Martin Carver and Chris Loveluck with Stuart Brookes, Robin Daniels, Gareth Davies, Christopher Ferguson, Helen Geake, David Griffiths, David Hinton, Edward Oakley and Imogen Tompsett
This review explores the research opportunities and priorities currently presented by the maritime sphere in the Early Medieval period. The archaeological study of Early Medieval maritime Britain has great potential to illuminate the history of seagoing and inland waterborne traffic, processes of fluctuating economy and ideology, and the changing relationship between the English and the sea.
Britain is an island and the understanding of the role of its maritime interactions and international traffic is crucial for our history and its modern appreciation. The period under review saw the change from terrestrial and regional to maritime and international, and is therefore pivotal for what went before and came after. Throughout this chapter we remain conscious that the period under discussion is 600 years long and one of archaeology’s major duties is to identify and explain the changes that occurred (Fig 6.1). Accordingly we have here included historical gap-filling, social process, behavioural trends, and the expression of attitudes as being among the objectives of further research that are expected to bear fruit in studies of the Early Medieval marine historic environment.
Within the maritime zone we have included coastal areas, estuaries, and zones of former tidal creek systems, together with consideration of sea crossings and some river corridors. At this period the island was inhabited by a number of peoples, with slightly different geographical foci: the Anglo-Saxons (south and east), Britons (west and north-east), and Scots (north-west). This study focuses on England; however, since the subject is the investigation of maritime space, this necessarily involves not only Wales, Ireland and Scotland but also Scandinavia, the Low Countries, the Rhineland, and France. To help square this circle we focus on England but refer to the wider ocean peoples. It should also be noted that there is a long version of this chapter (Technical Appendix 5) available online (in which individual authorship of the different sections of text is clearly shown) that includes extensive case studies, in-depth discussion of the riverine system, artificial waterways, riverine economy, and rivers as boundary markers, the methods by which these broader research goals could be addressed, and discussion of cultural resource management issues.
Key research areas identified in this chapter include:
The coastline of southern England as it exists today was probably in place 3000 years ago (c 1000 BC), with localised variations thereafter caused by erosion and deposition related to sea-level change, tidal dynamics, climate change, and anthropogenic activities. Overall, the trend has seen net losses to the sea, with more friable coastal geologies most greatly affected, but in the shallower coastal waters of the Severn Estuary Levels, the Kentish Marshes, and the Fenland embayment, over 5090km2 have been reclaimed since the Roman period.
The rate of coastal erosion is significantly affected by climatic variation. Changes in prevailing winds and wind direction, ocean currents, prevailing sea temperatures, the occurrence of ice on rivers, lakes and seas, and general storminess all need to be assessed to calculate the extent and shape of landforms. The impact on the shape of the coast, the shingle bars, the configuration of tidal estuaries, and the viability of harbours can be profound, but may also be subtle, sporadic, and difficult to characterise within a restricted geographical region.
There are a few in-depth localised studies. The Romney and Walland Marshes have been the subject of long-term research identifying a detailed chronology of geomorphological and environmental change across the peninsula (Eddison 1995; 2000; Eddison and Green 1998; Eddison et al 1998; Long et al 2002). Further attempts to reconstruct regional coastlines during the Early Medieval period include the south coast (Harrington and Welch forthcoming), the Wantsum Channel and eastern Kent coast (Brookes 2007a), the Solent (Tubbs 1999, 10), and the Hampshire coast (Brooks and Glasspole 1928). In contrast, the area of the Levels and the Severn Estuary region appear to have been abandoned in the later Roman period and not fully reclaimed again until the 11th century (Rippon 2006, 80–1). There are also several detailed surveys of stretches of individual coastline, estuaries and major wetlands (Coles and Coles 1986; Coles and Coles 1990; Coles and Minnitt 1995; Minnitt and Coles 1996; Coles and Hall 1998; Crowson et al 2005; Hall 1987; 1992; 1996; Hall and Coles 1994; Hayes and Lane 1992; Healy 1996; Lane 1993; Pryor 2001; Silvester 1981; 1988; Waller 1994; Cowell and Innes 1994; Hall et al 1995; Hodgkinson et al 2000; Leah et al 1997; 1998; Middleton et al 1995; 2001; Van de Noort and Ellis 1995; 1998; 1999; 2000; 2001). These surveys are, however, multi-period, highlighting limited evidence of Early Medieval activity. Many of the surveys of wetlands were limited to studying land below specific contours, usually drained former wetlands, and did not necessarily explore sites on slightly higher roddons (islands) within marshes, or adjacent higher ground – areas which were used for permanent and seasonal settlement in certain regions between the 6th/7th and 10th centuries. It is the examination of the latter rather than the areas that were tidal or marshy that holds the key to understanding coastal and wetland exploitation and the importance of waterside location in the Early Medieval period (see Loveluck 2012; Technical Appendix). There is now a vast range of information available that should allow conclusions to be drawn about the location and changing character of the Early Medieval shoreline, and there have been a few finds which point to the potential of the resource. These include structures related to land reclamation (Rippon 2000; 2001a), and estuarine fishing.
Estimates of coastal erosion and change have also been attempted through the retrogressive analysis of published maps of the 16th to 19th centuries, recording terrestrial reference points, such as the Late Roman ‘Saxon shore’ fort of Reculver in Kent, or Warden Church on the Isle of Sheppey (Young 2004; Smith 1850, 1932). These suggest that, on the south coast, the erosion rate will have varied between 28m and 108m per 100 years (Valentin 1971). Between Selsey Bill and the mouth of the Cuckmere in East Sussex, the coastline in AD 400 may have been over 1700m further out (Goudie and Brunsden 1994, 48, fig 33). The coast of the Isle of Thanet and north Kent has lost land to a similar extent with estimates of up to c 3200–4800m for the same period (Brookes 2007a, 44). Between Folkestone and Dungeness in south-east Kent, however, the loss seems to have been far less, estimated at c 3–400m (Hole 1957; Young 2004). The same method of retrogressive map analysis has also been used to reconstruct the formation of shingle spits, barrier islands, and other coastal features (eg So 1963; de Boer 1996), which, despite being focused on Post-Medieval geomorphology, have considerable repercussions for the reconstruction of Early Medieval coastlines. In addition, survey of submerged offshore features can be used to extrapolate the shape of reconstructed landforms (So 1963; Dix et al 1998; Brookes 2007a). Further indications of coastal change are hinted at by written sources. For example, Selsey has over time been both separate from, and attached to, the mainland and is described by Bede as a peninsula joined by a narrow strip to the mainland (HE IV.16, Colgrave and Mynors 1969). Similarly, coastal erosion has accounted for the loss of a number of villages recorded in Domesday Book on both the south and east coasts (eg Brandon 1974, 117; Sheppard 1912, 49).
Different approaches are applicable to reclaimed coastlines. Grouped soil and drift units have been used to derive physical regions equating to the extent of floodplains against which archaeological distributions can be mapped (eg Allen and Gardiner 2000; 2006; Brookes 2007a; Hill 1981; Roberts and Wrathmell 2000), although these maps commonly suffer from the coarse scale of cartography and the lack of datable evidence. Field survey drawing on close-contour surveys, engineered structures and artefact scatters can be used to plot the extent of occupation at different times (eg Allen 1999; Reeves 1995). Survey data have also been compared with sedimentary analyses of the back-barrier marshes to reveal the complex evolution of palaeochannels and tidal creek systems underpinning marshland development (eg Waller 2002; Waller et al 1988; Burrin 1988; Spencer et al 1998). More recently, on the east coast of England, particularly in the Fens, and soon the Humber Estuary, LiDAR data has also been used (Challis 2004; Malone 2007; 2008; D Evans, pers comm). Again, topographical reconstruction from Early Medieval charters has provided important additional insights, recording for example the presence of early watercourses (eg Brooks 1988), harbours (eg Clarke 2012), and offshore islands (eg Gough 1992).
Coastal reclamation or land-claim was generally focused on areas of shallow coastal water. With their high levels of calcium carbonate, and sand, peat and clay structure, the soils of reclaimed saltmarshes, and to a lesser degree reclaimed tidal flats, were particularly desirable during the Medieval period (and beyond) as pastoral and agricultural land. Nevertheless, the construction of embankments and dykes, and the draining of land, always represented a significant investment of labour and infrastructural costs. It is likely that the most intensive and extensive coastal reclamation may be correlated with – but is not restricted to – firstly, periods of severe storm surges and sea-level change (ie during the Romano-British Transgression, AD 300–600, the Mid-Saxon Warm Period, AD 750–850, and the Medieval Warm Period, 950–1250 (Cracknell 2005, 2)), and secondly, the operation of large-scale surplus-producing economies in livestock and grain. However, detailed sediment surveys on the continental shores of the North Sea have shown that, in Flanders, drainage of already inhabited marshland landscapes was not linked to warm periods or sea- level change, but to the desire of the Counts of Flanders to gain greater control over the coastal marshes from the 10th century onwards (Baeteman 1999; Baeteman et al 2002; Tys 2003; Loveluck and Tys 2006, 154–7).
The three areas of the British Isles that witnessed the greatest extent of reclamation are the Severn Estuary Levels in south-west Britain, the Romney and Walland Marshes and Wantsum Channel in Kent, and the Fenland embayment on the southern North Sea coast, followed by the wetlands of the Humber Estuary and the Mersey and Dee estuaries. Evidence for drainage activities in the three former areas is well attested for both the late Roman and Medieval periods (Allen and Fulford 1987; 1990a; Eddison and Draper 1997; Rippon 1996; 1997; 2000; 2006; Cantor 1982; Darby 1983), and there are similar suggestions of Roman and Medieval drainage in parts of the Humber wetlands (Gaunt 2007).
Significant marshland reclamation also took place during the Early Medieval period, which may have followed a period when some low-lying coastal areas had been temporarily abandoned for settlement, between the 5th and 7th centuries, such as Romney Marsh, parts of the Fenland, and the Humber wetlands (Allen 1999, 16; Lamb 1995, 162; Cracknell 2005). The Humber wetlands were certainly exploited, however, both on the Yorkshire and Lincolnshire banks by adjacent settlements on slightly higher ground, such as Flixborough in north Lincolnshire, and Aldbrough in East Yorkshire (Loveluck 2007; 2012). In these areas the earliest sea-banks surveyed appear to date from the Middle to Later Anglo-Saxon era, the start of a long period of reclamation during the Medieval and Post- Medieval periods (Allen 1999; Silvester 1988; Lane 1993; Hall and Coles 1994). Written sources may support this claim: an uncertain charter of AD 772 (S108) refers to dykes on the Pevensey Levels. It is a mistake, however, to link a lack of formal drainage activity/dyke building with ideas of an absence of settlement in coastal marshland landscapes, since archaeological evidence is growing for significant and permanent occupation, in particular on sand islands within the low-lying coastal marshes of the North Sea coast of England, especially from the Humber Estuary to the Fens (Crowson et al 2005; Cope-Faulkner 2012; Loveluck 2012).
Reclamation has been viewed traditionally as driven by two principal objectives: to provide defence against flooding and as a means of increasing the amount of land available for agricultural exploitation. However, such environmentally deterministic and resource-based explanations may not have been the sole objectives. Drainage of coastal marshlands and waterways through dyke building also seems to have been linked directly to a desire on the part of landward-based political authorities for greater control over maritime-oriented coastal societies which had been difficult to administer for much of the Early Medieval period. This seems to have been the case in coastal Flanders and in the fens and coastal marshes of eastern England (Loveluck and Tys 2006, 161–2; Loveluck 2012; Loveluck forthcoming).
Towards the end of the 7th century coastal and estuarine ports (often termed emporia, or wics in an Anglo-Saxon context) began to develop around the English Channel and the North Sea (Hill and Cowie 2002; Coupland 2002). In England these trading sites have been identified at Hamwic, modern-day Southampton (Stoodley 2002), possibly Fordwich in Kent (Tatton-Brown 1984), London or Lundenwic (Hobley 1988), Ipswich (Scull 2009), and York or Eoforwic (Spall and Toop 2008). It has been assumed that Sandwich was a similar trading port though recent work suggests it was simply a beach landing place and that it did not have an international trading function (Clarke 2005; Clarke et al 2010). These wics had their counterparts in Holland at Wijkbij-Duurstede (Van Es and Verwers 1980) and in France at Quentovic (Hill et al 1992).
The English emporia were invariably located on a clear shoreline on a tidal river. Several of them, notably London and Ipswich, offered the potential for fostering regional trade upstream. Hamwic even offered a choice of rivers into the interior, though to what extent these were employed is an open question, since the absence of archaeological evidence for trade into the interior has led Palmer (2003) to question how typical Southampton is as a wic. Most of these rivers, however, could have permitted smaller vessels to travel down from inland regions, bringing goods to be traded with foreign merchants. However, the relationship between an emporium such as Ipswich and its hinterland remains far from clear.
Much has been written in the last 30 years about the emergence of coastal and estuarine ports from the 7th century onwards. These settlements have been classified and characterised using concepts borrowed from human geography and social anthropology: whether ‘gateway communities’, using the work of Hirth (1978); ‘ports-of-trade’, borrowing from Polanyi (Polanyi 1963; Polanyi and Polanyi 1978) and Renfrew (1975); or even ‘dream cities’, in seminal works by Hodges (Hodges 1982; 1989; 2000). All of these terms came with conceptual associations which viewed the settlements they described as ‘outside’ or something apart from the wider settlement and social hierarchies of their landward hinterlands. Gateway communities were viewed as trading settlements designed to exploit hinterlands, usually from a coastal location. Ports-of-trade were defined as liminal settlements founded on social and geographical boundaries by elite groups, with a view to controlling trade and wider socially embedded exchange, usually in objects classified as ‘prestige goods’. The concept of ‘dream cities’, as applied to the maritime and riverine central places of 7th- to 9th-century north-west Europe ascribed their existence and location to a conscious decision to locate beyond former Roman centres, perhaps influenced by monastic ideas of location apart from former centres (Hodges 2000, 86–92). However, such centres also existed in pagan northern Europe, where such an argument could not apply, and certain trading centres had an undoubted association with secular and religious authorities housed in former Roman townscapes immediately adjacent to them, for example at London and York (Malcolm et al 2003, 143; Kemp 1996, 76–83).
The ports-of-trade model espoused by Hodges, with amendments in the late 1980s stressing the importance of specialist commodity production and exchange at these centres, has been particularly influential during the last quarter century for the interpretation of the roles of the trading and artisan settlements around the Channel and North Sea, from the mid-7th to mid-9th centuries AD. The emporia were viewed as foundations by Frankish, Anglo-Saxon, and Scandinavian kings in order to consolidate and enhance their ruling authority. In particular, these central places were seen as entry- points for the controlled redistribution of luxury ‘prestige’ objects, which had social value due to their rarity. This was accompanied by the suggestion of a change in the organisation of production, both in the rural world and in the fabrication of specialist products at emporia (Hodges 1982, 50–6). However, at the time when these ideas were put forward, and generally accepted, comprehensive publication of much of the excavated remains from most emporia had not yet been achieved. Furthermore, detailed studies had not been undertaken of settlement patterns and exploitation of coastal zones adjacent to emporia, nor of relations between emporia and hinterlands in the interior, away from the coasts – apart from the suggested split functions between Hamwic-Southampton and Winchester by Biddle (1976, 114–15).
Moreland (2000) also highlighted the multiple spheres in which these settlements functioned, with their roles as foci for the redistribution of prestige goods via gift exchange probably existing alongside their role in specialist commodity production, exchange, and taxation. The likelihood that emporia contributed to profound transformations in the organisation of rural production and provisioning mechanisms or reflected changes that had already taken place was also stressed, not least in the provisioning of the dietary needs of emporia (ibid, 80–1). Significantly, Moreland also questioned the paramount role of kings and royal families as the sole controllers of the distribution of rare commodities derived from long-distance exchange, apparently channelled by emporia, but he still emphasised the role of elites in controlling surpluses and their transformation into imported goods via exchange (ibid, 101–3). Subsequent studies of import distributions in rural hinterlands of emporia, in the later 1990s and early 2000s, have further emphasised the likely channelling roles and links between predominantly elite rural centres and emporia, stressing connectivity between the ports and their hinterlands, and also the impact of the ports on the use of specific artefacts – such as coinage – in their surrounding regions (Palmer 2003; Naylor 2004).
The recent trend to stress the connectivity of ports with their landward rural hinterlands, especially via elite hierarchies and networks, has diminished the analysis of the port settlements themselves and of the archaeological signatures provided by their populations – in regard to their liminality or ‘otherness’ – compared to most contemporary rural communities (Loveluck 2012). In much of northern European scholarship, the emphasis on control of surpluses and exchange as a preserve of landed elites has also resulted in the presentation of merchants operating from these ports as highly subordinate clients, acting on behalf of secular and ecclesiastical patrons. In England, the potential for merchant seafarers to trade and make a profit, in addition to working for their patrons, has rarely been considered in the last twenty years, nor has their social background as people from coastal, seafaring regions. Only in recent publications in relation to Hamwic-Southampton and Lundenwic– London have the independence and profit-making abilities of merchants been considered, although to a limited extent (Birbeck 2005, 192; Malcolm et al 2003, 189–90). Works by Lebecq (1983; 1997) and Schmid (1991) for Frisia and the North Sea coast of Germany provide rare and now quite old studies of the social backgrounds of the seafaring and farming communities from whom specialist merchant households are likely to have emerged. Both envisaged potential for profit as a stimulus to the emergence of specialist seafaring traders, from the 6th and 7th centuries onwards. Similarly, the more recent work of Sindbæk (2007, 128–9) has stressed that a profit motive drove long-distance traders in Scandinavia, which stimulated a hierarchy of trading places as nodal points, not divorced from political support but alongside it. However, the origin of the long-distance traders and how such specialists developed was not discussed, nor was the social make-up of the nodal points themselves.
The pattern of nearly universal access to continental imports amongst the coastal social hierarchy between the Humber and the Fens, between the 7th and late 9th centuries (at least in terms of pottery and querns), has a number of implications for the understanding of the social dynamics of the North Sea coast (Fig 6.2). Firstly, it would appear that the trading centre at York did not exhibit any significant control over the actions of seafaring traders around the Humber Estuary, nor is there evidence of control of their actions as they sailed up the east coast. The very widespread occurrence of imported goods in the coastal margins suggests that if there had been an intention on the part of Anglo-Saxon royal powers to control access to imported goods using emporia centres, then they failed in that role. Secondly, certain distributions of imported goods also suggest the operation of different maritime connections and trading activities along the North Sea coast from the Humber to the Fens. This is perhaps reflected most clearly in the distribution of Ipswich ware. The excavations at Fishergate and other deposits from later 7th- to late 9th-century York have yielded comparatively little Ipswich ware, perhaps as few as 50 sherds, and the ware is hardly represented in areas between York and a concentration around the Humber. The widespread occurrence of Ipswich ware runs in a band only c 10km wide around the shores of Holderness and the Humber Estuary, and then extends down the east coast (Fig 6.3). There is no apparent distribution linked to sites of specific character or status in this coast and hinterland zone. A rank-related distribution may certainly be reflected in quantities of Ipswich ware, but it is undeniable that a far greater spectrum of the population along the coast had access to Ipswich ware, in comparison to the inhabitants of York and its immediate hinterland. This suggests the existence of different exchange networks operating via the coast and via the trading centre at York, even though the same seafaring merchants may have been involved in both networks (Loveluck 2012).
The existence of different trading networks may also be reflected in the use of coinage. Around the Humber, coinage was deposited at landing places and larger settlements from the end of the 7th century and the vast majority of the coinage deposited was struck in Frisia and northern France until the 730s, although, significantly, the earliest silver coinage struck in Northumbria, by King Aldfrith (AD 685–705) also has a concentration around the Humber and East Yorkshire coast, with discoveries at Whitby, and the Humber landing place at North Ferriby (Loveluck 1994; 1996, 43–5). Aldfrith died at the royal estate of Driffield, at the headwaters of the River Hull which leads into the Humber. His presence and coinage may suggest an interest in facilitating the trading activities of the Humber zone. Interestingly, silver coinage dating from between the late 7th century and the 730s is currently very rare in York, with only one Frisian issue being found at Fishergate, prior to a predominance of Northumbria issues (Kemp 1996, 66; Naylor 2004).
This difference between York and the coastal zone indicates, firstly, that the Humber Estuary was the major contact and exchange zone prior to the foundation of a trading centre at York (Loveluck 1996, 43–5; Naylor 2004), and that the distinctiveness of the populations of the coast was maintained via direct maritime connections, even after the Fishergate settlement at York was in existence. There is a greater quantity of continental pottery at Fishergate, however, when compared with the coastal settlements, and this may reflect a greater concentration of foreign seafarers operating in York from the mid-8th century onwards. In this context it is also interesting to note that the coastal concentration of Ipswich ware is evident on the Humber from the early 9th century, on the basis of current excavated sequences (Loveluck 2007). This could reflect different mariners operating around the Humber or differential choice on behalf of foreign seafarers in terms of what to trade.
In the 10th and 11th centuries there was a fundamental shift in the nature of maritime settlement. With the socio-political changes of these centuries (namely, Scandinavian elite presence in the urban centres of eastern England, the creation of the West Saxon Kingdom of England, and the Danish and Norman Conquests) major port towns became fully integrated with their rural hinterlands, at the same time as the scale of maritime-orientation and freedoms of coastal populations diminished overall. This seems to be a pan-North Sea trend. In Jutland, direct exchange contacts with foreigners were refocused on royal port towns, and former complex coastal settlements were redefined within a process of rural manorialisation (Loveluck 2012). The 10th- century changes at Flixborough can be attributed to a similar redefinition of urban-rural relations; likewise, in coastal Flanders, the populations of coastal marshes lost their long-distance contacts at the same time as the onset of major land reclamation and the growth of new port towns, such as Bruges, directly sponsored by the Count of Flanders (Loveluck and Tys 2006, 162). Hence, diminished long-distance contacts coincided with increases in the demonstrable power of royal and regional governments over the worlds of their coastal margins.
The towns, especially major sea or river ports, became the principal locations for artisan and trading activity, producing finished goods for their surrounding regions in a way that had not been the case with most of the earlier emporia (with the exception of Ipswich, in relation to Ipswich ware pottery and, potentially, its contents). This resulted from a combination of more developed governmental structures, with towns as administered regional central places, markets, and taxation collection points, as well as preferred locations for trade, on the part of seafaring merchants. For example, in 10th-century York, while under Scandinavian rule, the concentration of secular political patronage and ecclesiastical patronage (from the Archbishops of York) resulted in very wealthy artisan and resident, or transient, merchant populations. The remains from the Coppergate excavations illustrate this point, with their concentration of iron workers, and gold and silver workers, amongst other crafts.
Also found within these same 10th- to 11th-century artisan/merchant tenements were riding gear and weapons (spears, arrowheads, and sword furniture) and the reused Coppergate helmet, along with items denoting integration within Scandinavian trade routes to the Orient, in the form of silk and Islamic coins (Hall 1981; Hall et al 2004; Ottaway 1992; Tweddle 1992; Walton Rogers 1997). By AD 1000, London was the object of twice yearly visits by merchants, known as ‘Esterlings’ (the easterners), who paid their port tolls in large quantities of pepper from Indonesia or the Malabar coast of India (Keay 2006, 108). By the 11th century, therefore, it was this Early Medieval ‘globalisation’ that set the major port towns and their societies apart from those of the countryside.
Research on the wics thus needs to be broadened and deepened in the light of these ideas. It is not a given that these settlements were short-lived royal instruments; we should expect them to have a longer life, a more plural participation, and more active networking with inland markets. There is a need to publish the results from Ipswich and to refocus research on the waterfronts of all known wics. Those that are still unknown (eg Sandwich, Fordwich) are attractive research targets. Overseas, British archaeologists have much to learn from the early beach-markets being defined in Scandinavia and the settlement patterns in the Frisian mudflats. These studies probably provide the most immediate analogies to the genesis of maritime trade in Britain.
In contrast to the network of wics and larger port towns set out above, the smaller anchorages and landing places of the Early Medieval coastline are less well understood. Landing places, ie specific points of embarkation and disembarkation for seagoing, coastal, and estuarine ships and boats, have proved highly elusive in the archaeological record due to changes in the nature, perception, and use of the edge of land. Excavated evidence for landing places, and suggestions of landing places from artefact scatters, has tended to be found in estuarine, inlet, fen-edge, and riverine locations rather than on modern-day coastal and beach locations, although suggestions of Early Medieval landing places in all the latter locations have been found around the coast of Britain.
Overall, the evidence from which we might reconstruct harbourages, landing-places, and foreshore activities for central southern and south-eastern Britain in the Early Medieval period is fragmentary and in many places dependent on Late Roman evidence (various examples of which are cited in Cracknell 2005). Present-day coastal inlets may have been longer, and suitable beaching places on river estuaries may well have been further inland from the sea 1500 years ago. A case-study from the River Thames has demonstrated that the tidal head of the river – and thereby the range of easy navigability – has moved slowly upstream, ie westwards of the City, since Saxon times, but that there were also brief periods (as in the late 10th to 11th century) when river levels also swung back again (Thomas et al 2006). Similarly, sea-level changes mean that archaeological horizons containing evidence for maritime activities may be preserved today on dry land, in the intertidal zone, or below sea level. Around the coasts of south-west Britain, the locations of landing places in sheltered sea coves and on beaches are consistently suggested, whether below sheltered headlands, such as Tintagel, or partially covered under modern dune systems, as at Bantham Ham, Devon (Fox 1955; Silvester 1981; May and Weddell 2002), and Gwithian, Cornwall, amongst others (Nowakowski 2004; see also Tompsett in Technical Appendix 5).
The evolution of individual harbours (including an assessment of the Early Medieval topography) has been carried out in piecemeal fashion, examples of which include Dunwich (Chant 1986), Langstone (Allen and Gardiner 2006), Brading and Pagham (Wallace 1999), Sandwich Haven (Brookes 2007a; Clarke 2012), and Broad Water (Kerridge and Standing 1987). Unfortunately, excavation of many of the major wics (eg Saxon Southampton, York, and Ipswich) has not included the excavation and survey of related harbour areas, a trend which continues with the sample excavations and surveys of beach/ dune and tidal creek-sites (eg Sandtun, Bantham Ham) which were possibly only occupied seasonally. Along the North Sea coast of eastern England, from the Humber Estuary to the Fens, concentrations of late 7th- to mid-8th-century sceattas and pottery at North Ferriby (on the north shore of the Humber) and similar concentrations at Halton Skitter and South Ferriby on the south bank, suggest beach trading sites in these locations. In addition, a series of fen-edge and river landing places are also becoming apparent (Loveluck 2012). For example, a 7th- to early 8th-century logboat and reveted wooden trackway from a landing place have been excavated at Welham Bridge, East Yorkshire, on the landward edge of fenland waterways that would have led into the Humber (Allen and Dean 2005, 91– 3; D Evans, pers comm). A second wooden revetment for a jetty landing place has also been excavated at Skerne, on the River Hull, close to Driffield, and it may have been linked to the royal estate centre there (Loveluck 1996, 44–5; Dent et al 2000). The actual locations of exchanges with mariners were probably the beach sites, and coastal and riverine boats, such as the Welham Bridge logboat, could have been the principal method of dispersion of goods around the coastal zones. However, in some instances seafaring merchants may have moored directly at riverine landing places close to estuaries. Flixborough certainly had watermills below the settlement on the River Trent in 1066, and these are likely to have been combined with jetties, akin to Skerne (Loveluck 2007, 86). It is less likely, however, that seagoing ships sailed up the River Hull as far as Skerne, although seagoing ships were certainly based in Beverley in the 12th century. The locations of the contact between the inhabitants of small marshland hamlets and mariners are less easy to predict, but the situation of the hamlets on tidal channels near river estuaries, as at Fishtoft, Lincolnshire (J Rackham, pers comm; Cope-Faulkner 2012), and West Walton, Norfolk (Crowson et al 2005), suggests that ships moored for the night or to reprovision would have been visible from some distance and contactable via tidal creeks and riverine and coastal boats.
Studies focusing on material culture may be of use here. The distribution of imported objects as chance finds and excavated items provide some indication of the range and spread of overseas contact, as well as – potentially – the routes that this traffic took. Numerous studies exist which apply this principle at the regional scale (eg Huggett 1988; Welch 1991; Harrington and Welch forthcoming; Hines 1984), but the same technique has also been used to pinpoint the location of individual maritime entrepots, in eastern, southern and western England, namely in southern Lincolnshire, East Anglia, Hampshire, the Isle of Wight, and Meols, in the Wirral (eg Pestell and Ulmschneider 2003; Griffiths et al 2007). At present most of the possible Early Medieval landing and/or trading places recorded by the Portable Antiquities Scheme (PAS) appear to be riverine rather than coastal. Twenty sites have been recorded across the country with four or more sceattas and ten or more pieces of other Early Medieval metalwork. None of these sites is within 2km of the contemporary coast and several are as much as 50km distant. A few sites are well known, such as Coddenham (Suffolk) and Heckington (the ‘South Lincs productive site’), but most have received little study (see Davies 2010). Material culture studies should also help in improving the underlying theoretical framework. The expectation that Early Anglo-Saxon landing places, Middle Anglo-Saxon beach-markets or trading ports (Pestell and Ulmschneider 2003), and Late Anglo-Saxon shipyards or fishing ports (Barrett et al 2008) all have different archaeological signatures has not yet been addressed (Ilves 2009), but the PAS database alone suggests that a multi-disciplinary approach may have significant and immediate benefits in this respect.
There is evidence of a variety of subsistence activities and coastal ‘industries’ in the coastal landscapes of Early Medieval England, including mixed arable and pastoral farming, salt production, and fisheries of different kinds, as well as coastal wildfowling. For example, within the marshland landscape along the east coast from the Humber to the Fens, the results of the Fenland Survey (Hall and Coles 1994) and subsequent excavations indicate a landscape of farmsteads or small hamlets sited on sand islands within less well-drained marshland, sometimes located in proximity to tidal creek waterways (see Technical Appendix 5). The landscape of small hamlets and farmsteads, dating from the 7th to 10th centuries (and later) excavated at Gosberton, in the Lincolnshire Fens, and the settlement on a sand spur adjacent to a tidal channel at Fishtoft, near Boston, Lincolnshire, provide the best examples of such settlements to date (see Crowson et al 2005; Cope-Faulkner 2012), and these hamlets appear to have been permanently occupied, with mixed farming economies suited to saltmarsh environments. Due to the environmental conditions, however, there was a significant bias towards the raising of cattle, sheep, and horses, reflected by the preponderance of young and sub-adult animals at the Gosberton sites, although barley, a salt-tolerant cereal, was also grown.
The circumstances of their living environments, therefore, resulted in coastal marshland communities with a predisposition for the production of specialist products and the need for exchange for their daily needs. A recurrent pattern of iron smithing was also found, probably exploiting a bog- iron ore source, and possible hints of salt production were also identified (Crowson et al 2005). At Fishtoft definitive evidence of salt production dating from the 8th and 9th centuries was recovered, in the form of large quantities of securely stratified briquetage (E Morris, pers comm; Cope-Faulkner 2012). There remains, however, limited archaeological evidence for Early Medieval salt production, in contrast to the extensive Domesday evidence for salt production in 1066. The only other excavation of a coastal salt-producing site with good stratigraphic evidence from this period, dating from the late 9th to 11th centuries, comes from Marsh Chapel, in the Lincolnshire sea marshes, a location which also suggests a focus on animal husbandry (Fenwick 2001). The occupants of these marshland hamlets, who concentrated on the raising of livestock and production of salt, with more limited cereal production and inshore fishing, may have been loosely incorporated into estate structures or may have been free proprietors. Whatever their tenurial relations, however, they would have needed to enter into exchange or redistributive relations to support aspects of their dietary, raw material, and other life needs. Importantly, if they were tied to estate structures this did not diminish their ability to profit via direct maritime exchange with mariners (Loveluck 2012). Of particular importance on the eastern and south- eastern coasts were the sea fisheries. The chronology of the development of sea fisheries, and the intensification of their exploitation, is complex and less well understood than recently thought. Detailed studies of the development of deep sea fisheries in the North Sea and Channel had suggested that their exploitation may have been linked to stimulation under Scandinavian influence. The work of James Barrett and others certainly demonstrates that there was a huge change in the scale of exploitation of deep sea fisheries in the Northern Isles, with Scandinavian settlement, acculturation, and hegemony (Barrett et al 2001; Barrett and Richards 2004). However, fish like cod, haddock, and whiting were already exploited by the native inhabitants of sites in the Shetlands, such as Scalloway, as a result of inshore fishing before Scandinavian influence changed the scale of exploitation (Cerón-Carrasco 1998, 112–16). The onset of the greater exploitation of sea fish seen to a certain extent in 10th- and 11th-century coastal and riverside towns in England cannot be attributed to Scandinavian influence alone, however. Recent excavations at Mid- to Late Saxon rural settlements on the Channel coast of England, at Bishopstone, Sussex, and Lyminge, Kent, by Gabor Thomas, have yielded significant quantities of large members of the gadid family – cod, haddock, ling, whiting etc which could have been derived from deep-sea rather than coastal fishing from the 8th to 9th centuries onwards (Rebecca Reynolds, pers comm). Other coastal and estuarine settlements, such as Flixborough, Lincolnshire, also exhibit freshwater, estuarine and deep-sea species, with little change in fish consumption patterns between the 8th and 10th centuries (Barrettt 2007), whilst the perhaps seasonally occupied coastal port at Sandtun, West Hythe, in Kent, dating from the 7th to 10th centuries, has produced a full range of marine and estuarine fish, including cod, herring, haddock, ling, and whiting, demonstrating coastal fishing at least (Gardiner et al 2001; Hamilton-Dyer 2002, 256–61).
Fishermen supplying settlements such as Bishopstone, Lyminge, Flixborough, and presumably those living at Sandtun, seasonally or permanently, could have been exploiting the larger deep-sea fish when they came close to inshore waters, although this may be significantly underestimating fishing abilities in deep water between the 7th and 9th centuries. The evidence collated by Barrett undoubtedly shows an increase in fishing for deep-sea species during the 10th and 11th centuries, which may also be linked to sustaining growing urban populations, an increasing concentration of shipping in 10th- and 11th-century ports, and hence also growing markets for fish within the context of religious observance (Barrett et al 2004). In some circumstances, the presence of large numbers of cetaceans, in the form of porpoises or dolphins, can be explained as a social marker of the status of settlement inhabitants, rather than a reflection of subsistence-based consumption. The consumption of perhaps up to 30 dolphins at Flixborough appears to be a practice of the secular aristocratic phases of that settlement’s history, in the 8th and 10th centuries (Loveluck 2007; Dobney et al 2007). Gardiner has also noted that access to beached cetaceans tended to be a preserve of elites, when such action could be policed (Gardiner 1997).
Herring fisheries, in particular, were also an important part of the late Early Medieval economy at least from the 10th/11th century (see Pelteret’s discussion in section 2.3 of Technical Appendix 5); herring was transported up-river by boat when necessary, as a herring-processing factory in York, dating from the early 10th century, indicates (Cramp 1967, 18–19). Campbell (2002) has explored the implications of the evidence in Domesday Book for herring renders, which are recorded for the shires of Kent, Surrey, Sussex, Norfolk, and Suffolk. The quantities are large: Dunwich for example supplied 60,000 herrings to the king. The total renders for East Anglia in 1086 amounted to 164,900 herrings, which Campbell conjectures could have amounted to a total catch of well over 3,000,000 fish. Moreover, Campbell suggests more than 5 tons of salt would have been needed to conserve the herring catch (Campbell 2002; Morely and Cooper 1922, 4). Campbell also provides plausible evidence that the inhabitants of the inland, riverine settlement of Frostenden in Suffolk used boats to reach the sea to fish along the coast. The herring and other fisheries would therefore have been at the centre of a vast web of economic enterprises and social relationships, and the connections between these deserve further research.
Recently, work has been done through documentary sources on the development of fishing settlements, particularly in the south-west (Fox 2007), but further studies of the archaeological evidence of these communities and industries are required. In addition, it appears that in a number of locations sea fishing was combined with other fisheries, notably estuarine fish traps. At Tidenham in Gloucestershire, for example, 65 basket weirs produced catches including sturgeon, herring, and even porpoise (Murphy 2009, 47). Further evidence of the role of certain species as status-markers is reflected in the reservation of porpoises and sea- fish for the lords of the manor of Tidenham between c 950 and 1066. Archaeological surveys have identified the widespread construction of fish weirs in the Severn Estuary and north Devon, although only the weirs in Bridgewater Bay have been dated, to AD 932 and 966 by dendrochronology (Groves et al 2004). On the east coast, radiocarbon dating of fish traps, on the Essex coast at Bradwell-on-Sea, in the Blackwater Estuary, and in Suffolk at Holbrook Bay on the Stour and on several sites in Norfolk, suggests an intensification of activity in the 7th to 9th centuries (Murphy 2009, 48). Such sites are increasingly being identified in other locations on the English coast (see Cowie and Blackmore 2008; Cohen 2003; Strachan 1995; 1998; Wallis and Waughmann 1998), but though there are a number of types of fish trap, including composite stone/timber traps (and numerous local variants), there is as yet no typology, nor any clear sense of how these fisheries fitted into the local or regional economic networks of the communities who built and used them.
There is also some evidence of the exploitation of shell fisheries during this period. In general, oysters predominate in Early Medieval deposits (Murphy 2009, 50), though mussels are also common in waterfront deposits at Whitefriars Street, Norwich (Ayers and Murphy 1983). There is evidence for the management of oyster beds during the Roman and Medieval periods (Winder 1992), but there is so far little evidence of this during the Early Medieval period and the collection, potential management, and consumption of shellfish during this period would benefit from further research. Finally, coastal wildfowling was also practised. Baker (2005) notes the presence of the bones of wild geese, duck, coot, small waders, and a harrier in domestic deposits from fenland sites in Lincolnshire, and the occasional consumption of seabirds is also demonstrated at monastic sites such as Hartlepool (Loveluck 2007; Rackham 2007). Again, large-scale exploitation of coastal and marshland wildfowl seems to have been a marker of secular elite exploitation and control of landscapes, as reflected at Flixborough in its likely secular elite phases, in the 8th and 10th centuries (Loveluck 2007). Exploitation of wildfowl at monasteries or monastic estate centres seems to have been more limited to occasional exploitation, as reflected by the huge decrease in wildfowl at Flixborough at the end of the 8th and through the 9th century (when literacy and items such as window glass were also present). The eating of seabirds at Hartlepool seems to have been very occasional.
For more than a century from the 790s the Vikings used the waterways in and around Britain to pose a considerable threat to Anglo-Saxon communities. Approaching from the sea, the Vikings were able to mount surprise attacks on vulnerable coastal and estuarine sites, causing devastation and striking fear into the hearts of contemporaries, not least the chroniclers who recorded these actions. By the later 9th century, sea and riverine routeways were used, sometimes in conjunction with overland routes, to stage sustained attacks on the various Anglo- Saxon kingdoms, with the ultimate aim of conquest. At various times during these campaigns offshore islands and peninsulas (eg Mersea, Shoebury, Benfleet, and Sheppey) were used as Viking bases. The intensity of these attacks forced those who were eventually to prevail against the Vikings, such as the West Saxon kings, to implement new military arrangements to counter the threat, including the creation of a series of fortified sites across southern England. In order to understand the strategic importance of these defences it is necessary to appreciate not only their context relative to the land and sea, but also the nature of the menace posed by water- borne Vikings.
Archaeological and toponymic evidence suggests that a range of sites linked to coastal defence and intelligence existed, at least by the 10th century (Baker and Brookes forthcoming). At this time the Roman pharos of Dover was refurbished, perhaps in order to provide early warning of a threat crossing the Channel. Further lookouts are evidenced by place-names containing the elements weard and *tōt. Perhaps supporting this system were a number of further structures which could serve as convenient observation points, notably freestanding stone or timber towers and ringworks, as well as the turriform (tower-like) churches of the 10th and 11th centuries. Several examples of these private defences are known from the south and east coasts: excavations at Bishopstone in Sussex, overlooking the mouth of the River Ouse, have revealed the cellar of what was probably a substantial timber tower in the 10th/11th century (Thomas 2005); at Jevington, overlooking the mouth of the Cuckmere, is the still extant 11th-century turriform church of St Andrew (Taylor and Taylor 1965); and on the River Adur is the enigmatic ringwork of Old Erringham, dated through coin evidence to the late 10th century (Holden 1980).
Discovery of the archaeological remains of ships and boats dating from the period between AD 400 and 1100 has been exceptionally rare. To date, no Early Medieval shipwrecks from submerged marine contexts have been retrieved and only one possible wreck site of this period is known around the British coast – a possible Viking shipwreck is suggested by the recovery of a sword-guard from the Smalls Reef, off the coast of southern Pembrokeshire (though this may be an artefact lost overboard and requires further investigation). The sword-guard was decorated in late 11th- to early 12th-century Urnes- style decoration. The wreck site of the possible Viking ship had been eroded over time, and had subsequently been overlain by the wreck of the steam ship Rhiwabon, which sank in 1884 (Redknap 2000, 58–9, 87).
However, remains of ships and boats have been found on land and within sediments of estuaries, rivers, and former tidal channels. The nature of the remains fall into two categories: firstly, the remains of complete and partial ships and boats that had been deliberately interred within the context of funerary and burial ritual; and secondly, the remains of ships and boats that had been abandoned at landing places or in tidal creeks and sometimes reused in landing place revetments. The best-known ship remains from the Early Anglo-Saxon period are the clench nail and degraded wooden outlines of the late 6th- to early 7th-century ships and boats that were interred within the context of ostentatious furnished burial at Sutton Hoo and Snape, in Suffolk (Carver 2005; Filmer-Sankey and Pestell 2001). To these can be added the boat rivets/ clench nails reflecting reused fragments of clinker- built vessels from 6th-century graves at Mill Hill, Deal, and Minster Thorne Farm, amongst others in Kent with isolated clench nail/rove finds (Brookes 2007b, 14–15). Other examples of fragments of timber held together with clench bolts have been found in graves along the east coast of England, dating from the 7th to 9th centuries, for example at Dover-Buckland, Kent, Caister-on-Sea, Norfolk, and Castledyke, Barton-upon-Humber, North Lincoln- shire (Brookes 2007b, 16–18). These burials have been described as ‘pseudo-boat burials’ (Brookes 2007b), although they could also reflect the reuse of timber derived from boats rather than a similar tradition to the boat-burials of early 7th-century Suffolk or the boat-burials of the Early Medieval period found in both pagan and Christian contexts in northern Germany and Scandinavia (Schön 1999, 76–9; Birkedahl and Johansen 1995, 160–4).
Unlike the discoveries based on clench nails and soil marks, a few more complete finds of ships and boats have been recovered in waterlogged deposits, in the same locations as excavated landing places. That is to say, the wrecks have tended to concentrate in harbour silts and around the margins of river estuaries and their coastal marshes. The excavated wrecks fall into two types: seagoing and coastal vessels, represented by the Graveney ship (Fenwick 1978a), found in the mud of a former tidal creek in Kent; and smaller boats, seemingly designed for river and estuarine transport, located in beach and estuarine foreshore locations.
There is evidence for ship and boat repair at a number of other landing places beyond Graveney. Repair would certainly accord with the wood- working tools associated with the excavated jetty on the River Hull, at Skerne, East Yorkshire (Loveluck 2000, 227–37), and the wood-working tool hoard from Flixborough, Lincolnshire, a settlement known to have been linked with a landing place and two mills on the River Trent (Darrah 2007, 60–1; Loveluck 2007, 82; Ottaway 2009, 256–66). The relationship between the location of wrecks and riverine and estuarine landing places and waterfronts is further reinforced by other discoveries of ship fragments, for example, the recovery of an early 11th-century side rudder from a waterfront in London (Goodburn 1993, 57–9), and a fragment of ship planking, dated between AD 920 and 1080, from a revetment on the bank of the River Usk, at Newport, Gwent (Redknap 2000, 60).
Examples of smaller boats have been recovered from Welham Bridge, East Yorkshire, and the harbours at Poole and Langstone on the south coast. All of these small vessels were logboats. The Welham Bridge logboat had been broken up and built into the revetment of a landing place, not far from the known Anglo-Saxon settlement focus at Holme-on- Spalding Moor, on the edge of the marshland and tidal inlet called Walling Fen that ran into the Humber. The boat is likely to date from the 6th century (Fig 6.1), whilst the wattle trackway of the landing place, for which the boat had formed part of the revetment, was dated AD 530 to 690 (Allen and Dean 2005, 91–3). The Early Medieval logboat from Welham Bridge represents continuity in the use of this form of estuarine vessel around the Humber from the Iron Age – the Hasholme boat was discovered in a very similar fenland edge/tidal channel location nearby (Millett and McGrail 1987, 69–125). The Langstone logboat has been dated to c AD 500 (J Satchell, pers comm), but accurate information on the Poole logboat is not available; it is thought to date from the 10th or 11th century (see Hinton in Technical Appendix 5).
The Anglo-Saxons and Britons were surrounded by three seas. Each of these seas had been regularly crossed since the Bronze Age, and arguably since the Mesolithic, so we are not concerned with people discovering new land but with an evolving practice (see Marcus 1980; Cunliffe 2001; 2008). The conditions are different in each of the British seas and, as a consequence, so is the means of navigation and the skill sets required of mariners as well as (it is often assumed) the technology of the boats. Yet given the limited number of ship and boat finds and the variety of ships and boats in use, the navigation skills and seafaring experiences of Early Medieval mariners are not well understood archaeologically. Instead we have a rather broad, possibly essentialised, picture of seafaring and navigation during this period. The variations and plurality of vessels used and the relationships between different maritime technologies and different seafaring ‘knowledges’ – locally, regionally, and nationally – require further study. What is known is drawn from a variety of iconographic, documentary and experimental archaeological sources (eg Crumlin-Pedersen 2006; 2010; Carver 1995a), as well as the wrecks and boat find record.
In the Irish Sea region it appears that seafarers maintained the boatbuilding tradition of leather stretched over a wooden frame, assumed to have originated in the Iron Age or earlier. These boats were light, easy to portage and keel-less, making rapid way to windward under sail, but needing to be paddled in any other direction. No early boat of this kind has been excavated archaeologically, but the form survives in the Welsh coracle. Severin’s (1978) experimental craft, the Brendan, sailed successfully from Ireland to St Kilda and Iceland. On the Irish Sea, travel appears to have been characterised by short-haul journeys between beaches on rocky foreshores and islands, and there are numerous inshore lakes and narrow necks of land inviting portages. Journeys offered few long runs and numerous byways. From Early Medieval Irish documents we pick up tales of navigation by island-hopping and, when the land ran out, seafarers followed the geese – by sight and sound – north in spring and south in autumn (Marcus 1980, 9–10). The natural axis is north–south, one which, as Cunliffe has emphasised, provided an ideologically unified community from the coasts of Spain and Brittany to Ireland, Wales, western Scotland and the Northern Isles (Cunliffe 2001, 558; Carver 2009) (Fig 6.4). In the western seas, navigation was aided by dead reckoning since, although visibility is famously capricious, there are a large number of islands. The littoral cultures of the Irish Sea suggest frequent interaction over a long period, although the actual movement of people has not yet been the subject of stable isotope analysis.
The earliest boats in the North Sea region are also thought to have been made of stitched and caulked hide stretched on a frame. By the 4th century, however, boats with hull-first construction in timber planks are evident. The planks were stitched together and caulked, and then the timber frame supporting the benches was lashed inside this shell. By the 7th century (at Kvalsund and Sutton Hoo) the hull was fashioned from overlapping planks fastened by iron rivets, but even in the Viking- period ship from Oseberg, the frame was still lashed to the hull. Nydam did not have a step for a mast, but Kvalsund and later Viking ships had a massive seating amidships, in which a mast could be set and the ship rigged for sail. All the boats were steered using a steer-board, essentially an oar bound with roots to a wooden boss on the right hand (starboard) side (see Crumlin-Pedersen 2010, chapters 2–4). The North Sea was fiercer than the Irish Sea, with longer hauls, where areas of protected water take the form of long inlets – firths in the north, estuaries in the south – which we presume attracted and canalised deep-water traffic. The winds are variable but appear to constitute a home-blowing system in favour of Scandinavia (Carver 1990) (Fig 6.5). There are numerous folklore references to mariners following fish, geese, and the mother swell and, when near land, of listening to the characteristic noise of the sea breaking on a particular piece of coast by putting one’s ear to the gunnel. Assumptions may also be made about Early Medieval use of the sun and stars to gauge latitude, because it is clear that the peoples of Britain were in contact throughout the Roman period with mariners from the Mediterranean where such navigation was routine. The use of a compass in the North Sea remains uncertain (Thirslund 2007).
The principal preserved ship finds of the period, the Nydam, Kvalsund and Oseberg, provide an iconic succession, forming the basis of a projected evolution of ship technology: boats that were rowed, boats that may have had a sail and boats that did have a sail (Crumlin-Pedersen and Trakadas 2003; Crumlin-Pedersen 1997a, 18–20; Marcus 1980, 35). However, despite the popular notion that the sail was invented by the Vikings, sails would have been seen round Britain since the Iron Age at least and certainly during the Roman period. The 4th-century Roman Blackfriars I had a sail and McGrail (1995) has proposed a whole succession of ‘Romano-Celtic’ boats which were flat-bottomed cargo carriers with sails plying the Channel in the early 1st millennium AD. In fact any boat can be sailed in a following wind but only in one direction. The word for sail (segh*), and by implication the technology of sailing, existed in Celtic and West Germanic languages before the Anglo-Saxons were at sea in the 4th century (Their 2003; Sayers 2004). Although the basic idea of erecting a sail was not challenging, the key step was learning to use steering and rigging so that the ship could go in other directions than windward. The Edda, a replica Viking ship based on the Oseberg burial-ship, capsized while attempting to tack in 1988 (see Carver 1995a for further discussion).
The ability to tack was a key factor in the social use of the sea. If a ship could not make to the wind, then it required a large complement of rowers to move it, which meant the ship was then full of crew. If it could make to the wind, even a little, then a small crew could take a large ship, with cargo, across the sea. The ability to tack was significant, therefore, in the development of small groups of entrepreneurs or traders and the rise of cargoes and trade. Tilley (1994) makes the same point in his discussion of sailing in the Mediterranean and recent discussion about when this was achieved in the North Sea favours the 9th to 11th centuries. Crumlin-Pedersen suggests that large Nordic cargo ships carrying bulk cargoes were plying the waters of the North and Baltic Seas at least from the 11th century and points out that the Romano- Celtic cargo vessels were, of course, much earlier (Crumlin-Pedersen 2000; McGrail 1995). However, since we have so few vessels this debate remains open, although we can suggest that, while sail was always a possibility from the 4th century, the opening up of the oceans to long-distance cargo-carrying by sailing to windward was probably, in the main, a contribution of Viking seamanship.
This emphasis on the large ships has tended to obscure the roles of numerous short journeys in small boats. We have seen some of these in burials of the 6th and 7th centuries, as at Slusegaard and Snape, where they appear as shell structures about 3m long, probably of bark, and later as faerings (four-oared vessels). Some of these were found with a larger ship, the Gokstad, implying a role as ‘dinghies’ to make landfalls in shallow water (Seal 2003). It is legitimate to imagine that the rivers, lakes, and estuaries were thronging with these personal craft, small and light enough to be carried by their crew when the water ran out. However, in general, the use of water-transport in the waters around Early Medieval Britain depends too heavily on imagination. We need direct evidence from more boats, of both inland and seagoing type, to understand how they were handled, landed, ported, and sailed.
Nevertheless, it is clear that Early Medieval seafarers could travel from one point to another across the sea and, at 15–20mph (13–17 knots), did it rather faster and carrying heavier loads than was possible for terrestrial transport. As such, coastal journeys may well have been the norm for English, Frisian, and Scandinavian seafarers (see Crumlin- Pedersen 2010, chapter 5).
In general sailing times were not prohibitive. Jarrow is two hours from Wearmouth and it would have been possible to reach Budle Bay, landfall for Lindisfarne and Yeavering, in a day from Wearmouth. From the Humber to the Forth, the coast of greater Northumbria could be sailed in a yacht in 32 hours or about three days. There is no need to conceive Jarrow and Lindisfarne as significantly separated by geography: Northumbria was a kingdom that could be unified by boat. At the same time, there is no prima facie reason to deny the feasibility of direct crossings out of sight of land – the so-called blue water crossings. If the weather was reasonable the crossing could be fast by comparison with land journeys (Carver 1990). Fair winds in the right direction were mostly to be encountered in summer, and thus we can expect that blue water voyages were largely seasonal. However, these observations about the ease of passage through the seaways are largely conjectural since, unlike Scandinavia, the coasts of Britain have not been subjected to intensive experimental voyages.
With a combination of ships, boats, blue water crossings, and manoeuvres in estuaries and tidal creeks, it is likely that the Anglo-Saxon water-world was a busy place, full of local and overseas traffic. However, when we examine the terrestrial evidence for the movement of people and goods across the sea we find that it is specific and temporary rather than general and continuous. Although the potential existed for sea travel all round the island, it appears it did not actually happen: there were preferred routes and distinctive attitudes that changed with time. Other factors, therefore, may have been at work and the imperatives for change must lie elsewhere in the realm of ideas and politics with its alignments and alliances.
In Early Medieval research three concerns in particular have focused attention on the shape and character of the coastline and of maritime spaces: (1) continental contact during the migration period; (2) North Sea traffic and maritime exploitation at the time of the wics; and (3) seafaring in the Viking Age. All three themes incorporate questions and assumptions about the likely routes of maritime connectivity and seasonal variation in traffic, the location of nodal points or articulations in the transportation network (eg harbourages, transhipment points, isthmuses), and the cultural phenomena which accompanied these engagements (Westerdahl 1994). Cross-cutting these questions are issues about seafaring capabilities and boatbuilding technology.
There were a number of imperatives persuading people to cross the sea – migration, invasion, trade, enslavement, and religious mission – and, currently, archaeologists have difficulty telling the difference between them. The debate about Anglo-Saxon immigration continues. Some accede that large numbers of Germans arrived on the east coast of Britain in the 5th century, as implied by Bede’s narrative and the similarity of grave goods either side of the North Sea (Cunliffe 2001, 454; 2008, 419), while others suggest the Britons realigned themselves for political or ideological reasons with their neighbours (Lucy 2000; Hills 2003). A further group sees a few Saxons invading first the land and then the gene pool, to create a DNA descendancy related to that of northern Europe, without having to invoke large numbers crossing the sea (Thomas et al 2006).
We could try to escape this pendulum swinging between single causes by looking at the sea rather than the land. If sea travel were feasible and frequent in the 1st millennium AD then the natural targets of research are not territories at all, but maritime spaces (Carver 1990; Crumlin-Pedersen 1991b; Westerdahl 1991; 2006). The western, eastern and southern seas that border Britain were social arenas themselves with their own agendas and historical trajectories. We can see this, for example, in the finds of imported Mediterranean red-slip ware and amphorae of the 6th century, which illuminate a route up the Irish Sea (Cunliffe 2001, 481) (Fig 6.7). It was succeeded in the 7th century by imports from Aquitaine – but still following the same western seaway. Technically, a ship with a cargo of Mediterranean pottery might have ‘turned right’ at the Scilly Isles and appeared in London or York as the Romans did before them, but they did not – this pottery does not appear in Anglo-Saxon England in any quantities that could allow us to believe in a supply (see Vince 1990, 7, 11; Watson et al 2001, 55, for sherds of c 400–500 AD Mediterranean wine amphorae found at Billingsgate). This pottery was part of a unifying project, and where pots could go, so could people.
The significance of this is that it may alter the ‘missionary’ narrative applied to the cultural, religious, and technological connections of Irish and Welsh communities during the early period. If the seaway was operating, then ‘Irishmen’ and ‘Britons’ might have visited the Mediterranean whilst Mediterranean peoples would have been visitors to the courts of Connaught, Powys or Dal Riada. Similarly, Campbell (2001) argues that the west of Scotland was not invaded by a rush of Irish, bringing Irish kingship, Columba, and Christianity. The Irish and the western Scots were simply the same maritime people in contact with each other since the Bronze Age or before. The course of history is therefore determined not by a migration, but by the ideas of the indigenous people, stimulated by travel, visitors, and imported red plates.
In the eastern sea, materials such as glass were travelling across the sea with a certain pattern between the east coast of Britain, western Scandinavia, and the Rhineland. There was also a maritime system operating in the east, leaving Britain as a land of two halves with their backs to each other. In this part of Europe this period was an age of maritime communities not migrations, in which the Scots or Irish on the one hand and the Frisians and Angles on the other were building confederations connected by trade, intermarriage, and belief. As Hills has long insisted and Loveluck is showing anew, the 6th century was a period of multiple exchanges between centres all along the North Sea coast and in the Danish archipelago. Only in the 8th century does the axis of exchange shift to cross- Channel, exemplified by the distribution of sceattas. If this point is valid, than we need to account for the fact that we ended up with England, Wales, and Ireland, rather than a Northern Irish Sea or a southern North Sea kingdom.
According to a recent collection of papers entitled The 6th Century, the changes in territorial allegiance and the upsurge of maritime traffic were attributed largely to the rise of the Merovingian kingdom, although it is less clear what caused the rise itself (Näsman 1998; Wickham 1998). Whilst this appears to underestimate the vigour of sea travel in the 5th and 6th centuries, it can be accepted that, by the year 600, the combination of Frankish ambition, the use of the Roman Empire as a model, and the Christian missions were provoking the formation of land- based territories. In Britain these were normally Iron Age and Roman territories redefined, Kent and East Anglia, for example, being successors to Roman civitates (Carver 2011). Both Kent and East Anglia were shortly to acquire Christian leaders, taxation and the wics.
As the English retrenched from their membership of the maritime community, there was a brief and unusual flowering of ship burials in East Anglia (Fig 6.8), for example at Sutton Hoo, Snape, and Caister. There was no clear tradition of ship burial in Britain, so we must look to other imperatives and contexts to help explain why, at this time, an investment in burial ships in East Anglia was thought to be desirable (Carver 1995b). Among the more convincing explanations is the role of the ship in a wider shared cosmology, which iconographic studies suggest might go back to the Bronze Age (see Bradley 2006). Recently, Henderson (2007, 299–300) has shown that Atlantic peoples invested in monuments on seaward promontories from the Neolithic well into the Medieval period. Therefore, these ship burials may have indicated a decision to reify a set of ideas that were already present in the common mind but did not need monumentalising until the appropriate moment. In the light of what was to come as a facet of the Christian kingdoms, namely the tight control of the wics, it is even possible to see in the ship burials of East Anglia a farewell to the freedom of the seas and a long metaphysical relationship with the ocean.
The construction of the wics in the 7th and 8th centuries (eg Lundenwic, Ipswich, Hamwic and Eoforwic) represents a significant social transformation, in which certain places are targeted for travel, with, we must assume, a consequent reduction in casual exchanges off creeks and at beach-markets. Though the method of loading and unloading was still tidal, making use of a river beach like the Strand at London, the object was to increase revenue. This can be implied by features such as the ordered street plan at Hamwic, the provision of storehouses at London, the possible foreign cantonments at Ipswich, and the provision of cuts of meat – as if to a garrison – at York.
The success of the venture in increasing cargo in the 9th century is conventionally signalled by the transfer, by Alfred, of the landing point in London from the Strand into the old Roman city of London. There may also be an ideological reason for this, as there was in the creation of the burhs, which emulated the network of Roman towns. A move back into the Roman capital may have indicated realignment with the ideologies and social structures of Rome as perceived by the ruling elite. There were also practical advantages; the use of a refurbished Roman dock meant that cargoes could be landed whatever the state of the tide. This new landing strategy implies that heavier vessels were plying the English seas, although few examples have been found this side of the North Sea. Following Crumlin- Pedersen (2000; 2010) we can see the merchants of Alfred’s time advancing towards large deep-water ocean-going vessels, whose masters had begun to face the challenge of sailing near the wind with a square sail, presumably making use of a massive keel and the dead weight of the hull. However, at the same time the 10th century in England represents another peculiar archaeological hiatus: the apparent lack of international imports at a time of outstanding wealth. The first London waterfront is dated by dendrochronology to the late 10th century, yet Vince (1994, 114) found that before c 1000 the numbers of imported sherds ‘could be counted on the fingers of one hand’. He argues that London could not have been active in international trade between AD 886 and 1000, suggesting ‘that the inland towns of southern England mainly came into existence as forts in the 9th century, developed local marketing roles in the 10th and early 11th century and only later became part of the network for distributing goods to the coast in one direction
and circulating imports inland in the other’ (Vince 1994, 114; see also Astill 2000; 2006). There may, therefore, be an international trade network to find off the coasts of 10th-century England which involved English as well as Norse entrepreneurs and brought exotic goods to land in unexpected places.
Vessels capable of long journeys from England to north Germany, Denmark, and Scandinavia certainly existed, as we learn from the voyages of Ohthere (Ottar) and Wulfstan. The journeys recounted by the Norwegian Ohthere to King Alfred showed the viable routeways of the fur trade, the principal route being the North Way, ie the coast of Norway (Bately and Englert 2007). Ohthere’s ship has been judged by modern maritime experts to have resembled that found at Gokstad (ibid, 115). It is interesting to compare these journeys, with their informed comments on the peoples of the North, the Finnas and the Beormas, to the rather different perception of maritime space revealed by the unique contemporary map that has survived as BL Cotton Tiberius B V f.56v (Fig 6.9). Here we have a strange concoction which does not seem to belong to the world of the well-informed navigators that Ohthere and Wulstan knew (Hill 1981, 2–3). The map is thought to derive from a Roman original copied in the 9th century and modified in the 11th century to reflect Archbishop Sigeric’s journey to Rome in AD 990 via Pavia, Verona, and Lucca (Barber 2006, 4–8). It refers to biblical cosmology, showing Noah’s Ark, the crossing of the Red Sea, and nine of the twelve tribes of Israel. Its geography reflects that of Orosius and may have even shared a scriptorium with the production of the Old English version of the Orosius within King Alfred’s ambit at Winchester (Bately 2007, 21). These appear to represent the changing interests of the English intelligensia.
Drawing on analogies from virtual reality, Foys suggests that the Early Medieval mappa mundi is best understood as a datascape, ‘a cartographic product that need not have correspondence with any real place on earth, but rather with imaginary places and circumstances made to seem real enough by an appeal to aspects of visual perception’ (Foys 2009, 120; see also O’Donnell 2009, 475–6). This was not a map for navigators, but was an expression of cosmology. Perhaps the most interesting aspect of this for our discussion was the apparent loss of connection with the Scandinavian seascape, now replaced with the imagined tribes and wonders of the ‘East’. Although, as Foys points out, there was some recognition of Scandinavia, the coasts of the Channel, the North Sea and the Baltic are now mainly hidden in a fog of ignorance, which is in contrast to the earlier descriptions of Ohthere and Wulstan, not to mention a presumed knowledge of the Anglo-Saxon homelands and nearly 200 years of Viking voyages in the three English seas and the north Atlantic (Marcus 1980, 41). Scandinavians appear to have been deliberately excluded from the new Anglo-Saxon world view. Away from the court and the cloisters, no doubt merchants and fishermen still routinely risked their lives, but the ideology of the new elites appears to have domesticated the insular sea space and made of it a literary conceit.
Maritime networks of the Early Medieval period were shaped by population pressure, trade ambition, political necessity, and ideological competition and we have an established ‘narrative’ of these large-scale changes. Yet there remains significant archaeological research potential to clarify and nuance this historical narrative of changing maritime communities and networks. Archaeology’s greatest contribution to this debate is probably the development of stable isotope and biomolecular techniques to map the movement of peoples and artefacts (especially organic artefacts) across the sea. However, there is a need to apply these methods systematically to the three ‘seas’ of Britain. There is also much to be gained from theoretical studies, anthropological and archaeological, on how to distinguish trade from migration and other mechanisms by which ideas are exchanged, resulting in similar artefacts and practices appearing on different shores.
Maritime communities and identities have been touched upon several times in the discussion above. Section 6.4 highlights the idea of shifting, broad maritime networks and, in some sense, maritime communities. However, it is Section 6.2.1 that high- lights the research potential of looking more closely at the archaeological characteristics of English maritime communities and exploring their nature and even identities – in more detail (see Loveluck 2012; and Technical Appendix 5).
If we remove the idea of the paramount role of coastal emporia in the control of socially embedded exchange in coastal zones, as the data begin to suggest that we should, then it becomes necessary to re-evaluate the nature of the merchant and artisan communities that lived permanently or periodically at the emporia, between the mid- to late 7th and late 9th centuries. The past emphasis on their subordinate role to royal authority and landed aristocracies has resulted in a lack of attention paid to the archaeological characteristics of the people who lived in the emporia communities. Yet, there are striking traits (see Loveluck 2012). Section 6.2.1 highlights how the distinctiveness of the populations of the coast was maintained through maritime connections. Loveluck highlights the potential of detailed analysis of material culture in this context. He questions, more specifically, whether the greater quantity of continental pottery in Fishergate (compared to coastal settlements) in the mid-8th century suggests a concentration of foreign seafarers in York; or whether concentrations of Ipswich ware on the Humber from the early 9th century reflect different mariners operating around the Humber, or differential choice on behalf of foreign seafarers in terms of what to trade (2007; 2012).
During the 8th and 9th centuries weapons were relatively abundant amongst the artisan and trading tenements at both Fishergate, York, and at Hamwic-Southampton, as was evidence of riding gear, which may suggest the ability to move around certain land routes quickly, in addition to maritime and river routes (Rogers 1993, 1428–32; Loader et al 2005, 53–79). Furthermore, in the refuse pits associated with the artisans and traders, imported glass vessel fragments of the finest quality, sometimes with reticella trails, were excavated (again at Fishergate, York, and over a thousand fragments from Hamwic-Southampton) (Hunter and Heyworth 1998; Rogers 1993). It would appear, therefore, that a significant number of merchant and artisan households had access to the material culture of warfare, mobility, and luxury drinking normally associated with the highest secular aristocratic households at their rural estate centres, like Flixborough in the hinterland of the Humber, and Portchester Castle, in the hinterland of Southampton (Loveluck 2007; Cunliffe 1976). What set rural aristocrats apart from the merchant and artisan populations of the emporia was not their use of different items of portable wealth; rather, the highest rural elites were marked out by their control of the resources of agricultural territories and, especially, rituals of dominance over landscapes and coastal seascapes: activities such as hunting, wildfowling and targeting of specific feast species, such as cranes and dolphins in the case of the 7th–8th centuries and the 10th century at Flixborough (Loveluck 2007). In contrast, the roles of the artisan and seafaring communities of emporia were defined by a much greater use of coinage, a broader usage of imported commodities in their everyday lives, and a greater ethnic diversity (see Technical Appendix 5 for further discussion).
Clearly, there is potential for this kind of in-depth approach to the material expression of difference and identity in the Early Medieval coastal zone, perhaps in a way which is not yet possible in terms of a specific mariner/seafaring identity (as identified in other chapters). It would be of considerable value, therefore, to consider these aspects of maritime settlements and coastal communities in more detail in future research.
It is possible to conceive of the English coasts as presenting a series of navigational challenges that demanded detailed local knowledge of tides and sandbanks – the story of St Wilfrid’s ship blown off course and stranded off the Sussex coast in AD 666 and threatened by pagan wreckers until rescued by the tide comes to mind (VW chapter 13). Suitable landing places and beaches today may well have been inaccessible in the past due to extensive salt- marshes, although former tidal channels could have given alternative access points. Coastal routes should perhaps be seen, therefore, as discontinuous and patched into a web of different transportation possibilities, some of which facilitated exchange and trade with landward interiors and major socio-political entities, whilst others connected the inhabitants of coastal margins (sometimes difficult to access from landward directions) with wider networks, which reinforced both the liminality and connectivity of coastal populations (Loveluck 2012 and see below).
Use of the inland and coastal waters was influenced by attitudes as well as the economic and political agendas of our Early Medieval forebears (see 6.4.3 above). Sources from both eastern (broadly speaking Anglo-Saxon writings) and western Britain and Ireland (mostly Welsh and Irish Saints’ Lives) display links with the sea, especially in relation to travel on the part of ecclesiastics. Notably, however, the latter seem to have travelled via existing maritime networks and infrastructures, hence some of the Early Medieval written evidence can provide at least a ‘misty window’ through which Early Medieval maritime-oriented societies can be glimpsed.
The Anglo-Saxon writers were clerics, usually writing from monasteries, and their works express a duality of view in relation to the sea, coastal margins, and mariners. For example, works such as the 8th-century Life of St Guthlac presented the marshland fens and east coast of England (and their occupants) as ‘liminal’, on the ‘edge’ of the inhabited world, as desolate wastelands, and the beginning of the realm of demons (Felix 87, [trans] Colgrave 1956; Coates 1998, 58). This liminal view of the edge of land and its role in religious polemic, where saints battled demons, has clear echoes in the heroic poem Beowulf. The perception of a desolate waste between land and sea as presented in Beowulf could reflect a generally held elite view of the low-lying wet margins of eastern England, from the 7th to 8th centuries, albeit expressed through the filter of a Christian cleric who committed the poem to writing ([trans] Heaney 1999, x–xi). Above all, those who described the coastal margins of eastern England as liminal wastes wrote from the perspective of land- holding authorities who judged value on the basis of potential for arable cultivation (Loveluck and Tys 2006, 162).
The second representation of the watery edge of England also comes from clerics, but in this the connectivity provided by the coast, and more particularly coastal ports, is stressed. Again, this connectivity is reflected in passages from Beowulf (lines 161–300, Heaney 1999, 8–11), and in Bede’s Ecclesiastical History and Altfrid’s Life of Liudger. The latter works provide famous descriptions of key port centres, housing transitory or permanent merchant communities of foreigners, often Frisians, notably in London and York.2 These ports were gateways to, and meeting points with, those from foreign lands, and also peaceful venues for interaction between the Christian and pagan worlds – Frisia was largely pagan in Bede’s day, despite the activities of the Anglo-Saxon missionary, Willibrord, from the 690s AD (Parsons 1996, 30–48). Indeed, the extent of maritime connectivity between eastern England and Frisia probably encouraged the Anglo-Saxon missions, although with Frankish assent. From the end of the 8th century, however, we also see the presentation by churchmen, such as Alcuin, of the seaways as conveyors of death and destruction, primarily as a result of raiding or organised invasion by pagan ‘northmen’ from Scandinavia (Alcuin, Ep. No. 20). Nevertheless, despite the seaways conveying danger, Rose (2007, 1–3) has recently observed that a very significant proportion of surviving Anglo-Saxon poetry, predominantly written down in the 10th and 11th centuries, demonstrates an intimate link between the sea, seaborne travel, and the Anglo-Saxon mentality.
The connectivity and freedom of movement provided by the seaways is further stressed in sources relating to Scandinavians in England between the 8th and 11th centuries, whether in the context of trading, raiding or organised campaigns of seaborne conquest. The travels of Ohthere (see above) provide an indication of the ports of call for one Norwegian chieftain-come-merchant, ranging from the trading centres of Skiringsaal (probably Kaupang, Vestfold, Norway) and Hedeby (Haithabu, now in Schleswig- Holstein, Germany), to the North Sea and Channel coasts of England (Bately and Englert 2007). Entries in the Anglo-Saxon Chronicle and other sources provide abundant evidence of Scandinavian seaborne warfare, from the attacks at Lindisfarne and Portland on the North Sea and Channel coasts in the late 8th century to the more organised raids and campaigns of conquest during the 9th to 11th centuries around the coasts of England, Wales, Ireland, and Scotland (ASC, [trans] Swanton 1996; Redknap 2000 et al).
The duality of liminality and connectivity on the part of coastal dwellers and seafarers glimpsed in the textual sources is abundantly reflected in the growing archaeological signatures of Early Medieval coastal societies in the maritime regions around the English coast discussed here. Exactly what constituted ‘coastal’, however, depends on one’s perception of the ‘edge of land’. In the Early Medieval period, the ‘edge of land’ included marshland landscapes with their islands and tidal creek systems situated between land and sea (Westerdahl 2000, 15–17). The balance between the liminal and linking roles of the coast, its waterways and societies, also changed significantly during the course of the Early Medieval centuries, associated with transformation in the roles of ports/towns and new socio-political circumstances (Loveluck 2012).