The Oxford Handbook of Neolithic Europe

by Chris Fowler

  These difficulties are perhaps most readily addressed in relatively small catchments or sub-catchments, where archaeological and palaeoenvironmental records can be compared to landform and sediment archives in the immediate vicinity (Hoffmann et al. 2007), and where the potential for intermediate sediment storage and reworking is greatly reduced. Such studies have reported evidence of late Neolithic valley colluviation, as well as alluvial fan and (or) floodplain alluviation linked to the onset of deforestation and localized arable cultivation for localities in Britain (e.g. Brown and Barber 1985; Evans et al. 1993; Bell 1983; French et al. 1992; Collins et al. 2006; see also review by Macklin 1999), western France (Macaire et al. 2006), loess-covered valleys in southern Germany (e.g. Kalis et al. 2003; Lang 2003; Hoffmann et al. 2007; Fuchs et al. 2010), and Poland (Klimek 2003). Neolithic catchment disturbance has also been inferred from accelerated rates of inorganic sediment accumulation in some lake sediment records, including sites in Britain and Ireland (Pennington 1978; Edwards and Whittington 2001), Germany (Zolitschka 1998), and the French Massif Central (Macaire et al. 2010).

  A broader perspective on Neolithic interactions with river environments may be obtained from countrywide reviews and comparisons of Holocene valley floor development throughout Britain (Johnstone et al. 2006; Lewin et al. 2005; Macklin et al. 2006, 2009; Brown et al. 2013), Spain (Thorndycraft and Benito 2006), Poland (Starkel et al. 2006), Germany (Hoffmann et al. 2008; Fuchs et al. 2010), and France (Arnaud-Fassetta et al. 2010) (Fig. 2.4). By exploiting the growing number of published and well-dated catchment landform and sediment records and adopting an increasingly robust approach to selecting, interpreting, and analysing 14C-dated colluvial and alluvial sequences (cf. Johnstone et al. 2006; Macklin et al. 2009), these studies indicate that the geomorphological impact of anthropogenic land-use change is seldom widely evident until the marked intensification of woodland clearance and agricultural activity from the Bronze Age and later periods. Rather, Neolithic channel and floodplain environments experienced relatively little direct human intervention, often maintaining a cover of alder-dominated woodland and wetland habitats (e.g. Knight and Howard 2004; Tipping 1998; Thorndycraft and Benito 2006) amidst meandering (e.g. Starkel 2002; Dambeck and Thiemeyer 2002) or anastomosing (Knight and Howard 2004; Brown 2008) channel systems.

  FIG. 2.4. The probability density function of alluvial radiocarbon dates for upland and lowland river catchments in Great Britain during the Neolithic.

  Adapted from Johnstone et al. (2006).

  However, countrywide and sub-continental scales of analysis show periods in the early–mid Holocene which experienced broadly synchronous phases of accelerated fluvial activity, and which have been linked to the emerging record of periodic shifts to a cooler and/or wetter climate (Figs 2.2 and 2.4). Macklin and Lewin’s (2008) synthesis of the British, Spanish, and Polish records identified four such phases in the Neolithic, centred on 7590 BP (Spain, Poland), 6790–6820 BP (Britain, Poland), 5540–5640 BP (Britain, Spain), and 4840–4860 BP (Britain, Spain, Poland). Enhanced Neolithic flooding was also evident in Poland at 8400, 6250, and 5920 BP, and in Britain at 4520 BP. In German parts of the Rhine, Danube, Weser, and Elbe catchments, Hoffmann et al. (2008) also found broadly corresponding phases of accelerated Neolithic activity centred on 7475 and 5640 BP respectively. An additional early Neolithic activity phase at 8200 BP appeared largely confined to colluvial systems in smaller catchments. The 8.2k event is also evident in French catchments as a period of enhanced frequency and/or magnitude of flooding in the middle Loire, and as increased fluvial activity in the Durant and southern Alps; a similar pattern of activity occurs c. 6300 BP, although the record from the southern Alps suggests valley floors here were incising at this time (Arnaud-Fassetta et al. 2010 and references therein).

  Both Hoffmann et al. (2008) and Arnaud-Fassetta et al. (2010) found Holocene fluvial activity phases in German and French valley floors, respectively, to show only limited correlation with those identified by Macklin and Lewin (2008) in British, Spanish, and Polish records. This is considered to reflect, at least in part, differing approaches to the classification of 14C dates and the analysis of frequency distributions, but for Arnaud-Fassetta et al. (2010) this contrast in the intensity of fluvial activity between mid-latitude European rivers and those in northern and southern Europe hints at a sub-continental tripartite division of European hydrosystems during the early–mid Holocene.

  Current research agendas focusing on multiple scales of analysis (both spatial and temporal) and the quantitative modelling of fluvial system response to environmental change will refine our understanding of these issues (e.g. Arnaud-Fassetta et al. 2010; Hoffmann et al. 2010). What is clear from current evidence is that Neolithic land-use activities were rarely sufficient to promote detectable changes to channel and floodplain environments in the middle and lower reaches of larger European catchments. However, Neolithic communities were accustomed to flood hazard and the inherent rhythms of river channel adjustments, especially with respect to meander migration and occasional channel cut-off. During phases of cooler and/or wetter climate they also witnessed a change in the frequency and magnitude of flooding, alongside a change in the rate and possibly the style of channel and floodplain development.

  CHANGING COASTLINES AND COASTAL COMMUNITIES

  Europe has a strongly maritime character, with heavily indented coastlines and many large peninsulas, offshore islands, and archipelagos. Few areas of Neolithic Europe would have been far from contact with their nearest coastline, even if that contact was an indirect one through trade or exchange over hundreds of kilometres, as evidenced by the movement of the marine bivalve Spondylus shell ornaments from the Aegean to Neolithic sites in central Europe (Chapman and Gaydarska, this volume). Coastlines also played an important role as sources of marine food, raw materials, and items of value or decoration; as a medium of communication, travel, and trade by sea; and as sources of inspiration for myth and metaphor.

  Marine environmental conditions cover an immense range, from the Arctic to the Mediterranean and from exposed Atlantic coastlines to the protected and tideless basins of the Baltic and the Black Sea. Productivity generally follows a north-west–south-east gradient. Shallow areas of continental shelf, mixing of the water column by tides and storms, and upwelling currents ensure high levels of marine fertility on north Atlantic and North Sea coastlines, and an abundance of marine mammals and fish. The extensive intertidal flats of large river estuaries and inlets support large beds of bivalve molluscs, and rocky shorelines have relatively abundant supplies of limpets and other gastropods. The Mediterranean is much less productive, with limited tidal movement and clogging of major river estuaries by rapid sediment accumulation, although all types of marine resources are available, from top predators such as monk seals and tuna fish to molluscs. Least productive is the eastern Mediterranean, where temperature gradients trap nutrients at a depth beyond the reach of photosynthesis. The Baltic and Black Sea are intermediate, with little tidal movement, but inflow of nutrients from the surrounding land.

  Although agriculture is generally regarded as the dominant mode of production, marine resources continued to be widely exploited. Palaeodietary reconstructions based on stable isotope measurements of human skeletons in parts of Britain and Denmark suggest that marine foods were ignored by some Neolithic people in coastal regions (Richards and Hedges 1999). However, the interpretation of the evidence is controversial (Bailey and Milner 2002; Milner et al. 2004, 2006; Hedges 2004; Richards and Schulting 2006), and archaeological sites show continuing exploitation of fish, sea mammals, and shellfish throughout the coastal regions of Britain, Scandinavia (Clark 1983; Lidén et al. 2004; Milner et al. 2004), south-west Europe (Boyle 2005; Milner et al. 2007), and the Mediterranean (Jacobsen 1968; Tagliacozzo 1993). Submerged Mesolithic and Neolithic fish weirs discovered in Denmark show that the Neolithic examples, as at Nekselø, extended several hundred metres out from the seashore and were larger and stro
nger than their Mesolithic counterparts (Fischer 2007). For many farming communities in coastal regions, marine resources could provide an important alternative during periods of the year when agricultural products were in short supply (Deith 1988; Milner 2001, 2002). Settlements in northerly regions beyond the range of reliable crop agriculture would always have depended on the sea for a major part of their livelihood.

  Seafaring played an important role in fishing and sea-mammal hunting, trade, and population movement. At least one pathway of agricultural dispersal into Europe followed the northern shorelines of the Mediterranean, implying seaborne movements. Occupation of Mediterranean islands and the British Isles required the use of seaworthy boats to import crops and animals and exchange raw materials, even if, as now seems likely, Neolithic colonists were not the earliest seafarers in Europe (Anderson et al. 2010). The distribution of megalithic tombs has been linked in some areas to the seasonal movements of migratory fish (Clark 1977). Dugout canoes and skin-covered frame boats were already used in the Mesolithic. The earliest timber-planked boats are recorded from the Bronze Age, but were probably also built in the Neolithic, with the sail most probably in use by the late Neolithic in the eastern Mediterranean (Broodbank 2010).

  Much of what might be learned about Neolithic coastal environments may be missing because of sea-level change (Pirazzoli, 1991). At the LGM, 20,000 years ago, the sea level was over 100m lower than present and additional territory amounting to some 40% of the current European landmass was exposed on the continental shelf. The loss of territory as sea levels rose with the melting of the ice sheets after 16,000 years had profound effects on the ecology, demography, and social geography of prehistoric Europe, offset to some extent by climatic amelioration and the opening up of new hinterlands. These changes were especially dramatic in shallow areas such as the North Sea (Coles 1998; Flemming 1998, 2004; Gaffney et al. 2009), with their biggest impacts during the late Upper Palaeolithic and Mesolithic. The eustatic (glacial meltwater) contribution to sea-level change was completed with a final sea-level rise of about 15m between 8,000 and 6,000 years ago according to global estimates from deep sea records (Lambeck 1995, 1996; Lambeck and Chappell 2001; Siddall et al. 2003), which overlaps with the early Neolithic in southern Europe.

  Additional geological processes affecting sea-level change likely affected Neolithic coastlines more widely. These processes include coastal subsidence or uplift at a local or regional scale in response to tectonic and volcanic effects, particularly in the eastern Mediterranean, and isostatic rebound or subsidence of coastlines following the melting of the ice sheets, particularly in northern Europe. Geophysical models provide estimates of crustal movement (Lambeck et al. 2006; Peltier and Luthcke 2009), but precise changes can only be established by dating local palaeoshorelines, using evidence of submerged archaeological sites, shoreline biomarkers, or sediments such as peat (Shennan and Andrews 2000; Stewart and Morhange 2009). All these sources show that changes of relative sea level continued to occur in many areas during the Neolithic, with variable impacts depending on local topography and bathymetry, and that an important part of the coastal Neolithic in many regions now lies submerged on the seabed, as for earlier periods (Benjamin et al. 2011).

  The most dramatic isostatic effects were in regions close to centres of glaciation. In Scotland and northern Scandinavia there was coastal rebound, with coastlines lifting as much as 200m in northern Norway. Around the southern rim of the North Sea and the southern Baltic, there has been ongoing subsidence, with a corresponding loss of coastal territory and settlements. In Denmark and along the Baltic shoreline of northern Germany, Mesolithic and Neolithic settlements are now submerged in several metres of water (Fischer 2004; Harff et al. 2007). Partially or totally submerged settlements and megalithic sites have been recorded on both the Atlantic and Mediterranean coastlines of France (Geddes et al. 1983; Prigent et al. 1983; Cassen et al. 2011). In the Bulgarian sector of the Black Sea, Neolithic sites submerged by tectonic subsidence have been recovered (Filipova-Marinova et al. 2011). In the eastern Mediterranean, the pre-pottery Neolithic B site of Atlit Yam, Israel, was a coastal village practising farming and fishing and is now submerged in 11m of water (Galili et al. 1993). A Neolithic site on the Aegean island of Aghios Petros, Greece, is partially submerged (Flemming 1983), and recent underwater surveys at Bova Marina on the Calabrian coastline of Italy have drawn attention to the loss of a significant increment of land during the early Neolithic (Foxhall 2005). These changes would also have modified the ecology and configuration of resources available on the local coastline, removed land of potential value for livestock and agriculture, and perhaps influenced Neolithic cosmologies and perceptions of landscape.

  More dramatic effects have been claimed in the Black Sea region by Ryan et al. (1997), who used the sedimentary record on the seafloor to infer a catastrophic flood event 7,200 years ago, supposedly resulting from the overtopping of the Bosphorus sill by sea-level rise in the Mediterranean. This in its turn is supposed to have caused widespread dislocation of low-lying settlements on the shores of the Black Sea and triggered the dispersal of farming communities into south-east Europe. However, the geological evidence and the likely human consequences are now not widely accepted. Between 20,000 and 7,200 years ago the Black Sea was a freshwater lake and the water level fluctuated through an amplitude of 90m or more in response to the variable inflow of water from the rivers to the north. However, the sedimentary record in different parts of the basin produces conflicting interpretations about the pattern and timing of these changes, and the re-connection with the Mediterranean may have been more gradual than implied by the ‘flood’ hypothesis (Yanko-Homback et al. 2007). Overall, the loss of land locally in different areas of Europe, even if not as sudden as claimed for the Black Sea or as extensive as the Mesolithic inundation of the North Sea Basin, most probably had cumulative effects that were recognized within the lifetime of individuals, and the collective memory and oral traditions of many coastal societies likely incorporated stories recalling an earlier time of more dramatic land loss. The marked concentration of megalithic tombs and monuments at the coastal extremities of Britain, in the Orkneys, the Isles of Scilly, and in many other coastal regions of Scotland, Scandinavia, and western Europe—some intended to be viewed from the sea rather than from land—attests to the powerful influence of Neolithic seascapes as an arena for day-to-day subsistence, a place of danger, a source of myth, and perhaps the ultimate resting place of the ancestors (Westerdahl 1992, 2005; Phillips 2004).

  EMPTY AND SYMBOLIC SPACES—HIGH-ALTITUDE ENVIRONMENTS AND SKYSCAPES

  Until recently it has generally been assumed that in the mountainous parts of Europe Neolithic archaeology was restricted to valley floors and the lower slopes (Bocquet 1997). Recent work in the French western Alps at Les Ecrins National Park and the Hauts Ubaye Massif has shown Neolithic activity as high as 3,000m altitude (Walsh and Richer 2006; Mocci et al. 2008; Richer 2009). High alpine grasslands or ‘meadows’ were not empty spaces and had symbolic/ritual importance (as suggested by rock art) possibly related to their utility for summer hunting (Richer 2009). The environment does not only comprise climate, soil, and vegetation, but includes aspects such as skyscape and subterranean spaces, both of which vary spatially. Monument construction is testament to a growing human interest in, and desire to record, astronomical phenomena (Hoskin, this volume), alongside many other cultural stimuli. Neolithic monuments aligned on astronomical events like midsummer and midwinter sunrise include wood or stone circles and rows, isolated megaliths, and some henges and long-barrows. These Neolithic structures appear to be geographically restricted to northern Europe and this could at least partly relate to variation in the seasonal skyscape, which is a function of latitude (i.e. seasonal variations in the setting/rising positions of the sun and moon). Whilst other factors are clearly also important, both environment and latitude must play a part in the ritualization of the external environment. In wooded areas of modera
te relief, the skyscape is only viewable in gaps or clearings and the use of distant horizon markers implies a clear line of sight from the viewing location (Brown 1997, 2001). This association of open, or cleared, areas with Neolithic monumental landscapes (or ritual complexes) appears to hold and the augmentation or manipulation of natural events may have ritual and social importance (Evans et al. 1999; Brown 2001). Studies around Stonehenge on Salisbury Plain, England, suggest partial clearance by the time both Stonehenge and Durrington Walls were being built (Allen 1995). The same is true for ritual complexes on Cranborne Chase in England (French et al. 2005; French et al. 2007) and possibly southern Brittany (Scarre 2001). A fascinating aspect is the extent to which the ritualization of natural phenomena may have been a formative part of tradition, as, for example, with the recent suggestion that the banks and ditches of cursuses monumentalized the tracks of small tornadoes through woodland (Meaden 2009). These environmental phenomena, constraints, and opportunities need to be considered in any attempt to regionalize Neolithic traditions. We also need to explore the possible effects of natural events on the perception and ideology of later Mesolithic and Neolithic peoples (Larsson 2003) in what was still a fundamentally natural vegetation cover until human activities became the dominant driver in the late Bronze Age (Odgaard and Rasmussen 2000).

  LINKING ENVIRONMENTAL CHANGE, CULTURAL TRANSFORMATIONS, AND COGNITION