by Tim Pestell
The idea of the ‘productive’ site has not yet made a great impact on the terminology of sites in the west of Britain, but there is no doubt that in its loosest sense it could be applied to sites with a large quantity of imported material. Tintagel, for example, is extremely ‘productive’ in terms of post-Roman pottery of Mediterranean origins. Imported material from less dominant sites in the South-West such as Gwithian and Bantham echoes Tintagel’s external contacts. The work of Charles Thomas and Ewan Campbell in mapping the spread of African and Phocaean Red Slip tablewares (‘ARSW/PRSW’, also known as A ware), East Mediterranean amphorae (B wares), and western Gaulish D and E wares, together with Continental glass beakers, shows that the mid-first millennium ad in the West was characterised by patterns of exchange which still reached well beyond north-west Europe (Thomas 1990; E Campbell 1996). Major secular defended settlements such as Dinas Powys in south-east Wales and Dunadd in Dál Riata, western Scotland, show evidence of indigenous production of high-status portable objects such as penannular brooches (e.g. Lane and Campbell 2000). As the present author argued in an article published in 1994 by the Gudme Project in Denmark (Griffiths 1994), it seems that the imports – the fine tablewares, wine, honey and colourful dyes represented by the ceramic evidence and glass – and the locally-produced metalwork, were part of a system of redistribution within tribal or kinship territories based on tribute and reciprocal obligations. Feasting and inauguration ceremonies would have provided a focus for consumption. The Mediterranean contacts and increasing wealth of the early Christian Church certainly played a role in the pattern of exchange, and we can point to good evidence for imports and production in the vicinity of monasteries, Whithorn being an obvious example (Hill 1997).
One group of sites which forms part of the exchange network in the West, but for which there is almost no historical information, are those loosely termed ‘beach markets’. At a number of marginal coastal locations around the Irish Sea, mostly characterised by sand dune landscapes, there have been finds of metalwork comparable in type if not in quantity to the ‘productive’ site assemblages known further to the east and south. At Luce Sands, Galloway, south-west Scotland (Cormack 1965), stycas and bronze Anglo-Saxon strap-ends (among a range of other earlier and later material) have been found in eroding sand formations on Torrs Warren behind the beach. Stevenston Sands, Ayrshire (Callander 1933) is a comparable site, if smaller in the range of objects. A similar situation where Early Medieval metalwork has been found unstratified together with a range of prehistoric items is documented at Murlough Dunes, Dundrum Bay, Co. Down in Northern Ireland (Jope 1966).
These are not recent finds, but careful re-examination of the assemblages and their context can be surprisingly informative. The eighth-and ninth-century metalwork finds, if abstracted from the wider assemblages of material from these sites, begin to look like a sparser version of the classic ‘productive’ site phenomenon. (The presence in the West of contemporary Anglo-Saxon metalwork is also attested by hoards, including the Trewhiddle Hoard from Cornwall (see above) and the Talnotrie Hoard from Kirkcudbrightshire, southwest Scotland.) However, where the sand-dune sites are concerned, the presence of earlier metalwork, including sub-Roman and postRoman penannular brooches, together with Iron Age and also later Medieval metalwork, confirms that the material dating to ad 700–1000 must be seen as part of a long-term pattern of importation of traded objects at these sites. Elsewhere in the ‘Celtic West’, the presence of coastal redistribution centres is indicated by varied ceramic assemblages at highly marginal sites, of which Dalkey Island off the Dublin coast (Liversage 1967) is a good example, although Longbury Bank in south-west Wales (Campbell and Lane 1993) should also be mentioned in this regard. This author’s interpretation for the coastal market sites as a group is that they represent small non-elite and largely undefended settlements which occasionally functioned within a very long-lived tradition of seasonal markets at these specific locations. Indeed, a seasonal function has even been suggested for Tintagel by Charles Thomas (Thomas 1988). These seasonal events, possibly taking place on set days in the calendar, would have drawn trade largely from within the region, but the presence even in small numbers of merchants from much further afield would have been an undoubted attraction for the elites and their representatives. If one is permitted to infer back from more recently documented traditions, the annual ‘wakes’ of north-west England would be a good example.
From the late eighth century onwards, the Vikings clearly played a role in circulating material culture, mostly but not exclusively of insular origin. In all but the earliest phase of the Viking presence, it would be simplistic to view raiding activity as the predominant theme in this context; many apparent market sites in the West which show evidence of sixth-to eighth-century material also show Viking influence. There was a tradition in contemporary Scandinavia of non-urban markets in boundary zones, where a number of metalwork-producing sites have been recorded (e.g. Larsen 1985).
One significant coastal ‘beach market’ site is Meols (pronounced ‘Mells’, from Old Norse Melr, meaning ‘sand dune’). Situated on the north coast of the Wirral peninsula facing Liverpool Bay and mid-way between the river estuary mouths of the Dee and Mersey, this site is currently the subject of a full site history and catalogue being written by the author and Dr Robert Philpott of Liverpool Museum (Griffiths and Philpott in prep.). Over 3,000 objects from the site exist in five museum collections, dating from the Mesolithic to the post-Medieval periods, but it is from the Middle Iron Age to the fifteenth century ad that Meols seems to have acted as a regional market. Most of the finds were collected by antiquarians in the period 1846–1900 as coastal erosion gradually exposed ancient settlement. The coastal retreat was stopped in the late 1890s by the construction of sea defences. Led by the Rev. Abraham Hume (Hume 1863), a handful of antiquarians searched for artefacts and made some interesting, if poorly-located, observations of burials, wattle round houses succeeded in upper layers by rectangular long houses, trackways, walls and fences. The surviving objects represent, of course, an unknown fraction of the original number, but leave no doubt as to the regional importance of the site. The antiquarian story is fascinating in itself and we have to remain constantly aware of the need to be source-critical, but for the most part the collections are coherent, with regional parallels, and their provenance at Meols seems secure. There has been a trickle of finds since 1900, mostly chance discoveries, but in recent years metal-detector users have begun to contribute some important pieces. All of these subsequent finds confirm themes already inherent in the antiquarian collections, and in some cases have provided identical objects.
The similarities and correspondences at Meols to our familiar idea of the ‘productive’ site concern the eighth to tenth century material (for discussions of some of this material see Bu’Lock 1960 and Dolley 1961). The coins, although not large in number, are a case in point. Two Frisian Primary sceattas, still unusual on the west coast (an area once described by Michael Metcalf as ‘the back of beyond’ in monetary terms: 1987, 365) were found here, together with a handful of ninthcentury Northumbrian stycas. Stycas are slightly more common in the North-West than sceattas, with hoards and single finds known from Otterspool (Liverpool), Lancaster, Grange, Kirkoswald and Carlisle (Pirie 1986b). Other characteristic Middle and Late Anglo-Saxon objects at Meols include a bossed silver disc pin, a group of bronze hooked tags, a probable ‘Winchester style’ openwork strap-end, and an iron disc pin which has been referred to elsewhere as a stylus, but which is now not regarded as such (T. Pestell, pers. comm.). There are also a number of biconical-headed and faceted bronze pins (Fig. 6.1). These are accompanied by Irish Sea regionally-derived material including an interlaced carved mount, a bronze hexagonal pyramidal bell and the largest single group of Hiberno-Viking ringed pins outside Ireland.
FIGURE 6.1. Some eighth-to eleventh-century finds from Meols, scale 1 :1 (drawn by Mark Faulkner).
There are close to thirty tenth-and
eleventh-century Anglo-Saxon silver coins from Meols (Dolley 1961), although the whole pre-Norman coin series looks small compared to the Roman and Medieval coins which number several hundred. This reinforces the point that the Early Medieval objects are an important but still fairly small part of a very long-lived assemblage which includes, apart from imported material, much of the standard equipment of everyday life in the form of iron tools, leather, Roman and Medieval pottery and even some manufacturing evidence in the form of crucibles and metal wasters. Meols may have acted as a regional entrepot for trade in Cheshire salt or lead mined in the Flintshire Hills across the Dee estuary (Griffiths 1996).
Further evidence for possible sites linked to Meols within the surrounding river systems comes from a small group of ninth-century Anglo-Saxon metalwork objects, including a bronze strap-end with an enamelled zoomorphic design and a Trewhiddle-style circular silver disc pin head, found by metal-detectorists at Hale, on the north bank of the Mersey (Philpott 1999), although their context has not been investigated archaeologically. At Altmouth, on the Lancashire coast across the mouth of the Mersey opposite Meols, some Roman and eleventh-century Norman coins have been discovered, but as yet no Anglo-Saxon material.
One of the aims of the current Meols publication project is to integrate more recent and reinterpreted finds. Several Iron Age dress pins have been rescued from century-old labelling as Medieval, and a curious glass ‘necklace’, presumably assembled by a Victorian museum assistant, has been disaggregated to reveal a collection of Roman, Anglo-Saxon and Medieval beads. One of the more fascinating twentieth-century discoveries at Meols, by a local man digging for lugworms on the foreshore in the 1950s, was a small pilgrim’s pottery flask or ampulla from the shrine of St Menas, Lower Egypt, dated to the sixth or early seventh century.
Recent metal-detecting activity around Meols has added some significant pieces to the existing assemblage. An ‘Aucissa’-type Roman brooch from the first century AD, found on the foreshore, is identical to one in the antiquarian collections. Further recent finds have cast some exciting new light on long distance contacts. Indications of Mediterranean links in the Iron Age come in the form of three Carthaginian drachmae recorded among the nineteenth-century finds (three other Iron Age coins, two of Armorican and one of British origin were also found). These contribute to a small but important stock of evidence for Iron Age trade in the region (Matthews 1999). The St Menas flask is an indication that Mediterranean links via the Atlantic seaboard of western Europe (cf. Cunliffe 2001, 421) may have survived throughout the Roman period, obscured perhaps by the difficulty of detecting these distant links within the more general Roman influence on Britain between AD 43 and 400, or alternatively were resurrected in post-Roman times. From the hinterland of the site have come three sixth to early seventh-century Byzantine coins, including a follis of Justin I (518–27) a decanummium of Justinian I (527–65) and a follis of Maurice Tiberius (582–602). The coin of Justinian was a chance find, the other two were metal-detected (Philpott 1998). These add to the small but increasing stock of evidence for Byzantine contact in the Irish Sea region, including the probable inscription (c. 540) in tempore iustini consiliis at Penmachno in North Wales (Nash-Williams 1950, 93) and the recent discovery of an intaglio together with an insular penannular brooch at Cefn Cwmwd, Anglesey, during a road-widening scheme (White et al., in prep.). Traded objects imported directly via early Byzantine contacts in western Britain (for a discussion, see Fulford 1989) have often been dismissed as an over-interpretation of a small, if exotic, group of material. Other explanations, such as modern loss, have been sought for the presence of this material (e.g. Boon 1991). However, as further pieces come to light, we are surely right to begin cautiously to appreciate its significance within the patterns of exchange in western Britain (cf. Dark 2000, 162–3) and to see its role in fostering and continuing markets, and thereby the market tradition, which in some cases attracted traders bringing the type of material more commonly associated with ‘productive’ sites. To stretch this argument still further, it may even be that some of the Byzantine high-status metalwork found in Anglo-Saxon contexts may have found its way there via these western routes.
It is perhaps no accident that the Isle of Anglesey (Ynys Môn), forming the north-western part of Wales and at the centre of the Irish Sea region, has been the scene of the discovery of relevant material. A historic centre of power within the Welsh kingdom of Gwynedd, its relatively fertile and low-lying pastures have contributed to the island having an impressive range of archaeological sites of most periods, which form a significant concentration in western Britain. Until the 1990s the period 600–1000 was not well represented (Redknap 2000). However, there has been considerable metal-detecting activity in Anglesey over the past decade, to which the National Museums and Galleries of Wales has been quick to react. In an example of beneficial co-operation between metal-detectorists and museum archaeologists, one result has been the discovery of a major new site of the Early Medieval period, unknown before 1989. In the early 1990s, from Glyn, Llanbedrgoch, just inland from the east coast of the island near Benllech, metal-detector users reported eighth-and ninth-century coins and metalwork to Dr Mark Redknap of the National Museum. Dr Redknap initiated a major archaeological investigation which continued annually until 2001 (supervised in part by the present author in 1997–9).
One of the first indications that the metal-detectorists had stumbled across an extraordinary site was the discovery of an Anglo-Saxon penny of Cynethryth, wife of Offa (757–96) and a number of lead merchants’ weights. There was also the discovery by this method of several ninth-century coins, namely a penny of Wulfred of Canterbury, two deniers of Louis the Pious, one of Charles the Bald and one of Pepin II of Aquitaine. The subsequent discovery of a small collection of pins and brooches pointed to a Viking or immediately pre-Viking context. Also found were some classically Irish Sea Viking material such hacksilver and a bronze hexagonal bell, which emphasise the trading links of the site and which are close parallels for finds from Peel Castle on the Isle of Man and Meols.
At one of a small number of locations investigated archaeologically (with the active on-site assistance of the original metal-detectorists who discovered the material), geophysical survey followed by excavation began to reveal a defensive ditch forming an enclosure around a small spring. Subsequent seasons have revealed that this enclosure, fortified in places by a massive wall, was intensively occupied during the ninth and tenth centuries, with at least three (and possibly up to six) timber buildings of long-house type, extra-mural inhumations, extensive middens and metalworking hearths (Fig. 6.2). These latter two contexts have produced two Northumbrian stycas, and elsewhere the excavation has produced tenth-century Anglo-Saxon silver pennies and an Arabic dirhem. When begun, this excavation was designed principally to investigate the context of the Viking-period metalwork, but gradually more and more indications of pre-Viking activity have been discovered. It is now clear that the spring at the centre of the site has been the location of activity since the Neolithic period, and there is a sequence of artefacts (including a polychrome bead and a seventh-century bird-headed brooch) supplemented by radiocarbon dates, and a possible round house stratigraphically below one of the long houses, indicating settlement in the pre-Viking period. A nearby scatter of Roman pottery, some Late Roman coins in the form of a clipped siliqua and a barbarous radiate, together with the proximity on the same farm of the known sixth-to eighth-century settlement site of Pant y Saer (Edwards and Lane 1988, 99–101), go far towards completing a long-term context for the eighth-to tenth-century imported metalwork. The site was almost certainly a native Welsh site which experienced significant influence, and probably settlement, by Irish Sea Vikings from the later ninth century.
FIGURE 6.2. Llanbedrgoch site plan (drawn by Tony Daly © National Museums and Galleries of Wales). Key:
(1) Definite buildings;
(2) possible buildings;
(3) ditch;
(4) wall;
(5) areas of excavation up to 1999;
(6) bowl hearths (ninth century);
(7) metalled/paved surface;
(8) burials.
Despite the Anglesey site not being in an immediately coastal situation (although it is only one kilometre from a sheltered coastal inlet), the artefactual parallels between Llanbedrgoch and Meols are striking. Both seem to have been the setting for more-or-less permanent settlement in the Early Medieval period, but the trading contacts evident in the metalwork assemblages in particular point to an elevated role as market sites, or a very close connection to a market location. Given the chronological range emerging from studies of these sites, it may well be that we are detecting here long-term traditions of seasonal markets at territorial and topographical boundary zones. These may relate, possibly in the case of Meols at least, to a basic or ‘prime’ commodity which was available here along with metalwork for personal ornament, salt and lead being the most likely materials.
There is no evidence that there is any difference in the strength and quality of archaeological response to the discovery of ‘productive’ site type material in the west as opposed to the east of Britain. Indeed the Portable Antiquities Scheme, for example, presently covers at least as much of the west as the east of England and Wales, with only Devon, Cornwall and the English southern border counties in the Welsh marches exempted in the West. Therefore, we have to accept that there are simply far fewer obvious candidates for designation as ‘productive’ sites, and the general circulation of seventh-to tenth-century Anglo-Saxon metalwork is markedly lower. However, as illustrated above, there are a number of market-influenced sites in western Britain which conform to some of the general characteristics of ‘productive’ sites, whether in their method of discovery or artefact assemblage. A sufficiently broad and flexible attitude to the concept (perhaps seeing such sites simply as reflecting non-urban market activity in liminal zones) may well bring the western sites within the remit of the debate. There is a clear difference in scale of the characteristic metalwork assemblages between West and East, which must in part be due to a lesser or more diffuse Anglo-Saxon cultural influence in the West (cf. Hines 2000). Braudel’s longue durée is a useful interpretative backdrop to patterns of trade and external contact in those parts of western Britain bordering the Irish Sea region, which Mackinder once referred to as the ‘British Mediterranean’ (Mackinder 1902, 20). Consideration of artefact patterning across several millennia from the Iron Age through to the Late Medieval period is a key interpretative tool here, and might very well stand as a conclusion relevant to other areas.
