J D Hill and Steven Willis with Rodrigo Pacheco-Ruiz
Other than as the boundary that has to be crossed, the sea does not feature large in much archaeological writing on the later Bronze Age and pre-Roman Iron Age. It is crossed to import metals and other objects, ideas, and people, whose presence or absence is seen to have played a key role in changing the societies, cultures, and political economies of this long time period. Yet there is relatively little direct consideration of crossing the sea or the impact of the sea on people’s lives and ideas, although there is growing evidence for the changing nature of exploitation and settlement in coastal areas. This chapter, however, argues that despite the often limited and difficult evidence, people’s use of the coast and the sea serves as a barometer for identifying and understanding wider and deep-rooted changes throughout British societies across the period.
This long period (Fig 4.1) is marked by considerable social, technological, and economic changes, major shifts that have implications for the nature, scale, and organisation of maritime and coastal activities. The bulk of the maritime and coastal evidence for this period is from the land, be it objects that have or may have travelled by sea in hoards or on sites, or evidence for the changing exploitation of resources and/or settlement on or near the coast (including a corpus of river and estuarine vessels), although, notably, study of the use and settlement of the coast has been concentrated in areas that have seen systematic wetland or coastal survey. There is considerably less direct evidence for maritime activities themselves; the period begins after finds of sewn-plank boats such as the Dover Boat and ends before the ship and boat finds and harbour works of the Roman period. While it includes significant maritime finds such as the Salcombe Bay Bronze Age material and the image of a sailing ship on a very late Iron Age coin, as this chapter stresses, direct evidence for maritime activities is slight.
The Middle Bronze Age (MBA) in Britain is often regarded as a transitional era during which we see the marked decline of funerary and ceremonial monuments which had characterised preceding millennia. At the same time settlement sites and field systems become more archaeologically visible; indeed it is an archaeology of settlements and land divisions that dominates this c 1500-year period as a whole. Alongside this, the Middle and Late Bronze Ages are characterised by the deposition of large quantities of bronze objects, individually or in hoards on land and in water. In addition, the Late Bronze Age (LBA) sees considerable evidence for the importation of metal for the production of weapons, tools, or cauldrons. In some areas there appears evidence for both a settlement hierarchy and a social hierarchy, dependent in some form or other on the control of the supply of bronze. It is the end of this tradition of depositing metalwork that defines the start of the Iron Age in c 800 BC. The Early Iron Age (EIA) in some areas is comparatively difficult to detect, with often ephemeral settlement evidence, few formal burials, little hoarding and little evidence for imported objects. The Middle Iron Age c 300 to 100/50 BC (MIA) sees the start of a rise in population, with an increased prominence of settled communities with mixed farming. The record suggests a lack of social differentiation (at least as marked by material forms), distinct regional cultural expressions and identities, and few discernible imports. The Late Iron Age c 100/50 BC onwards (LIA) sees marked social changes in some regions, with more visible burials in some areas, the adoption of coins in others, and considerable evidence for cross-Channel trade, diplomacy, and movement of peoples in some areas.
There has been relatively little direct study of aspects of maritime or coastal activities for this time period. Nonetheless, the sea has played an implicit central role in explaining change in the period, with trade and exchange with continental Europe – its presence or absence – being seen as key factors in explaining how later Bronze Age and Iron Age societies functioned. Contact across the sea to parts of continental Europe is therefore central to the grand narrative of this period of British history, which ended with a seaborne invasion. However, this grand narrative largely ignores other seas and crossings, such as contacts across the Irish Sea, with islands, and along the coast. Cunliffe’s corpus of studies has frequently considered contacts and their significance (Cunliffe 1987; 1988; 2000; Cunliffe and de Jersey 1997), while more recently Henderson has examined Atlantic contacts (Henderson 2007). Various papers by McGrail have summarised the corpus of vessels of the period and the possibilities of crossing the Channel from a seafaring point of view (see Section 4.3.1 for details). The Bronze Age material from the Salcombe Bay and Moor Sands sites in Devon, and their contexts, have been discussed by Needham and Giardino and by Yates (Needham and Giardino 2008; Yates 2010), though this and MBA material found at Langdon Bay, near Dover, warrants further interpretive work. Clark’s publication of the Dover Boat and its wider contexts is relevant to this chapter (Clark et al 2004), as is Van der Noort’s (2012) study of the North Sea. There is less published research on cultural and symbolic aspects of the sea and coasts, but see Willis 2007 and Dobney and Ervynck 2007.
The key priorities for deepening our understanding of people and the sea in this period are clear: we must expand our research and gather more evidence. Achieving these goals will require innovative ways to utilise the available evidence to offer maritime perspectives for this period of history, a process that is not the same as listing evidence for activities on, in or next to the sea, or even just focusing on boats. Any future renewed research focus must take into account two key issues. Firstly, that the considerable social and economic changes seen from the start of the MBA across Britain to the Roman conquest of southern Britain are reflected in different levels and potentially types of use of the coast and the sea. Secondly, in concentrating on the maritime and the coastal we might distort, or inflate, the economic, social, and cultural importance of both for communities at this time. We must try to consider the perspective of those societies we are studying. Thirdly, there is a pressing need to develop novel and collaborative ways to think with the current sparse evidence for later prehistory and the sea in Britain.
There is, therefore, a need to address the following broad research issues in future studies:
As with other eras, establishing and considering where the coastline was at different times during this period is fundamental. Knowledge of ‘coastal gain or loss’ is vital as it has implications for how we understand basic information such as settlement and artefact distributions. In contrast to earlier periods the degree of sea-level change and coastal erosion and evolution was less dramatic, though in some places the scale of change was marked. However, also in contrast to earlier periods, there has been very little research focus on these questions for later prehistory. Due to the variable nature of these changes (as discussed in the previous chapter), this may be best explored via regional rather than national studies. In addition, changes in winds, tides, and weather patterns require further investigation. Equally, subsequent environmental changes that affected later BA and IA coastal and estuarine environments need to be considered, such as the Witham Valley which has seen massive silting and drying out since later prehistory (Catney and Start 2003).
Areas of coastal accretion at this time include the East Anglian Fens, while submergence occurred in the Outer Hebrides and around the Scilly Isles (Angus 1997; Barber 1985; Ritchie 1966; Robinson 2007). Along eroding coasts, for example, Holderness and East Anglia, the coastline of the 1st millennium BC has been entirely lost. In parts of Norfolk and north Kent there might have been a loss of c 2km of land since the Roman period (Murphy 2009; Moody 2008, fig 18). The Thames Estuary is likely to have undergone an especially complex development of deposition and erosion (Williams and Brown 1999). Consequently, it may not be possible to produce a definitive map of the coast for this entire period. Certain areas, however, have seen effective modelling of the development of the coastline in later prehistory. These studies are to varying degrees speculative (or will be until better data are available and our understandings refined). Case studies for Thanet (Moody 2008) and Romney Marsh (Eddison 2000; Eddison et al 1998; Eddison and Green 1988; Long et al 2002), amongst others, are instructive, as they model the changes and discuss the likely archaeological correlates of these changes. More work of this kind is needed.
This section briefly reviews the evidence for settlement on the coast, the use of coastal resources, like salt and pasture, and/or those from the sea, such as fish. Inevitably much detail from specific regions or sites will be lost in this general treatment. This broad picture is, of course, subject to regional and local variation and requires significant investigation in the future.
Throughout this period farmsteads and small villages, land divisions, and larger enclosures such as hillforts increasingly dominate the terrestrial archaeological evidence. It also seems likely that seasonal movement to coastal areas for salt-making and grazing was common, even if the details and intensity change over time. Such transhumance might also have been associated with procuring other resources (for example, peat-cutting, fowling, the collection of eggs and reeds, and perhaps pottery and salt production). It would be valuable to ask if such seasonal visitations to coastal areas, rather than necessarily permanent year-round settlement, were the context into which journeys by sea were also fitted.
In the MBA and LBA different patterns of settlement close to the coast are seen in different regions. Few settlements have been excavated immediately on the coast, but there are notable exceptions such Trethellan Farm, Cornwall (Nowakowski 1991). Wetland use and exploitation is evident in eastern England from the MBAin settlement, economy, and in monument building, artefact deposition, and barrow cemetery locations, notably at Flag Fen/Fengate (Pryor 1992), the Lower Witham valley, and the Fen edge in Lincolnshire (Willis 2007, table 2; Chowne et al 2001; Field and Parker Pearson 2003; Chowne in preparation; cf The Fenland Management Project; Catney and Start 2003). In the LBA in the Thames Estuary, ring forts and other potential high-status settlements emerge, perhaps linked to the control of bronze trade with continental Europe.
Following the marine incursion of the early 1st millennium BC (see Section 4.1), the nature of human activity on the coastal and estuary margins alters in eastern and south-eastern England. The incursion may not be the sole cause of change but it coincides with a wider contemporary matrix of social, economic, and environmental changes. From the 500–300s BC the intensification of settled mixed farming regimes may have resulted in wetland habitats being socially and economically redefined. Systematic research is needed in this respect, although provisional studies such as that by the Humber Wetlands Survey, in Holderness, appear to verify such a trend (Van de Noort and Ellis 1995). In the Fens, from the Early Iron Age, sea-salt extraction occurs on the silt fens and settlements occur on the gravel fen-edge terraces and islands, but there is no settlement in the fens and marsh proper (Evans 1997; Daniel 2009). Nothing like the ‘lake- villages’ at Glastonbury and Meare in Somerset, nor the crannogs of Wales and Scotland is known here (though such sites were not estuarine).
From the Iron Age there seems to have been little or no settlement orientation to the sea. Settlement sites in England were generally located inland, away from the marine margin. This pattern has yet to be firmly verified but some points can be highlighted. Allen and Gardiner, for instance, in summarising the results of the Langstone Harbour survey, note the presence of the shrine and hillfort (Tournerbury Camp) on Hayling Island but observe that overall: ‘The Iron Age is actually poorly represented within the harbour itself ’ (Allen and Gardiner 2000, xxi and 214–20). This seems broadly paralleled in the north-east of England (Tolan-Smith 2008), though there are a few exceptions such as the settlements at Tynemouth, South Shields, and Foxrush Farm, Redcar (Jobey 1967; Hodgson et al 2001). Several settlements, mainly LIA, are known in the hinterland of Poole and Christchurch harbours (Calkin 1965; Jarvis 1992; Cunliffe 1987; Cunliffe and de Jersey 1997).
In contrast, north-western Britain has an extensive coastline in relation to its area and, given the unsuitability of much of the interior for intensive agriculture, it is not surprising that there has long been a different pattern of settlement and practice in this region. Similarly, in parts of Wales, Ireland, and the south-west peninsula a greater proportion of settlement at this time was located relatively close to the sea. Whether this reflects a greater concern to be close to the sea or is more a consequence of a lack of feasible alternatives needs to be considered. Henderson (2007) suggests expediency will have inclined peoples in northern and western areas to settle in coastal localities and hinterland margins. Geomorphology, topography, climate, latitude, and soils were key (though not exclusive) influences in the nature of food production and the siting of settlements. However, phenomenological aspects to this settlement also need to be considered; for example, Parker Pearson (Parker Pearson et al 1996; 1999) has argued that brochs were located on some western Scottish isles close to the sea for symbolic reasons.
Settlement patterns, like actual settlement forms, were often highly regional throughout this period. Equally, some areas of the English coast have seen far more intensive archaeological investigation than others. This is particularly true of wetland areas, such as the Fens or Humber that have seen EH-funded surveys. Both mean that the changing patterns of settlement use are well documented in some parts of England, but that these patterns need not correspond to those in other areas. One consequence of the concentration on wetland archaeology in recent decades is that comparatively less is known about settlement in ‘dry’ coastal areas.
Coastal promontory forts (sometimes referred to as ‘cliff castles’) require separate discussion. Some of these sites, perhaps the majority, were initiated and ‘occupied’ in this period, though often they saw subsequent episodes of activity and might sometimes enclose earlier burial monuments. They are a feature particularly of parts of the coastline of Wales (eg Pembrokeshire), Scotland (eg Dumfries and Galloway and the northern coast), and the south- west English peninsula (Cotton 1959; Griffith 1988; Murphy 2002; Nowakowski and Quinnell 2011), but examples occur elsewhere along the coast of southern Britain and in the north-east of England, as at Flamborough Head. Forts also occur adjacent to the sea, as at Seaford (East Sussex), Worlebury (Somerset), Holkham (Norfolk), and perhaps Dover (Ashbee 2005); Hengistbury Head might also be seen as a site of this type.
There has been a general absence of work at such sites in England (and Scotland), which is a significant research gap (Richard Hingley, pers comm; Hingley 1992), and Murphy notes that artefacts from such sites are infrequent finds (2002, 52), and systematic environmental sampling has not been undertaken. Crucially, generalisations about these sites are problematic. Cliff castles or promontory forts for example can be of very different sizes, as well as varying in date, biography and functions, the latter still being open to debate. Some might be seen as enclosed or more heavily defended farmsteads located on a promontory; others may have been built as refuges. Some have areas of flat ground inside the bank or rampart suitable for habitation; others enclose rocky outcrops. Notably, some sites have access to beaches and thus the sea, as with the fort at the Mull of Galloway. However, there are many other instances where there is no access for kilometres, including in locations without more recent coastal erosion, for example the forts at Earn’s Heugh, north-west of St Abb’s Head, Berwickshire (Baldwin 1989, 151). The suggestion that some were used for ceremonies and rituals needs to be tested more rigorously. These sites are numerous but inadequately understood, and as a class of coastal monument they are conspicuously under-explored, and poorly dated and characterised.
While the movement of objects, plants, animals and people by sea took place around Britain and Ireland throughout this period, it is difficult to identify where these voyages began and ended (Matthews 1999). While it is possible to identify general areas where vessels may have sailed to or from, actual ‘ports’ or ‘harbours’ remain elusive. Ships and boats in this period did not need quays, waterfronts or hard landing places and could land on beaches or in shallow coastal areas (McGrail 1990b; 1993c; 1995); nor was there a need at this time for loading and unloading equipment or infrastructure. These sites need not have been permanent settlements. Indeed, cargoes may have been loaded and consumption of particular items been ‘immediate’ on the beach itself. There is little direct evidence for port and harbour facilities, and few coastal sites can be confidently identified as ‘ports’ or formal ‘landfalls’ in any form. The few there are dominate the literature and there is certainly scope for further work (Wilkes’ (2004) research in south Devon and on potential coastal port locations on the south coast offers one useful model).
It is possible, however, on the basis of the distribution of objects on land to identify a range of broad locations that were important landfalls in the period. These locales are primarily identified on the basis of concentrations of non-British and Irish objects in excavations, surveys and as chance finds (Fig 4.2). Examples for the LIA include Meols on the Wirral (Matthews 1999; Griffiths et al 2007), Redcliff/North Ferriby on the Humber Estuary (Cunliffe 2005; Crowther et al 1990), South Ferriby in North Lincolnshire (Cunliffe 1991a), Merthyr Mawr Warren near Bridgend (Cunliffe 1991b), the Isle of Portland (Taylor 2001), Mount Batten, Plymouth (Cunliffe 1988), Camulodunum, or a site nearby, in north-east Essex (Hawkes and Hull 1947; Niblett 1985; Hawkes and Crummy 1995), East Wear Bay, Folkstone (Parfitt 2012) and, arguably, Hengistbury Head, Dorset (Cunliffe 1987; 1991b; Sharples 1990; 1991; Fitzpatrick 2001). Notably, these putative LIA port sites do not seem to have been centres for redistribution (seen as a key role of more modern ports) or what has been termed ‘gateway communities’ (Cunliffe 1991a, 194). They received imports but seemingly did not circulate them into hinterlands (eg Hengistbury Head; Cunliffe and de Jersey 1997, 29, table 2). This suggestion, that they were not genuine entrêpots, requires further study.
There is evidence from the period for deliberately constructed mooring and landing structures in rivers and lakes, such as on the waterfront at Runnymede (Needham and Longley 1981) or jetties at a number of Scottish crannogs (Dixon 1994). Such structures may have been built more frequently from the LIA at coastal sites in southern England, including for example the waterfront hard gravel surface at Hengistbury Head (Cunliffe 1987) and unusual artificial moles from Poole Harbour (Markey et al 2002; Cunliffe 1991a; 1991b; Time Team, Channel 4, 8 February, 2004).1 There is significant potential in refining fieldwork approaches in order to identify these more elusive features better, and, whilst there is a little discussion of how such landfall sites fit within the wider taskscapes (see Sharples (1990) on LIA landfall sites, such as Hengistbury Head or Glastonbury and Meare, there is considerable scope for further more holistic interpretive work.
Sea-salt extraction is one of the few examples of how Bronze and Iron Age people used a coastal resource in any scale. Salt was a commodity of great significance both economically and socially, with implications for the patterns of community life and with power and political dimensions. Intensification of production occurred through the period, reflecting broader social and economic changes inland. The nature of the industries varies regionally in the type of record that remains (eg production sites or distributions). Much progress has already been made in mapping distributions and characterising industries but the scale of evidence is large and the activity widespread, with production sites known from the English Channel to Northumberland. There is a now a large and dynamic literature on salt extraction at this time. Important general sources for some production centres include: for Lincolnshire and the Fens (Baker 1960; 1975; Simmons 1980; Healey 1999; Lane and Morris 2001), for Essex (Fawn et al 1990; Sealey 1995), and for north-east England (Willis 1999; in press; Sherlock and Vyner forthcoming).
The earliest sea-salt extraction sites in Britain date to the BA, but these early sites are usually small scale, which militates against their identification (known sites are found in Lincolnshire (Palmer-Brown 1993), the Fens (Daniel 2009), Essex (Fawn et al 1990; Wilkinson and Murphy 1995, 157), Somerset (Bell 1990) and Hampshire (Powell 2009) and a probable site in East Yorkshire (Kelly and Richardson 2008; Richardson nd)). The IA saw phases of intensification of salt extraction from sea and estuarine environments around Britain. In some places such activity has left a strong archaeological signature, either in terms of production sites (for instance, in the form of brine evaporation pans, briquetage, and mounds of burnt debris) or through
the survival of ceramic salt containers (often referred to as transport briquetage) at consumer sites. The precise location of IA production sites varies: newly discovered sites at Loftus, Yorkshire, and Berwick-upon-Tweed, Northumberland, are on cliff tops, to which brine or brine-rich sand or mud was trasnported from the shore (Sherlock and Vyner forthcoming; Proctor forthcoming); direct location on seashores is possible, but most production sites were evidently located in shallow estuarine localities; salterns now found directly on coasts such as at Ingoldmells, Lincolnshire, were probably originally on creek systems (Warren 1932; Baker 1960; 1975; Aram 1993; Robinson 1993).
There is no doubt that this was an important industry tied into the agricultural cycle and entwined in the dramatic agricultural developments of the time (Morris 1994; 2007). Salt extraction was almost certainly a summer activity (Bradley 1975) and would have been a structuring aspect of annual routines, probably combined with seasonal movement to pastures in and around saltmarshes and estuaries.
In contrast, there is little direct evidence for collection of minerals from coastlines and cliff exposures during this period. Cliff quarrying for metal ores, which could have included tin (Penhallurick 1986), jet and shale and perhaps coal, clays and stones for querns, can be implied from artefacts of the period. Doubtless beachcombing was undertaken periodically. This may have been directed to the collection of specific materials such as amber on the shores of eastern Kent and East Anglia, as well as driftwood and sea coal, together with dead seabirds and sea mammals, and seaweed (Bell 1981; Murphy 1992; Smith 1999, 335; Huntley 2000). A rotary quern factory, active from the LIA, is known at East Wear Bay, Folkestone, Kent (Kellor 1989; Parfitt 2012). Here Greensand (Folkestone Beds) was evidently hewn from the sea cliff and shaped. The sea-cliff location combined the exposure of a suitable rock with possibilities for transport by boat of these weighty items. Similarly, Roe’s work on whetstones from Maiden Castle, South Cadbury, and other sites (Laws 1991; 2000) suggests that specific exposures of stone in the Plymouth area were used for making these items.
Study of sites of the period on the coastal and estuarine margins in England and Wales have found that they tend to have been specialised sites, often perhaps seasonal or temporary, where there is an association with salt production and/or grazing (of either sheep or cattle) on saltmarsh, alluvial grassland, the inter- tidal zone, and other marginal lands. Permanent colonisation of some of these areas may have arisen towards the end of the IA when population increase meant a ‘filling-up’ of the landscape, in some areas, with settled agriculture widespread.
Wilkinson and Murphy (1995, 165) and Sealey (1995, 71; 1997, 63) amongst others, have suggested that saltmarshes on the east coast of England are likely to have supported sheep flocks on a large scale during the IA (see also Major 1982, fig 7; Wilkinson and Murphy 1995, 150). Pryor (1996) has argued that the salt and freshwater margins of the Wash were used in the LBA, with the western edge of the fens (and presumably elsewhere), showing ditch systems and enclosures interpretable as large- scale flock management features (see also Gibson and Knight 2006). Evidence from the Gwent Levels of cattle grazing of inter-tidal margins includes hundreds of hoof impressions and rectangular buildings dating to the MIA (Bell et al 2000; Murphy 2002, 55; Rippon 1996, 23–4; Nayling 2002, 111; see also Chowne et al 1986, 184; O’Sullivan 2001 and Proctor 2009, 81 for other examples). Evidence to date suggests that use of saltmarsh/ wetland grazing in this period was intensive in at least several regions, but how widely this occurred in other regions requires further specific studies.
There is very little direct evidence for exploiting fish, shellfish, seabirds or marine mammals for food or other purposes from MBA to LIA sites in England. This is in line with the very low incidences of hunted animals or gathered plants on all sites of these periods. Fish bones of all kinds are rare on IA sites and there is little evidence for the routine consumption of fish in the IA of southern and central Britain, even those by the sea or rivers (see Dobney and Ervynck 2007; Jay and Richards 2007). There are more records of seafish on some unusual LIA sites in south-east England, reflecting changes in cuisine in this area just before the Roman conquest. Dobney and Ervynck show a degree of caution in interpreting their results, noting that taphonomic, preparation or recovery factors may need to be considered (Dobney and Ervynck 2007; Van Neer and Ervynck 1993; see also Evans 2003). Compared to later period assemblages, the scarcity of fish bones is very marked. It appears that people of later prehistoric Britain did not consume fish in any routine manner, though it is possible that fish were consumed in a way that does not leave a regular trace in the record (for instance, if they were consumed on shores or if remains were processed to produce ‘fish glue’). It has been argued that fishing was apparently unnecessary in central and southern mainland Britain where populations had alternative sources of food, or that fish were not eaten for cultural reasons (Haselgrove 1989; 2001; Hill 1995a; Willis 2007). These questions warrant further interpretive consideration.
In contrast, sea fish remains are recorded from some sites in west and northerm Scotland at this time, suggesting that their inclusion in human diets may have been more regular, though moderate (Nicholson 2004; Brown and Heron 2004). Consumption appears to have been rising from the LIA; at Dun Vulan on South Uist, extensive evidence for a fishing economy was identified (Parker Pearson and Sharples 1999), while at Bu Broch, Orkney, plaice and cod were recovered. Several species recovered on Scottish sites indicate sea fishing from boats but most fishing was probably conducted from the shore. Given the environment of northern and western Scotland consuming fish might be thought economically expedient, but Sharples (pers comm) has suggested the general infrequency of fish finds may mean its consumption was associated with status. This example serves well to illustrate the need for further research into how this order of evidence reflects the actuality of fishing and fish consumption among later prehistoric people, not least because similar aspects probably pertain for the Scilly Isles, where fish bones are known from Halangy Down and Bryher (Johns et al 2004).
Other marine and coastal animals follow the same pattern as fish. The evidence for shellfish consumption, both molluscs and crustacean, is slight. However, the use of shellfish in the period has not been subject to detailed synthetic study – and current evidence raises a number of questions. Notably, LBA and IA sites lying near the coast that might be expected to produce assemblages of marine molluscs, as with fish bones, yield very few (for example, see Wymer 1986; Murphy 1986, 296; Cunliffe and Hawkins 1988, 38; and notably Hambleton and Stallibrass 2000, 155). The evidence shows a pattern of limited exploitation at sites near to the coast and estuaries, with shellfish very rarely found at inland sites (eg Evans 2003, or Brewster 1963 for the exception). In addition, edible crustacea such as crab are rarely represented on IA sites, especially outside Scotland (see Bell 1977 for the exception). A note of caution is again prudent, since firstly in the past shells were not invariably retained during fieldwork, and secondly preservation is a factor as marine shells are susceptible to hostile soil conditions. Further, when swift transport inland may not have been routine, shellfish might have been consumed at or near their collection point, leaving no archaeological trace (Cunliffe and Hawkins 1988, 36). Overall, however, the pattern appears to be one of low consumption of shellfish, in marked contrast to the subsequent Roman era (see below).
Sea mammal and seabird remains provide an equally indistinct picture. Overall, it would seem that sea mammals are occasionally found on sites of the period often alone or in low numbers (eg Armour-Chelu 1991, 146; Gebbels 1977, 279–80), while at sites in coastal areas of northern and western Scotland bones are more common, perhaps representing a targeted resource (Mulville 1999; Dawson and Levy 2005). We can only speculate as to how they were obtained, and how they were utilised and valued by later prehistoric communities, both where they were rare and where they were more abundant.
There is some variation in the frequency of seabirds at settlement sites by the coast, but on the whole they seem not to have been used as a significant resource. Where seabirds (and other wild bird species) occur at sites in southern and central Britain, larger birds are represented with disproportionate frequency which may be explained by the use of their feathers in addition to their use as a food. Where exploitation is attested, selection of species is apparent (eg sea eagle, swan, kittiwake), and a connection with cultural life is generally interpreted (Partridge 1979; Harcourt 1979; Fairhurst 1984; Coy 1984; Evans and Serjeantson 1988; Parker 1988; Serjeantson 1991; 2006; Hill 1995a; Harman 1996). In western and northern Scotland the picture differs (Serjeantson 1988), most birds represented being edible species (Cartledge and Grimbly 1999). It would appear there was a cultural interest in seabirds and a wider systematic review might be called for to verify these apparent trends.
There is virtually no primary evidence for seagoing boats or ships from the MBA, LBA or IA (Murphy 2002, 55). The only evidence for shipwrecks are the well-known assemblages of bronze objects from Dover and Salcombe. In this context any new discoveries of ships, parts of ships, possible wrecks or lost cargoes, and representations of vessels are very significant. Due to the extremely scarce evidence for later prehistoric vessels, any interpretation of such evidence needs to be approached with care.
In contrast, a relatively large number of LBA and IA boats, including logboats, have been found in rivers, lakes, and estuaries in Britain. The Brigg ‘Raft’, a sewn-plank vessel dated by C14 to 825–760 cal BC (Wright et al 2001), and the extended logboat from Hasholme dated to the MIA (dendrochronology shows that the tree was felled c 322–277 BC, and as a logboat it will have had a life of c 30–50 years (Millett and McGrail 1987)), are not regarded as suitable for the open sea, although they could have operated in calmer estuarine waters (McGrail 1990a). These two finds come from the Ancholme and Foulness valleys, tributaries of the Humber (see Fig 4.3). The Humber bank at North Ferriby is also known for its corpus of boats largely dating to before this period (Wright 1990), while upstream more modest riverine logboats are also known (McGrail 1990a).
There are few clearly identified parts of seagoing boats from the period (such as timbers, rigging or oars), although finds such as Goldcliffe and North Ferriby, specifically Ferriby-5 (McGrail 2001, 187), show the potential. A wooden punt-pole/paddle dated 1255–998 cal BC was recovered at Canewdon, Essex (Wilkinson and Murphy 1995, 155). There are also two iron anchors. One from Aberdaron, Gwynedd, has been identified as a Greco-Roman type dated to the 3rd to 1st century BC (Boon 1977a; 1977b; Cunliffe 2005, fig 17.29) on analogies to Mediterranean anchors. However, as there are few early dated anchors from north-west Europe, the applicability of a Mediterranean typology remains uncertain. The other iron anchor (with chain) is from a hoard at Bulbury Camp, Dorset, dating from the LIA (Cunnington 1884; Cunliffe 2005, fig 17.29).
Wreck sites or discarded cargoes of the MBA are known at Langdon Bay, Dover (Needham et al forthcoming), and Salcombe, Devon, whilst recent discoveries dating to the LBA again at Salcombe are interpreted as a wreck (Fig 4.5), though no vessel remains are known (Needham and Giardino 2008; Yates 2010). Claims that the LIA Llyn Cerig Bach metalwork hoard represents a wreck (Roberts 2002) are not supported by the evidence and context. However, there are some finds of LBA metalwork, along with LIA pottery and amphorae, recovered from the sea around England (Matthews 1999; Parham and Fitzpatrick forthcoming) that may represent wrecks or lost cargoes, although few are precisely located.
At the same time images and models of any boat are extremely rare from this period. This is in itself not surprising as there is little representational art of any kind before the adoption of coinage in the late 2nd and 1st centuries BC in southern England. The few representations there are of boats have been frequently discussed by maritime archaeologists and others precisely because they may fill the gap in the actual evidence for boats themselves. The same applies to the few references to north-west European vessels in Greek and Latin texts that date from before the Roman conquest of southern Britain (Caesar, De Bello Gallico III.ii.13).
The models and images include the Caergwrle bowl (Davis 2010) dating to the MBA, the EIA Roos Carr wooden boat and armed human figures, from Holderness (Fig 4.6) (Coles 1990; Giles 2009), the LIA or early Roman period Broighter gold boat from Co. Londonderry (Warner 1991), and the very few images of seagoing sailing vessels on coins. The latter include a Roman warship’s prow with a cornucopia on a coin of Verica (coin type ICC95.3428: S8), c AD 10–40, and an issue of the British king Cunobelinus also c AD 10–40 (coin type VA1989:E8) depicting a ship with sail on one face (Muckelroy et al 1978; Sealey 1997, fig 8, pl 1). The latter ship is a distinctive high-sided type that appears to be seagoing and commercial. It is not a galley and by consensus is not necessarily Roman, perhaps Gallic or British. This image highlights the issues about representation of vessels. While important evidence in its own right for an early 1st century AD sailing vessel, and often invoked in discussions of the importance of LIA cross-Channel trade, it is rarely asked why this particular image was chosen or why it only occurs on a single coin type which is, itself, very rare.
The absence of complete seagoing vessels, parts of vessels or many images of boats poses a key challenge to understanding the history of boat- building traditions in this period. Evidence seems to suggest that sewn-plank vessels of the BA stopped being made at some point in this period, while hull-first plank-built vessels fixed with iron nails of the Romano-Celtic tradition were being constructed at the end of the IA and in the Roman period. This technological tradition was certainly contemporary with hide boats, as probably were sewn-plank boats, although it is not known if the latter were exclusively a western British and Irish tradition. The end of evidence for sewn-plank boats and the beginnings of the Roman-Celtic technological tradition has received much discussion, along with the possible history of hide boats. However, with little direct evidence to fill over 800 years of boatbuilding activity it is difficult to advance interpretations or set this into a cultural and historical context. In addition, narratives about later prehistoric boats are dominated by a focus on ‘technological development’, which is itself generally characterised as a linear progression from less complex to more complex boat construction – an evolution of boat technology. These ideas are being increasingly challenged by discussions of plurality, innovation, and tradition in contemporary ethnographic studies of small boats (Blue et al forthcoming; Lundberg 2003; Ransley 2010), whilst a few more recent studies have highlighted the interpretive potential of looking at the particular cultural and historical context of individual finds (eg Giles 2009; Ransley 2002).
Along with the lack of actual evidence for boats there is also little direct discussion of the capacities of LBA and IA vessels (McGrail 1990a), or the nature and experience of seafaring itself. The sailing abilities of vessels from this period can only be inferred from earlier and later finds of vessels. What little discussion there is of the organisation of voyages in this period, probable sailing routes, the ways currents, tides, seasonal weather patterns and coastlines were used is almost completely confined to a single paper by McGrail (1983; though see McGrail 1990b; 1993c) on LIA English Channel crossings.
One of the key characteristics of later prehistory in Britain is the ebb and flow of traded materials, technologies, and ideas (Cunliffe 1987; 1991a; 1991b; 2009; McGrail 1983; 1996; Clark 2002; 2004). The MBA and LBA saw considerable movement of metals and other resources around coasts, and across the Channel and Irish Sea (Yates 2010). Evidence for similar movements in the EIA and MIA is far less visible. By the LIA, however, exchange is much more prominent and an area of extensive scholarly attention, not least as these imports and exports are taken to relate to the cultural and structural changes seen in southern and central Britain at this time. The presence, or apparent absence, of raw materials and objects from outside Britain has been seen as of fundamental importance to the social and political organisation of British communities in all periods considered here. While they also often provide important chronological evidence for terrestrial archaeology, little specific attention has been paid to actual maritime aspects of these exchanges such as distance, time, and tide. There has also been very little focus on trade and exchange around the coast, between islands and across the Irish Sea.
While terrestrial perspectives on ‘international trade’ can be faulted for failing to consider the maritime, it is equally true that many discussions of cross-Channel trade and seafaring from maritime archaeologists operate from an anachronistic perspective on the nature of ‘trade’ in prehistory. The former means there is little detailed consideration to the practicalities of sea-crossing or the details of its organisation in later prehistory, while the latter means that much discussion of the maritime in later prehistory is effectively divorced from the societies undertaking the maritime activities. For both the conceptual shift from stressing maritime ‘trade’ to maritime-based exchange is potentially important. For example, there may have been relatively little commercial activity in later senses of the word. Rather, objects and raw materials could have moved by sea through webs of primarily kinship or social and political contacts which the moving objects and raw materials helped to sustain. Such networks need not presume that ‘imports’ require an equivalent return of ‘exports’.
The scale and pattern of trade and exchange between Britain and other parts of the Continent, including Ireland, varied across the period. However, a detailed synopsis of items traded is not attempted here since this would be a very long list; neither are all distribution models for the period documented. In the MBA and LBA bronze objects, scrap metal, and ingots of copper and tin were evidently moving across the seas in very large numbers, together with some gold, as seemingly attested by the recent Salcombe finds (Fig 4.5; Yates 2010; Needham et al forthcoming). The exact patterns of this trade and the contacts across the Channel, and how these fitted within larger patterns of the Atlantic BA, have been well documented (Clark 2004). Equally, contacts between Ireland and Scotland are a feature of the distributions of MBA and LBA metalwork. Similarities in artefact types and even the presence of continental European building forms point to levels of contact, marriage, and movement of peoples and ideas. To what extent disruptions of this complex political economy, which in Britain and Ireland depended heavily on maritime contacts, caused the end of the ‘Bronze Age’ is still a debated subject. Even so the movement of large numbers of bronze objects, such Armorican axes, continued for a still undetermined time into what is chronologically the EIA.
In contrast to the preceding centuries, the evidence for continental European objects or other ‘imports’ is very scarce for most of the IA. This has led many to assume there was a considerable reduction of maritime contacts, although this suggestion relies on visible ‘imported’ objects as the measure of maritime contact and activity. Evidence shows that contacts continued in this period, including imports of red coral inlay in metalwork (Dent 1982; Stead 1991); the voyage of Pytheas the Greek merchant who records a circumnavigation of Britain, having set off from Marseille in 320 BC (Cunliffe 2002); and, arguably in metalwork items (Andrew Fitzpatrick, pers comm). Other measures for this continued contact include similarities in some artefacts, including the following of broad changes in fashions of objects and art found in other parts of Europe, as with the application of red coatings to some pottery types found across southern Britain, mirroring similar contemporary practice in France and Belgium. Moreover, Matthews (1999) has reminded us of the possibility that many ‘archaeological invisibles’ were being traded near and far in this period, although little long-distance exchange seems to have been carried out in British societies at this time other than of querns, salt or raw metals.
Cross-Channel and other contacts between parts of southern Britain and continental Europe in the LIA have enjoyed a high profile (Cunliffe 1987; 1991b; 2005; McGrail 1983; 1996). There is clear evidence for increasing levels of contacts, trade, and exchange across the Channel from the 2nd century BC onwards. This evidence includes a wider range of materials than in the BA, including coins, pottery, and foodstuffs from the western Mediterranean and France/Belgium, and a range of other Roman material. The presence of such objects, along with literary evidence (notably Strabo Geography IV.v.2–3; Caesar, De Bello Gallico V.i.12), points to the changing scale and the social-political importance of trade and exchange of particular types of ‘exotic’ non-British objects in parts of south-east and southern England (Fitzpatrick 2003).
This social and political importance has led to a range of studies on imported material, and to some extent excavations of recognisable port and other sites involved in these exchanges, such as Mount Batten, Hengistbury Head, and in Poole Harbour (Parfitt 2004), along with the development of theoretical perspectives and narratives of change to invoke these exchanges as either a primary cause of social and political change or a feature and measure of such change. The significance of these emerging patterns has been seen in different ways (Haselgrove 1984; Hill 2007; Fitzpatrick 1993; 2001; Willis 1994), and how this trade was organised remains unclear. Strabo’s list of commodities leaving Britain at this time identifies grain, cattle, metals, slaves, hunting dogs, and hides, and points to the importance of archaeologically invisible traded commodities (Haselgrove 1982). However, although useful, uncritical use of Strabo’s list reflects other examples of uncritical use of classical sources and the implicit assumption of a balance of trade in the past. This is a multi- faceted subject that warrants renewed, innovative research and interpretation.
Attention has mostly focused on cross-Channel exchanges. Greater recognition of coastal exchanges, or exchange across the Irish Sea and to/within island groups, is warranted. There is some evidence for such activities, although their changing patterns and scale remain unknown; examples include Irish metal objects from Britain (Gerloff 1987; Raftery 1994). Matthews (1996; 1999), building upon the work of Morris (1985), has highlighted the coastal distribution of Cheshire Very Coarse Pottery (VCP) salt transportation containers along the North Wales coast from the Dee Estuary to Cardigan Bay, implying seaborne supply. Additionally, a few sites, such as Mount Batten, Plymouth, have been interpreted as key points in coastal trade (Cunliffe 1988).
Evidence from the Outer Hebrides also highlights the range of items that might have been traded, exchanged or given as gifts around the coast and to/from islands. The broch at Dun Vulan produced badger bones (a mammal that does not occur naturally in the Outer Hebrides) and wood from Rhamnus catharticus (sea buckthorn), a shrub with possible medicinal uses that today grows only in southern and central England (Mulville 1999, 169 and 265; Taylor 1999, 190).
There has been little attention in the literature to the organisation needed for the physical movement of objects across the sea such as boat size, sailing capabilities, the organisation of voyages, and the social structures required and necessarily sustained. This area represents a significant gap in research. Notably, there is little sense of how much voyaging to and fro is actually represented in the objects seen in the record. For example, all of the pre- Roman conquest Mediterranean amphorae found in Britain, despite the importance placed on them in archaeological interpretations, would probably require no more cargo space than three Blackfriars- size ships (Marsden 1990). Are we seeing evidence for large numbers of boats and ships crossing in the LBA or IA stuffed to the gunnels with cargoes or the movement of relatively small quantities of material at any one time? Who are the crews sailing these ships – are these merchants or kin visiting relations to feast, marry, and share gifts?
A key aspect highlighted by studies of cross-Channel exchange in the LBA and IA is the strength and nature of links at different times between communities that lived on opposite parts of the Channel, beyond simply ‘maritime trade’. This is to suggest that at some times and in some places individuals, families and communities may have felt closely tied to other individuals, families and communities from which they were separated by the sea but with whom they were in (fairly) regular contact. There was almost certainly some movement of people in both directions across the Channel and southern North Sea in this period, although the extent and nature of these movements of people are difficult to assess. Such movement, which may have included whole communities, needs to be set in the wider context of population growth and expansion of settlement into areas with low densities of permanent settlement that are a distinct aspect of IA Temperate Europe. More routine isotope analysis of skeletal remains may, in the future, shed light on these possibilities. Maritime contacts may have included links of trade and kinship, but perhaps also aggressive contact such as raiding, piracy or warfare.
One of strongest cases for a common social identity straddling the Channel comes from the LIA. From the late 2nd century BC to the Roman conquest there are similarities in burial forms, material culture, and other aspects of society between parts of south-east England and north-east France and Belgium. These include the adoption at a similar time not just of coinage, but of the same coins (so- called Gallo-Belgic coinage and potins) that are now known to have been made on both sides of the Channel (Haselgrove 1993). These similarities were in the past interpreted as evidence for an invasion of Belgic peoples, as mentioned by Caesar (De Bello Gallico V.i.12). Now they are seen as evidence of closely interlinked and to some extent commercially familiar and political interdependent communities on both sides of the Channel (Willis 1994). About far more than simply ‘trade’, these links were most likely fostered by the movement of groups of people and marriage partners in both directions and their articulation clearly depended on crossing the sea. Yet crossing the sea often seems incidental to archaeologists’ interpretations, while the actual lack of representations of ships and other maritime imagery on both sides of the Channel at this time is itself noteworthy.
This issue, again, has largely been addressed through cross-Channel or southern North Sea links – the traditional narrative of ‘Island Britain’. However, similar questions can be asked about linked communities separated by other bodies of water. To what extent, for example, might communities on either side of the Thames Estuary be seen as a single maritime community or heavily inter-penetrated and dependent communities? The strength and nature of links fostered by maritime communication at different times in the period need not mean communities on both sides of the water actually appear to be similar. The IA in particular is marked by often extreme ‘regionalism’, where different regions may have distinct object types, and also distinct settlement forms, burial and other social practices. For much of the IA, for instance, East Yorkshire had very different settlement forms, burials and object types from those on the other side of the Humber Estuary in Lincolnshire. The differences are such that it might initially appear that the two developed in isolation of each other. However, as Millett (1989; 1990) and others (eg Hill 1995a) have stressed, the need to maintain clear regional ident ties in this way is probably a response to the high level of contact they had with neighbours.
Studies of long-distance trade dominate the grand narrative for explaining change in this period, yet there has been little attempt to incorporate the specifics of seafaring into interpretations and models of trade and exchange. While cross-Channel trade has attracted much discussion, there has been far less attention to exchange across the Irish Sea or around the coast and there is a need to address these questions at regional and smaller, even estuarine, scales. As such, the following specific research topics emerge:
As reviewed in this chapter, the seas around England were regularly crossed from the MBA to the Roman conquest. Yet there is little physical evidence for how this happened, boats or maritime resources are not apparently seen as key symbols by later prehistoric societies, nor is there much evidence for a focus on exploiting the sea’s resources. As such, compared to contemporary societies in, say, southern Scandinavia or the eastern Mediterranean, it is hard to argue that these were ‘maritime communities’, or even ‘maritime-orientated communities’, or that they may have lived in symbolically charged ‘cultural maritime landscapes’ to similar degrees.
However, there is growing evidence for the possible religious and cultural resonances of the sea, or at least of things of the sea, at this time. This includes ritual locales in coastal locations and the manipulation of things from the sea in ritual contexts. Yet again, this evidence needs to be placed in the context of other such evidence from different BA, IA and Early Medieval societies to gain a better sense of the scale and nature of these phenomena. Moreover, because of the relative lack of visible evidence for these cultural practices, there has been little research in this area to date.
Middle and Late Bronze Age and Iron Age societies in Britain were aniconic. They left hardly any representations of people, animals or things, such as houses, tools or boats, in two or three dimensions. In contrast to, say, rock art in Scandinavia, there is an absence of images of boats, making it hard to understand how, if at all, boats were deployed as symbols and ‘things to think with’ in British societies. However, in a context where there are virtually no representations of things, the fact that there are two or three models of boats (Broighter, Roos Carr, and the Caergwrle bowl) from this period but no representations or models of wagons or chariots, despite the clear ideological importance horse-drawn vehicles had for these societies, might hint at the symbolic and cultural importance boats could have had. There are also a small number of very late Iron Age coins that have images depicting the Roman god Neptune or his attribute, the trident. These copy Roman prototypes and how these images were comprehended by peoples at the time is uncertain (Creighton 2000; Williams 2002).
That fish, shellfish, seabirds, and marine mammals were rarely consumed as foods or utilised in other ways has been noted in Section 4.2.6. Whether this represented a clear prohibition or just reflects a lack of interest has been discussed in the literature. For some, not eating freshwater or seawater fish demonstrates a clear cultural prohibition, grounded in religious ideas (Dobney and Ervynck 2007; Hill 1995b). Yet if fish were taboo, this applied to all fish and foods that lived in water, not specifically foods from the sea. Certainly, eating seafish and shellfish in some very late Iron Age communities in southern England was bound up with the active creation of new identities through how and what people ate, but those identities appear more about being exotic or Roman-like than specifically connected to the sea.
The question of whether the sea or any water was avoided or symbolically charged is also of relevance to the occasional presence of marine or coastal animals in ritual or structured deposits. In later prehistoric Britain wild animal bones are very rare on sites, but often any such evidence comes from structured or ritual deposits (Hill 1995b). There are a few examples where these include marine animals. For example, at the IA settlement at Slonk Hill, West Sussex (by the estuary of the River Adur), a large quantity of marine mollusc shells, mostly mussels, were placed across the bottom of an empty storage pit before an adult male human body was placed in the pit (Hartridge 1978), whilst IA burials from Knowe of Skea, Orkney, had pockets of shells in association (Moore and Wilson 2005). In addition, the claw of an edible crab from Bishopstone, East Sussex, probably came from a structured deposit (Bell 1977). These are all sites close to the sea, but there is evidence that some seabirds, specifically kittiwakes, were exchanged or moved inland in the MIA. At both Danebury and Gussage All Saints, bones from kittiwakes were found, potentially deliberately deposited as complete feathered wings (Hill 1995b).
Other clear evidence for representing the sea or aspects of the sea in ritual contexts is rare in these periods. Other than the probably deliberate, ritual deposition of the model boats mentioned above, there are no boat burials, nor deliberate deposits similar to Hjortspring in Denmark, nor obvious parts of boats in ritual deposits. The Caergwrle bowl probably comes from a wetland votive deposit, but its context is lost (Green et al 1980). The Roos Carr figures are another likely deliberate deposit in a wet part of the landscape, whilst the Broighter model comes from a hoard of gold objects (Warner 1991) and the iron anchor from Bulbury, Dorset, is part of a large hoard of iron objects, where it is probably the deposition of iron objects, not the anchor specifically, that was the focus of this likely ritual deposit (Cunliffe 2005).
In ideology and practice later prehistoric communities evidently had a complex, developed relationship with freshwater contexts (Fitzpatrick 1984; Bradley 1990; Hedeager 1992; Willis 1997; Buxton 1994). Rivers, lakes, and bogs were appropriate locations for the deliberate deposition of objects from the Mesolithic. Often seen as ritual deposits, from the MBA onwards weapons were placed in some rivers in England. From the LBA onwards cauldrons were also placed in lakes, bogs, marshes, and some rivers. There is evidence for human remains from rivers, such as skulls from the Thames (Bradley and Gordon 1988; Knüsel and Carr 1995; Bradley 2002, 56). As such, fresh water appears to have been an appropriate place for ritual offerings and at times even for the dead. The complex and changing meanings and symbolism behind these deposits are difficult to decode, but in general fresh water seems often to have been a place to communicate with other worlds (Bradley 2002; Green 1989; O’Sullivan 2001; Willis 1997). The unanswered question is whether this clear symbolic, ritual, and religious focus on water extended to the sea?
In various cultures the sea is linked with the dead, and fish with the underworld (Bradley 2002, 12). If a proportion of the dead were committed to British waters at this time, relations with the sea may have involved complex ideological aspects. For example, sea-salt extraction was practised and so if there were social prohibitions they appear negotiable in that respect. Hingley has pointed out (pers comm) that sea salt was a product of a transformation before use, and so perhaps this alteration was important in enabling (legitimating) its consumption (Hingley 1997). Perhaps one could take the salt but not the fish. Alternatively the situation may simply have been that the people of late prehistory wanted salt from the sea but, in an uncomplicated way, and were little interested in fish. If bodies of water were conceived of as a point of passage from one world to the other or a means of communication with the cosmological realm it may not have been the bodies of water themselves that were venerated, rather that they were a means to the sacred. In such a scenario, making salt need not be symbolically charged in the same way as eating fish may have been.
If freshwater finds of metal and other items of later prehistoric date are often interpreted as votive deposits, did people also deliberately deposit similar objects in the sea? There certainly are similar deposits in wetlands close to coasts and estuaries, although there has been less research on this if they come from areas that were regularly inundated by the tide. Notable examples of probable votive deposits close to the coast include the Broighter hoard (Warner 1991), while O’Sullivan notes likely votive items of LBA date from the Fergus and Shannon estuaries (2001, 127–8), and the Dagenham idol and the bronze figurine attributed to Aust-on- Severn, might constitute votive items deposited on coastal margins (though the latter may be Roman) (Drury 1980, 53; Ellis 1900). Again, the question to ask about such finds is whether these are deposits in (fresh) watery places incidentally close to the sea, or if their location close to the boundary between land and sea is of more importance. This transitional zone is often culturally and ritually charged in different societies (Mack 2011). IA coin finds from the foreshore in some areas, such as East Yorkshire (May 1992), may be votive deposits in this ‘charged’ zone, from a time period when there is also evidence for potential shrines in coastal locations (see Section 4.5.4). A significant number of IA coins have been found at the seashore in southern England (Haselgrove 1987), which raises similar questions; alternatively, like those found in Yorkshire, they might have been eroded from terrestrial, though coastal, deposits.
To date, there are no definitely identified deliberate deposits of BA or IA metal work from the sea. Did later prehistoric people throw objects from some cliff castles or place them in the sea from boats? There is little evidence for these practices, but very few later prehistoric finds have been recovered from the sea. There are small numbers of BA metal objects from the sea, which, along with finds such as those from Salcombe, are usually interpreted as wrecks or lost cargoes. Whether some of these might be deliberate deposits is hard to demonstrate. We can fill this vacuum of evidence with assumptions that the sea was a ritually charged landscape using analogies from other societies. These may be right, but for this to be more than speculation we might also need a more rigorous way to use such analogies, in the wider context of all the evidence for how people used the sea and its resources.
Around the coast of southern Britain, there are a number of IA shrines and Roman temples that lie in close proximity to the sea. These are probably of the LIA, but possibly with earlier origins. Whilst not constructed immediately by the sea, they seem to have been located in order that the sea be visible or adjacent. Such sites include Hayling Island, Hampshire (King 1990; King and Soffe 1991), Lancing Ring, West Sussex (Bedwin 1981), Worth, Kent (Holman 2005a; 2005b), and Heybridge, Elms Farm, Essex (Atkinson and Preston 1998), as well as less well-known candidates or those which have so far yielded only Roman evidence, such as Langford, Essex (Wallis and Waughman 1998, 227), Jordan Hill, Dorset (Lewis 1966), Brean Down, Somerset (ApSimon 1965; Bell 1990), and Lydney Park, Gloucestershire (Wheeler and Wheeler 1932); the massive LIA cliff-top enclosure at Bracquemont, near Dieppe, Yvelines, probably constituting an oppidum, had a Roman temple placed within it and so may be a parallel from the Continent (Willis 2007, 120). Most of these locations have elements in common: they lie on elevated ground, by river estuaries and/or points where rivers open into the sea. This latter aspect might have been of key importance in their siting, since such locations, on the joining of fresh water/ river and sea were significant places in other past cultures (Tilley 1991, 130–3; Willis 2007).
Shrines and sanctuaries of the pre-Roman era in Britain are of modest scale and typically represented by ephemeral features, as at Heybridge, Elms Farm, and Lancing Down (Atkinson and Preston 1998; Cunliffe 2005, 561–6). The features at Lancing Down were only encountered because works were being undertaken upon more substantive Roman remains (Bedwin 1981). In Britain, as in northern Gaul and Lower Germany, pre-Roman religious foci may be more detectable via associated material culture assemblages, particularly coin finds, rather than by other archaeological means (eg possibly Worth, Kent).
Each shrine site is likely to have been instituted in the light of specific local factors and considerations, so the degree to which one may generalise is uncertain (Willis 2007). Nonetheless, there appear to be some shared attributes in site location, likely to have held symbolic importance and their significance to local communities may be under- scored by their monumentalisation in the Roman era. However, there is, again, significant scope for important further research on these questions.
During the BA and IA different people used coast and estuary environments in different ways, but often through visiting from inland home-bases, not from permanent settlements. It is probable that these areas were not free and open to all. There may have been a recognised right for particular groups to undertake certain tasks and activities. To what degree coasts, beaches, and inter-tidal zones were ‘owned’ or controlled, if at all, is an opaque matter. Around much of Britain their economic potential is likely to have been low compared with that of the land, and so concern over ownership or control may have been limited or non-existent. Yet the importance of sea-salt extraction and pastures suggests there were exceptions (such as the Gwent and Somerset Levels), while there is considerable evidence for terrestrial landscapes being increasingly divided up and ‘owned’ in changing ways during this period (Hill 1995a). The way in which coastal and estuarine spaces were lived in and understood therefore remains unclear.
The religious and symbolic importance of fresh water is well recognised for later prehistory, but less is known about attitudes to salt water. There is some evidence for the religious importance of coastal sites, especially landfalls and the highest point where tides reach up rivers (Willis 2007). The following areas clearly require closer attention in future work: