Markets in Early Medieval Europe

by Tim Pestell

  At another level in the hierarchy are those rural sites with strong evidence for production, some of which may have had ecclesiastical connections. Brandon (Fig. 5.1:14), at which a church with attendant cemeteries has been unearthed, seems to have had a textile production role; a waterfront industrial area has been identified where dying and weaving materials have been found, associated with what seems to have been a wharf for vessels to tie up at (Carr 1985; Carr et al. 1988). However, there were non-ecclesiastical sites involved in specialised production. Wicken Bonhunt in Essex (Fig. 5.1:15) is one of the best examples of this. Although structurally it looks like a typical rural site, evidence for intensive meat production, in this case pig rearing and butchery, has been found. Its productivity seems to be reflected in the abundance of Ipswich ware, which made up 70 percent of the assemblage, and also imported Frankish wares which seem to have been shipped via London (Crabtree 1996; Wade 1980). The so-called Fen-edge sites near the Wash (Fig. 5.1:16), identified first by Ipswich ware scatters on the roddons (Silvester 1988), have subsequently produced evidence of specialist beef rearing. Some of this was doubtless for marketing on the hoof, but bone analysis and a salt pond found at Walpole St Andrew suggest that butchering and salting was taking place on at least some of these sites. Hay Green alone has produced over 1,000 sherds of Ipswich ware (Rogerson and Silvester 1986; Leah 1992, 55).

  There is also an increasing number of rural sites which provide evidence for neither an ecclesiastical connection nor specialised production but which appear to have been involved in trade in some way, simply as a result of their location on major trade routes. A classic example of this is provided at Lake End Road near Maidenhead (Fig. 5.1:17), where a large concentration of refuse pits contained imported pottery, lava quernstones and Ipswich ware but which, despite extensive excavation, has revealed no evidence of metalwork or permanent settlement (Blinkhorn 1998, S. Foreman, pers. comm.). Yarnton in Oxfordshire (Fig. 5.1:18) represents another apparently ordinary rural site which had access to the trade routes along which Ipswich ware and lava quernstones were shipped in the Upper Thames (Blinkhorn 1998). Sandtun at West Hythe, Kent (Fig. 5.1:19), is an inhospitable site on Middle Anglo-Saxon coastal routes. It has produced imported pottery and stone but very little metalwork. There is some supposition that the site was attached to Lyminge minster but it seems to have owed its imported assemblage primarily to its position as a natural harbour. Analysis of fish bones from the site suggests seasonal occupation (M. Gardiner, pers. comm.). Such small and sometimes seasonal sites were probably trading in occasional surpluses in textiles, or consumables.

  While the emporia have been touted as production centres, none of the household units at the wics would have looked out of place in the countryside, and production at the emporia, although diverse, does not seem to have been on a scale even close to the amount needed to support the trade which appears to have passed through them. Given the ubiquity of weaving evidence at rural sites, it would be hard to argue that many of the textiles reaching the emporia were not already in a finished state. Early sceattas associated with weaving debris at Shakenoak on the Upper Thames (Fig. 5.1:20), and lava querns and Ipswich ware found in weaving sheds at West Stow (Fig. 5.1:21), may support this (Blair 1994, 82; West 1985).

  The development of inland central places or areas with their own hinterlands

  As the archaeological and historical landscape behind the emporia of eighth-century England appears more clearly, the development of inland central places or areas, with their own hinterlands, are starting to emerge. This has been touched upon with regard to ecclesiastical sites and their interconnections. With artefactual evidence from individual sites, the Brandon finds assemblage illustrates how local, regional and long distance trade could come together at an ecclesiastical settlement. It has been suggested that the plant remains at the site suggest a consumer-type economy and certainly the raw materials for the textile production would have been acquired locally. The large number of oyster shells at the site must have come from the coast however, and the large range of imported finds attest to the association of the site with longer-distance trade networks (Carr 1985; Carr et al. 1988). This type of assemblage is not unique to religious sites: a similar assemblage of goods of varied provenance is visible at Lake End Road, which seems to have been supplied with Emmer wheat (a bulk import not known since Roman times), tidal and coastal fish and Continental pottery (S. Foreman, pers. comm.).

  It is also possible that multi-settlement focal areas were emerging in the Middle Anglo-Saxon period which developed more fully later. This is particularly visible in what became the Danelaw where, in the Middle Anglo-Saxon period, coin-rich sites are found near Thetford (at Middle Harling: Fig. 5.1:23) and Norwich (at Caistor St Edmund: Fig. 5.1:22), both of which later became urban centres (Wade-Martins 1997, 38; Rogerson 1995). The same may be true in the Middle Thames region in the Old Windsor/Staines area (Fig. 5.1:24); in the Upper Thames around Oxford (Fig. 5.1:25); and in the Bedford area (Fig. 5.1:26) (Palmer 1998).

  The distribution of lava quernstones from the Eifel region of the Rhineland to a large number of insular rural sites of all kinds is becoming increasingly apparent. This artefact distribution in itself points to the development of a multi-tiered system of distribution and exchange which mirrors indicators already highlighted at the level of individual settlements. Unfinished stones are known from the emporia, from Thetford and from Linford in Essex (Fig. 5.1:27) (Parkhouse 1997; 97, 102 and 104; Freshwater 1996, 42–5; Dunning 1956, 232; Barton 1968), which suggests that they were imported in block form and finished nearer to or on the sites at which they were found. These lava querns, despite being better than their native competitors, were still utilitarian items whose distribution indicates how widely they were traded. This would make sense, as such bulky and friable items would have to have been traded in large quantities to make profit viable. Here then, we may see artefactual evidence of bulk trade in agricultural goods which tallies with the evidence for the increasing mobilisation of resources in the English countryside at all levels.

  Evidence for growing specialisation and organisation

  Economic specialisation at sites such as Wicken Bonhunt and Brandon has been mentioned, and what appear to be iron production centres at Ramsbury (Fig. 5.1:28) (Haslam 1980), Romsey (Fig. 5.1:29) (Scott 1993 and 1996), and Little Totham (Fig. 5.1:30) (Adkins 1989) are also known.

  There is also evidence that at some sites the arrival of traded goods accompanied a change towards a more commercial economy. Pennyland (Figs. 5.1:31) has been regarded in this way by Paul Blinkhorn (1999, 15–16). In the Early Anglo-Saxon period the settlement existed as a fairly typical, unenclosed, scattered hamlet. The late seventh to early eighth century saw the first use of domestic enclosures, though the artefactual evidence suggests that at this stage the settlement was still engaged in a broad-based subsistence economy. However, when Ipswich ware started arriving at the site, apparently in association with lava quernstones, the enclosure system appears to have been abandoned for domestic use in favour of stock rearing and the domestic buildings retreated to a smaller cluster to the west. Blinkhorn suggests this indicates a major shift in the economy towards a less labour-intensive and more profitable stock rearing function (ibid.; Williams 1993).

  The Fen-edge also appears to show an expansion of rural activity. This area had been exploited in the Roman period but then fell out of use, partly due to climatic conditions. The Middle Anglo-Saxon period saw fresh advantage being taken of the good pastureland, communication routes and saltings at least for seasonal commercial purposes (Silvester 1988; Leah 1992). Another example of the reorganisation of resources can perhaps be discerned at Old Windsor. From the middle of the eighth century, imported goods, notably pottery, started to reach the site. By the end of the third phase of the settlement in which these contacts first appeared, there was a large three-wheeled mill in existence, contemporaneous with the introduction of a man-made leat and large-scale deforestatio
n of the area, which would seem to indicate increased exploitation of the agricultural resources of the region from the late eighth century (Wilson 1958).

  We can see that, particularly in the counties around London and Ipswich, evidence for what appears to be a quite complex and multitiered system of trade is emerging. This is represented by a range of sites, among which those termed ‘productive’, played a prominent but far from exclusive role. It further seems that geographical considerations, previous trade routes and the role of minsters and ecclesiastical settlements as central places had a major role to play in the shaping of these hinterlands. The growth of trade in goods, both agricultural and industrial, was a feature of the countryside as well as the wics. However, one of the most interesting conclusions to emerge from this survey is that, when we look at Hamwic and Hampshire, we are dealing with a different situation.

  Anomalies in Hampshire

  The hinterland of Hamwic does appear to have exhibited an unusual degree of royal control in its development. In retrospect, the clues were there from the start. The wic was laid out c. 700, after the other two emporia in question had started trading from more modest origins. The context of this foundation is also significant, coming as it did immediately after Csdwalla had captured the Jutish provinces on the south coast, providing the Gewisse with direct access to the sea. After this point, the tribe of the Gewisse starts to be recorded in documents as the West Saxon kingdom (Yorke 1989). Furthermore, the West Saxon kingdom was the first to organise its lands around a manorial system of royal estates with associated minsters. Hamwic itself, according to Hase, may have represented one of these estates and the recent discovery of a high-status cemetery close to the emporium may support this proposal (Hase 1988, 45). Such observations accord to an extent with the suppositions made by Hodges and those deductions which have been made from the excavations within the emporium itself.

  Certainly, the survey outlined in this paper highlights trends in the development of the Hampshire hinterland which are substantially different to those seen in the remainder of the study area. There are many fewer ‘productive’ sites in Hampshire than in other regions (Ulmschneider 2000a; Palmer 1998) and those sites which were associated with production and the movement of goods to the emporium, such as cattle from Bishopstoke (Fig. 5.1:32) and iron from Romsey, appear to have been tied in to the estate system (Nenk et al. 1992, 233; Scott 1993 and 1996). Furthermore, very few imported goods are known from rural sites in Hampshire.

  Communication routes were still of undoubted importance in the shaping of the hinterland, for instance Roman roads are clearly important in the distribution of coinage (Fig. 5.3). However, Hamwic had a very narrow hinterland. A re-examination of that area shown in Figure 5.2 illustrates that the Hampshire hinterland stands out rather pitifully in contrast to the extensive navigable networks spreading out from London and Ipswich. The rather restricted geographical extent of this late-formed kingdom may have been another reason for the royal impetus required in the control of the economy of the area.

  There are indications of some limited rural trade in Hampshire, as would be expected, for instance in iron bars for smithing, oysters and local hand-made pottery; but these sorts of trade do not seem tied to the emporium and there are no rural sites which seem to have boomed in response to increased trading opportunities provided by the development of the wic (Palmer 1998).

  FIGURE 5.3. Coin-productive sites (excluding hoards) and Roman roads in Hampshire (after Ulmschneider 2000a, Map 21).

  Even the distribution of the Hamwic-minted sceatta coinage, Series H, is anomalous, the vast majority of examples having been found inside the wic (Metcalf 1993–4, iii, 321–33). It has been suggested that this represents a positive balance of trade, but in the light of the other evidence, all it may represent is that this currency was primarily used as an internal token of exchange within the emporium and that monetary transactions had not spread into the hinterland.

  Hamwic is the best-studied wic site, at least partly due to its archaeology having been less disturbed than that of London or Ipswich because it did not survive beyond the ninth century. This early demise may be due to the artificial nature of the original foundation: the emporium was unable to survive the disruption of its Continental trade because it had no strong tradition of inland trade to fall back on. In this light, it is ironic that it has become the type-site for the English emporia.

  Conclusions

  In a survey of limited scope one can only hope to raise possibilities and questions rather than solid conclusions. This may be a healthy thing, as for too long we seem to have been concerned with debates over semantics and, as several contributors to this volume point out, a rehashing of a fairly limited dataset that has been known for decades. With increasing data becoming available, in terms of extensively excavated sites, the publication of rescue excavations, and increased numbers of single finds, now seems to be an opportune time to try out new multi-disciplinary approaches to Middle Anglo-Saxon trade.

  II

  Trading and ‘Productive’ Sites in the British Isles

  CHAPTER 6

  Markets and ‘Productive’ Sites: A View From Western Britain

  David Griffiths

  Discussions of ‘productive’ sites in Early Medieval Britain tend to focus exclusively on the eastern and southern counties of England. Yorkshire, Lincolnshire, East Anglia and the South-East account for almost all of the known ‘productive’ site locations. Rarely mentioned, by contrast, are the areas towards the western side of Britain, commonly referred to as the ‘Celtic West’ but which were affected in varying degree by Anglo-Saxon settlement (south-west and north-west England, Wales and its borders, and south-west Scotland), which have far fewer sites. This pattern begs the question as to whether this distribution pattern is real, or simply a reflection of greater metaldetecting and/or archaeological recording in the East and South-East. If, as argued here, it is to a large extent ‘real’ in that it broadly indicates the spread of sites rich in Anglo-Saxon non-ferrous metalwork in the seventh to tenth centuries, is this a reflection of economic ‘backwardness’ in the West? Or, is the relative lack of such sites more to do with a lower (but still detectable) circulation in the West of the characteristic Anglo-Saxon metalwork of the eastern sites, rather than a genuine absence of market activity? Following these basic questions, it seems a useful point to ask whether we should hold to a rigid definition of the ‘productive’ site as a wholly Anglo-Saxon material-culture phenomenon, and thereby make it a distant factor in understanding contemporary market sites in the West. Alternatively, should we adopt a definition which puts the widest possible range of indicators of market activity at the heart of our interpretation, and make the presence in sufficient numbers of deterministic ‘productivesite’ Anglo-Saxon metalwork types a secondary consideration?

  This brief survey of western Britain illustrates the marked difference between East and West in the quantity of ‘productive’ sites, if strictly defined as the presence of medium to large assemblages, perhaps over fifteen objects, of Middle to Late Anglo-Saxon period metalwork and coins (assuming one accepts the validity of the term at all – cf. Richards 1999b and this volume). West of a line drawn between (approximately) the Dorset/Hampshire border on the south coast and the English/ Scottish border on the east coast at Berwick upon Tweed, there are very markedly fewer sites than to the east. Interviews by the present author of finds liaison officers within the Portable Antiquities Scheme financed by the Heritage Lottery Fund, have confirmed the general pattern. Although in mid-2001 the scheme has not yet covered the whole United Kingdom, Dorset and Somerset, the English West Midlands, Wales and the North-West (defined as Cheshire to Cumbria) are under the remit of this programme. However, the picture in the West is not completely negative.

  Metalwork of the eighth to tenth centuries ad is known from a number of locations in western Britain (see below). There is considerable metal-detecting activity, and this has led to some interesting discoveries
. One such is the collection of material, including bronze Anglo-Saxon strap-ends and a pin, together with ironwork and a fine ninth-century openwork bronze plaque, from Asby Winderwath Common, near Appleby-in-Westmorland, Cumbria (Youngs 1999). Possibly indicative of a market site on a trans-Pennine routeway, the Asby Winderwath material shows that there is some potential for the eastern distribution of sites to extend northwards and westwards. In the Midlands, a site producing a range of Anglo-Saxon metalwork at Bidford-on-Avon in western Warwickshire (Wise and Seaby 1995; A. Bolton, pers. comm.), also pushes the distribution pattern westwards. A recent study of Germanic-Celtic interaction in the post-Roman period by John Hines includes a useful survey of some recent Anglo-Saxon finds in the South-West (Hines 2000), but unlike the Asby Winderwath material, for instance, these are largely individual discoveries, and cannot be said to resemble the traditional Anglo-Saxon ‘productive’ site artefact pattern. It is remarkable, if also ironic, that the great ninth-century hoard from Trewhiddle, Cornwall, should have become so emblematic in studies of contemporary Anglo-Saxon metalwork styles, yet was discovered in a region where comparable site finds have been so meagre.

  Regarding the distribution of sites producing metalwork of the eighth to tenth centuries, however, one significant difference seems to be that such sites in the West are predominantly coastal or near-coastal, whereas most ‘productive’ sites in the East are situated away from the sea. Two sites in particular which the author has been closely involved in researching, namely Llanbedrgoch (Anglesey, Wales) and Meols (Merseyside, north-west England) form relevant case-studies. The situation in western Britain challenges us to be flexible in our definitions for markets and ‘productive’ sites, and suggests that a longer-term perspective is necessary.