Markets in Early Medieval Europe

by Tim Pestell

  Could these entries reflect even earlier connections? (Fig. 7.3) When the West Saxon king Csdwalla conquered the Jutish Isle of Wight in 686, two young princes are reported by Bede to have fled to a place called Ad Lapidem (Historia Ecclesiastica, iv. 16), which has been identified as Lepe in the New Forest, in what would have been the adjoining mainland territory of the Jutes (Yorke 1989, 90). Estate ties between the island and mainland are also attested in a charter of 826, which records the grant of a large estate at Calbourne (see above) to the bishop of Winchester. One therefore must assume a fairly regular network of contacts and flow of goods between parts of royal and ecclesiastical estates in the two areas, and it seems likely that other landing and probably market places would have existed.

  Perhaps one of the most interesting questions is how and why this ‘productive’ site came into being. At present it appears from the coinage that the market would most probably have been established, or started to function, fairly shortly after Csdwalla’s conquest of the island in 686, in the immediate aftermath of which a quarter of the Isle was given ‘for God’s use’ to Bishop Wilfrid (Anglo-Saxon Chronicle, s.a. 686; Historia Ecclesiastica, iv. 16). Patrick Hase has long argued for the existence of an early mother church for the island at Caris-brooke, the ancient rights of which are preserved in the Cartulary of Carisbrooke, and the parochia of which by the twelfth century still appears to have comprised the vast area from sea to sea between Northwood and Shorwell/Chale (Fig. 7.2; Hase 1975, 323–33; Hockey 1982, 1–13, map in cover). Could the ‘productive’ site have been associated in some way with the foundation of an early mother church in the area? The suggested date of its foundation would certainly be consonant with that of five other mother churches around the Solent in the late seventh to early eighth centuries (Hase 1988, 45–8 and fig. 9). Unfortunately, in the absence of any corroborative written or archaeological sources, the evidence must remain highly speculative and entirely circumstantial.

  There is no further information about the area until Domesday, when Bowcombe was the site of the most important royal manor and hundred on the island, encompassing a church, mill, salt-house, and a toll worth thirty shillings. All the tithes of Bowcombe belonged to this church, to which were also attached a second mill and twenty smallholdings, inhabited by bordars (DB Hampshire, fos. 52b and c). Both the presence of the bordars as well as the toll all point very strongly to the existence of a market at Carisbrooke in the Late Anglo-Saxon period (Dyer 1985, 100–1).

  Where would this market have been located? There is certainly no evidence at present to suggest a continuation of the ‘productive’ site. So did it shift to another site? The only Late Anglo-Saxon coin so far known in the area is a penny of ^thelred II (985–91), found at Carisbrooke Castle car park. However, the find is said not to have been in situ, and may have arrived in soil carted from Newport ( Jones 1959, 157–9). As the excavated remains of substantial timber buildings attest, the later castle site was clearly important during the Late Anglo-Saxon period (Young 2000, 53–5 and 191–2), but other places also need to be considered, not least areas in the vicinity of the priory church, on the opposite site of Lukely Brook.

  Finally, a third line of enquiry must focus on the relationship between the island and the Continent. It is now becoming increasingly clear from its burials that perhaps already by the late fifth, but certainly during the sixth century, the ‘Jutish’ Isle of Wight exhibited considerable wealth and wide trading connections (Arnold 1982a; Morris and Dickinson 2000). Equally significant are the historically and archaeologically well-attested links between the Jutish island, the Jutes of Kent, and the Continent, at a time when both Kent and the Isle of Wight appear to have held a monopoly of southern cross-Channel trade (Historia Ecclesiastica, i, si5; Huggett 1988; Welch 1991). But what were the routes in operation, and how did they change over time? Their continuing use is now reflected both in archaeology and in early written records which, at least from the late seventh or early eighth centuries onwards, point to the growing importance of travel routes to Northern France and the Seine area (Le Maho, this volume; Johanek 1985, 222–5 and 234–44), Willibald’s ship perhaps being bound for the important early fair at Saint-Denis (Bauch 1984, 40–1 and 136–7).

  In conclusion, there is now evidence for a major ‘productive’ site evolving in an area that commanded important coastal routes and cross-Channel links from the earliest times. Access to these, as has been argued elsewhere, was probably one of the foremost reasons for the final West Saxon conquest of the formerly Jutish areas by the late seventh century (Ulmschneider 1999, 36–8). By the early eighth century, a number of trading posts existed in the area (Fig. 7.3), including the major emporium at Hamwic, the landing place at Hamblemouth, the ‘productive’ site on the Isle of Wight, and possibly other places, such as one recorded as ‘South Hampshire’, perhaps south of the New Forest (Rigold and Metcalf 1977, 47). In addition, Metcalf has suggested that the minting place for the West Saxon coinage of Series W may have to be sought in this area or a little further west (Metcalf 1993–4, 152–7 and 684), while ninth century coins are now emerging from a site at Eling Creek, Totton (Dunger 1997; EMC 1999.0098). Are we therefore beginning to observe a network of markets in operation, each seemingly bound to a major inlet?

  More research is needed, particularly on the chronology and interaction of these sites, before any such pattern can be proposed more confidently. Future research will also have to try to explain two other important observations: First, why, despite much metal-detecting activity, have ‘productive’ sites on the scale of those in the eastern counties not been discovered in the south of England (Ulmschneider 2000a, 107 and maps 5 and 21)? Second, how can the outstanding wealth of the Isle of Wight ‘productive’ site be explained? Apart from Hamwic, no other sites anywhere near its scale are known along the south coast, or within a fifty-mile radius inland (EMC, December 2000). Probably the most important explanation, cross-Channel trade, has already been suggested, but other factors also need to be considered such as the strategic position of the island, which provided important access to Winchester and the Thames Valley – a fact not lost on the Vikings (Anglo-Saxon Chronicle, s.a. 897 [896], 998 and 1001). Ultimately, part of the answer may also lie in the nature of Southampton Water and the Solent itself. Could some of these multiple markets have supplied ships or fleets anchored in and around the greatest natural harbour on the south coast?

  CHAPTER 8

  The Early Anglo-Saxon Framework for Middle Anglo-Saxon Economics: The Case of East Kent

  Stuart Brookes

  Introduction

  Recent contextual approaches to Early Medieval socio-economic development have stressed the important relationship between trade or exchange and political power in the formation of increasingly hierarchical structures of social dominance. Given the dearth of archaeological evidence for the nature of these early changes, such approaches are marked by the application of social theory to numerous multi-disciplinary lines of argument, to model wider developments on the basis of recognised social features. As such, the identification of seemingly specialised commercial or trading settlements (wics or emporia) during the seventh century, has been linked to wider theories regarding the increasingly centralised economic and political power of emergent polities during this period (for example Hodges 1978; 1988; 1989a and b; Arnold 1988; Carver 1989; Scull 1992 and 1993).

  In part fuelled by the importance of such trading sites and imported artefacts in archaeological literature, many interpretations of the social processes during the period have privileged the competitive exchange and consumption of foreign exotica as a causal dynamic to institutional change. The distribution pattern of these objects has been argued to represent a ‘prestige-good exchange’ apparatus operating between the kinship-dominated societies of the Early Anglo-Saxon period. As gift-giving and loyalty in military service formed the basis of social relations, a monopoly of the trade in luxury goods ensured political dominance within the competitive peer-
polities. The coastal zone (including Kent, the Thames Valley and the East Anglian headland) in which these objects are primarily clustered, and the evolutionist models proposed by analogy with the later emporia sites within this geographical area, have combined to emphasise early Kentish material within macroeconomic interpretations. In part explainable by the inevitable shipping patterns around the East Kent headland, Kent has historically been presented as the focal node of impact for diffusionist patterns of settlement, trading connections, state formation, and Christianisation. Certain documentary sources attest to an East Kent/Frankish connection, in the form of Frankish claims of authority over Kent, royal alliances and tribute extraction (Wood 1983), and these have often been utilised to explain the particularly dense patterns of imported goods in the county in the Early Anglo-Saxon period (Huggett i988; Welch 1991). Additionally, some evidence for inter-regional trade is supported by eighth-century toll remissions at the ports of Sarre and Fordwich (Kelly 1992). Despite little concrete archaeological evidence for comparable trading settlements therefore, Kent has often been presented as the first kingdom to adopt monopolising control over long-distance trade, a move that placed the sixth-and early seventh-century kingdom in a pre-eminent position over its neighbours.

  Such models of Early to Middle Anglo-Saxon socio-economic development suggest a number of further areas of investigation. The reconstruction of the spatial framework of increasing inter-(and by implication, intra-) regional trade and exchange suggested by the works of Hodges for example, can be explored both through the distribution of commodities in space (Renfrew 1975 and i977) or the arrangement of hierarchies of settlement with respect to the tenets of central-place theory (Davies and Vierck 1974; Arnold 1988, Fig. 5.5; Aston 1986). Carol Smith’s (1976) thesis on regional economic systems, from which Hodges drew much inspiration, also showed through a number of empirical studies, that the regional organisation of hierarchical central-places was crucially tied to modal networks, particularly in disarticulated, periodic market systems (ibid., 26). Following Stine (1962), Smith argued that, within a predominantly agrarian economy, traders need to be mobile in order to maximise the consumer demand range beyond the minimum range of goods traders need to offer to survive. Periodic markets, fairs and other specialised centres by their very nature therefore, are tied to trader mobility. Sedentary commercialised settlements, Hodges’ ‘type B’ emporia, could only occur when the consumer range increased enough to support permanent traders. These sites, be they emporia, inland settlements or markets, are expected to occupy key nodal positions within a region; straddling environmental zones, major routeways and frontiers, to maximise the demand catchment-area (Hodges 1989a, 52–3).

  Straight models of such economic central-place hinterlands show a decrease in demand density with increasing distance from the commercial centre. Close to markets, high demand density and transport efficiency enable continuous commercial activity (Plattner 1976). As the distance to the market increases, the zone of viable commerce is determined by a ratio between the periodic trader’s minimum daily income threshold and transport costs. Finally, a zone is defined, where demand falls below the trading threshold, and the transportation costs are so expensive, that even itinerant trade cannot be sustained. A model of key regions comprising the trading hinterland implies that the comparative pattern of consumption of individual communities is expected to vary regionally. Gateway communities or elite residences should demonstrate markedly different patterns of consumption and access to wealth markers than contemporary productive settlements. As part of this phenomenon, there should also be some evidence for cultural features associated with increasing specialisation, such as production, provisioning and distribution. Finally, institutional changes of the type suggested by Hodges, should also manifest themselves in increasing forms of centralised control.

  Corridors of movement, settlement and exchange

  Although this economically deterministic model runs the risk of oversimplification, there is some evidence for such restricted commercialisation in Early to Middle Anglo-Saxon Kent. Reconstruction of the region’s Anglo-Saxon transport geography presents a useful framework to examine the siting of many of the important settlements of the period. Drastic geomorphological change to the landscape of East Kent, as a result of Holocene sea-level changes and local coastal responses, means that many of the settlements of interest are now up to several kilometres inland (Fig. 8.1). Of particular importance amongst these, are those around the former Wantsum Channel in north-east Kent which have been continually argued to represent comparable, though possibly earlier, specialised trading sites to those excavated in other parts of England. In addition to such evidence of past riverine and coastal movement, the network of Roman roads radiating out from the civitas capital of Canterbury, and some Iron Age and Prehistoric trackways, continue to be preserved in part as contemporary routeways (cf. Margary 1946 and 1948; Knox 1941). Finally, the evidence for Anglo-Saxon and Medieval detached pasture, in the form of -den place-names, High Medieval lists of extra-manorial demesne and other charter evidence, has prompted some authors to posit a number of possible ancient droves existing as trackways and sunken lanes linking the northern coast with the Weald in the south-west (for example Everitt 1986, 36 Map 1).

  FIGURE 8.1. Map of East Kent, showing the reconstructed coastline c. 800 and its relation to some of the sites and roads mentioned in the text. The digital elevation model of East Kent was produced in ARCVIEW at 50m pixel resolution from 10m Ordnance Survey digital contours provided by the Digimap Project (http://edina.ac.uk/digimap/).

  Spatial analysis in the form of Kolgomov-Smirnov testing of Early Anglo-Saxon cemetery distribution in East Kent suggests that these were significantly structured in relation to existing routes of communication, i.e. Roman roads, navigable rivers or the coast. Approximately 85 percent of cemeteries are seen to lie in highly visible locations within 1.2km of these routes, while the location of the remaining sites can be argued to correlate closely to predicted droves, linking thirteenth-century estate centres with their appurtenant Wealden pig-pasture. Although the antiquity of these latter routes is uncertain, the case argued by historians (e.g. Everitt 1986; Witney 1976), the number of denns mentioned in Anglo-Saxon charters from the mideighth century onwards (such as Sawyer 1968, Nos. 24–25, 30, 33, 37, 123 and 125), the correlation of these routes with computerised GIS ‘least-cost path models’, and places-names containing the element -ora (Brookes forthcoming), all suggest a clear tendency for sites to be located close to roads or routeways. Given the conspicuous use of above-ground markers in Kentish cemeteries (in the form of secondary interment in prehistoric barrows or Roman monuments, as well as primary burial in Anglo-Saxon barrow cemeteries or cemeteries with other visible grave structures – Shephard 1979; Hogarth 1973) and the close association of burial throughout the period with the prevalent routes of communication, a deliberate rite can be suggested, determined in the first instance by selection of places in the landscape visible from the patterns of movement of the living.

  The evidence from the distribution of Early Anglo-Saxon cemeteries offers a picture of fossilised social movement from which to compare the pattern of hierarchical locales. In this regard, Tatton-Brown’s Kentish gazetteers of early towns (1984) and ten ‘old minsters’ (1988) finds close topographical correlation with both this pattern of Early Anglo-Saxon mortuary structures and Late Roman settlement. Utilising further documentary sources and the distribution of archaeological complexes, Everitt (1986) has also argued for a number of ‘seminal place’ estate-centres, forming the nucleus of colonising tenurial holdings in the North Downs and Weald. His hypothesis has stressed the importance of diverse economic conditions in the make-up of early estates by looking at the ancient patterns of land usage in terms of their detached lands and ancient common rights (f Witney 1976 and 1982; Everitt 1979 and 1986). Significant links between the more intensively cultivated lowland estates in the northern coastal fringe and Holmesdale valley and depe
ndent appurtenances in the Weald and North Downs have been argued to reflect not merely the pattern of economic settlement, but also the underlying basis of later administrative structures. Thiessen Polygons constructed around these settlements confirm the contrasting economic zones of agrarian resources underlying their topographical shape, and the importance of their location, almost equidistant, along the routes of communication (Fig. 8.2).

  If Everitt’s interpretation can be accepted (1986, 339–41), that a relatively established system of estates with its origins in the pattern of Romano-British settlement, formed the basis of seventh-century ecclesiastical foundations, the importance of network utility becomes clear both as a component of Middle Anglo-Saxon economic organisation and as a link to earlier patterns of ostentatious consumption. In the purely functional terms of an integrated regional system, nodal settlements could be predicted at the junction of major modal networks, where the costs of transference are minimised. As central-places for tenure spanning a variety of agrarian resources, and the further administrative implications such a pattern of estates suggests, manorial centres therefore needed to be sited at the nexus of inland routeways. Similarly, the link between trader mobility and periodic markets stresses spatial organisation circumscribed by geographical and social restrictions. The importance of Sarre, for example, has been seen with respect to its location at both the major Roman crossing point and at the conflux of the double tidal waters of the Wantsum Channel (Brookes 1998, 29–32). Fordwich’s significance has been stressed due to the settlement’s position at the tidal head of the Great Stour, and its probable role as the emporium for the nearby villa regalis at Sturry and Canterbury itself, while also offering the most likely point of trans-shipment between the sea-going Channel and coastal vessels, and the inland riverine modal network. Equally, Sandwich’s strategic position at the southern entrance of the Wantsum, adjacent both to the Roman roads to Dover and Canterbury and the large sheltered haven of the Meacesfleote, offered it topographical attributes which would secure its importance until well into the medieval period.