or stone walling to provide support for the wooden superstructure.
Fragments of fired clay with wattle impressions (and one example of decorative moulding) indicated that some buildings had internal walls, but without dating evidence it was not possible to say whether these belonged to the early medieval settlement or to the underlying Neolithic use of the site. There was no direct evidence for roofing materials, but straw, reed, and rush thatch is likely, along with heather, bracken, gorse, and turf.
It is possible to reconstruct some of the buildings in more detail, however. One structure was identified by the remains of sill walls, as well as the fire-fractured kerbstones and reddened clay of its central hearth, which was surrounded by a narrow trample zone. Probably a residential hall, the building’s interior was divided into two levels. The south-facing end with its gable-entry door was paved with irregularly shaped limestone slabs incorporating rotary quern fragments, while the northern end, closest to the fire, had been dug out to a depth of 0.5m (1.6ft), creating a sunken area whose flooring was provided by the natural limestone bedrock. There were areas adjacent to the hearth that were clear of debris – indicating, perhaps, the location of benching raised above floor level for sitting and sleeping – and the date for the hearth’s last firing provides important evidence for the building’s final occupancy, suggesting that it had ceased to function as a dwelling by about AD 970. Small-scale baking was probably carried out on this and other domestic hearths, but the settlement also had a dedicated bakehouse whose construction alongside the spring, with a large, purpose-built oven, indicates the need to produce bread for a larger number of people as the site’s use began to evolve.
The excavation report (see ‘Further reading’ below) concludes that the earlier enclosure surrounded a late Roman farmstead – not much different in character from the small-scale rural settlements of Anglesey and north-west Wales that we find in every period dating back to prehistory. Its population was not as reliant on the Roman army and state for its economy as contemporaries in southern Wales, and was much less impacted by the end of Roman imperial control.
The expanded settlement of the 7th to 9th centuries, meanwhile, was strikingly different in character. Its size alone – five times larger than some early princely sites in north-west Wales, and comparable to royal-estate centres in other parts of Britain – demonstrates a radical change of use and a more complex economy. This is borne out by the middens full of charcoal, animal bone, and discarded artefacts dating from this period. What had been a small farm appears to have become a ‘functioning estate centre involved in the production, consumption, and possible redistribution of animal resources’. The zooarchaeological evidence shows a mixed economy based on rearing livestock – in particular, cattle – and the trading of surplus animals, raw materials, and manufactured items. Before long, though, the site developed even further, as the coin evidence (mostly from metal-detectorist finds from the plough soil) demonstrates.
Answers in the artefacts
The earliest coin from the site is an 8th-century penny of Cynethryth, wife of Offa of Mercia (r. AD 757-796). Adding to this picture is a small hoard of Carolingian pennies from Cae Howel, a field to the west of Llanbedrgoch, which consists of eight Carolingian coins minted in Melle, western France, alongside a penny of Archbishop Wulfred of Canterbury (AD 805-832). This collection was deposited or lost c.AD 850, a date that coincides with the first recorded Viking raids on Anglesey. Other coin finds from the enclosure include two 9th-century Northumbrian stycas, three fragments of 10th-century Islamic silver dirhams, and two mid-10th-century silver pennies. There was also bullion, and among the hacksilver were fragments of ingots, arm-rings, sheet silver, and a pendant, with one stratified find dated to the mid-9th century. The pieces range in weight from 6.7 grams to 0.3 grams, and some show scratches and nicks – a practice used to check for silver-plated objects that were being passed off as solid silver.
Various lead weights, which would have been vital for establishing the value of the hacksilver and other trading activities, were recovered from the plough soil within the enclosure. They were widely dispersed across the settlement and did not derive from a single set. Two are particularly ornate: one is decorated with a square cross motif that was originally filled with red and yellow enamel, and the other incorporates a gilt copper-alloy brooch terminal, decorated with a chip-carved interlace knot. Collectively, the weights form the largest assemblage of their type from any early medieval site in Wales, with other examples being isolated discoveries like those found at the possible beach markets at Freshwater West, Pembrokeshire, and Brynsiencyn, Anglesey. Similar weights are known from Viking winter camps such as Torksey in Lincolnshire (CA 281), Repton in Derbyshire (CA 100), and Aldwark in North Yorkshire (CA 426), and from settlements at York, Dublin, and Waterford. The coins, weights, and cut-up silver represent accidental losses – the hacksilver is small and lightweight and probably represents a fraction of what was used across the site – but they paint a vivid picture of commercial activity within the settlement, using both coins and bullion, perhaps alongside other items or modes of exchange.
Among the other notable finds were numerous brooches and brooch-pins, dress- and hair-pins, rings and hooked tags (used as garment- or purse-fasteners), buckle plates and complete buckles of metal and bone (some decorated with zoomorphic designs and inlaid enamel), as well as strap ends, glass and stone beads, and cross-shaped pendants (perhaps worn as an expression of religious belief). There were also several small bells (which are usually interpreted as horse harness-trappings, necklace fittings, or hawking equipment), parts of composite combs, and numerous knife blades – as well as whetstones for sharpening them. Eight decorative mounts – one bird-shaped and others with interlace and animal heads – were perhaps intended to decorate bowls or drinking vessels, and there were fragments of metal rims from wooden cups, as well as a shard of glass from an early medieval vessel.
Domestic finds included locks and keys; pottery and lamps; decorative bosses, finials, and suspension rings; as well as structural fittings. Craft tools and mould fragments, as well as pieces of fired-clay and part-forged items, spoke of metalworking taking place within the enclosure. Possible workshop areas were identified by the density of the smithing slag recovered from topsoil and middens. Ironworking was dominant, but other residue indicates that gold, silver, copper alloy, lead, and tin were also being worked. A particularly unusual find was a scrap of lead sheet, of a kind used as a cushion between hammer and die when transferring a raised relief design on to soft gold or silver.
As for other industrial activities, a range of broken and abandoned objects recovered from the middens suggest that all stages in the process of textile production took place on the site, including the preparation of raw fibres, dyeing, spinning, weaving, finishing, cutting, and sewing. Waste materials and semi-finished items show that red- and roe-deer antler was being processed too, and there were also various awls associated with leather-working. Food production was represented by sinkers for fishing nets and many quern-stone fragments, some in the form of partially worked roughouts, showing that they were being produced on site.
Interpreting the evidence
The excavations had yielded abundant evidence for the site serving a dual purpose as a trade hub and as a centre of craft and industry. The form and ornamentation of various items indicates that they were being brought from distant parts of England, Wales, Scotland, Ireland, and Scandinavia, reflecting the extended lines of communication and the distances being travelled by merchants as they journeyed between the ports and landing sites of the Irish Sea. Not all contacts were friendly: the creation of the enlarged enclosure at Llanbedrgoch about AD 625-655 could be seen as a response to the invasion of the kingdom of Gwynedd, of which Anglesey formed part, by Edwin, King of Northumbria, in AD 629. The destruction wrought by his army perhaps stimulated a desire for the greater security afforded by a nucleated settlement with a rampart.
The site’s subsequent growth as an Atlantic seaboard trading centre and manufacturing hub, meanwhile, can be linked to the foundation of Dublin as a longphort, a defended port for Viking longships, in AD 841, and its evolution into an international entrepôt on the River Liffey. Slaves, walrus ivory, furs, and amber, as well as wine, silver, and wool, were transported to Dublin before being distributed along the shipping routes connecting Iceland, Norway, Scotland, Wales, and England. Perhaps Llanbedrgoch was a staging post within this network, located about a day’s sailing from Dublin and two days from the Isle of Man, another Viking stronghold.
Many of the artefacts from Llanbedrgoch can be paralleled at other trading sites including Meols, Chester, York, and Dublin. As with similar trading centres in England and Scandinavia, Llanbedrgoch was located along a sailing route, taking advantage of a natural landing site above Red Wharf Bay (Traeth Coch – literally ‘Red Beach’). This natural sandy bay offered a sheltered haven for boats crossing the Irish Sea to Dublin (129km/80 miles to the west), or up and down the west coast of Britain. It provided easy access to inland Anglesey, too, and a bridgehead to mainland Wales. The topographer Samuel Lewis described this spot in the 1830s as a ‘safe and commodious bay… in which large sloops may ride in perfect security during the severest gales’. Unusually, though, it seems that trading transactions were mainly taking place within the settlement rather than on the beach, which raises the possibility that the site was also being used for toll collection on ships and cargoes. The strategic position of this location allowed its inhabitants to keep a watchful eye on boats approaching Red Wharf Bay and at Benllech Sands, to the north, and perhaps to control mercantile access to inland Anglesey from that point.
Those who controlled Llanbedrgoch – arguably the rulers of the kingdom of Gwynedd, from their royal court in Aberffraw, some 40km (25 miles) away on the south-western coast of Anglesey – enjoyed a significant source of income and access to the long-distance trading networks of the Atlantic seaboard. Inevitably this also sparked conflicts between Gwynedd and the powerful leaders of Dublin and the Isle of Man, which are recorded in Welsh and Irish annals as beginning in AD 854, as each sought to gain mastery over the region’s trade, with Anglesey becoming a prime strategic target for Viking raids.
Llanbedrgoch itself shows no clear evidence of burning or conflict, although the team did find a cluster of five burials – comprising two adult males and three adolescents – just outside the enclosure. They were radiocarbon dated to the 10th century, and isotope analysis indicates that they were not local to the area (CA 276). One of the men had a shallow cut-mark on his skull that might have been inflicted in an act of violence that led to his death. Elsewhere, the Welsh annals provide a record of hostile activities affecting Anglesey that became all the more intense after the defeat of Dublin’s Vikings by Irish forces in AD 902. Forced to abandon their longphort, the Vikings headed for Scotland and the Isle of Man, and – as mentioned at the start of this article – one group, led by Ingimund, headed for Anglesey. The 11th-century Annals of Ireland say that he held a place called Osmeliaun, but after he was driven away by Cadell, son of Rhodri, he sailed on to the Wirral peninsula. Raiding and conflict continued, however, and it is possible, as a result of the disruption to trade, that the site began to decline from c.950.
A ‘non-place’ as a new type-site?
In seeking to sum up the site, Mark Redknap makes the case for its unique character. Its form is currently without close parallels in Wales, though it bears an affinity to a network of related sites located close to sheltered havens at nodal points of connection within the maritime landscape of the Irish Sea. Places serving as outlets for surplus production and an import channel for goods not produced locally have been variously described as trading sites, ports of trade, proto-towns, wic, and emporia, but Llanbedrgoch fits none of these categories. Instead, it shows that many different settlement types developed in Europe between AD 500 and 1000, some with specialised roles as centres of administration or as the base for religious communities, traders, estate owners, aristocrats, and power-brokers. Some of these would grow into towns, but it was Llanbedrgoch’s fate to be abandoned, with the enclosure reverting to agricultural land worked by dispersed farms. The stonework of the ramparts was recycled for use in laying out the later field boundaries, and there appears to have been no memory of the buried site. Two field names, however, were recorded in the 17th century – Cae yr Myrddyn Llwyd (‘Field of the Grey Ruin’) and Cae yr Merddyn Mawr (‘Field of the Large Ruin’) – which suggest that parts of the rampart might still have been visible at that date.
Llanbedrgoch could be defined as a ‘non-place’, in the sense that it had no known or recorded identity prior to its rediscovery and excavation. Might there be others? Until recently, Tintagel was considered to be unique as a high-status coastal site controlling the trade in Mediterranean prestige goods such as wine and olive oil, but now we know of several other similar places, including Dinas Powys in south Wales. Perhaps future discoveries will show that Llanbedrgoch, too, was not one of a kind.
Further reading:
Mark Redknap (2026) From Farm to Viking-Age Trading Centre: the 1994-2012 excavations at Llanbedrgoch, Anglesey (Oxbow Books, ISBN 979-8888571972, £48).
All images: Amgueddfa Cymru, unless otherwise stated
