The Oxford Handbook of Neolithic Europe

by Chris Fowler

  Equally, the middle Neolithic Pitted Ware culture of southern and eastern Sweden was distinguished by the presence of decorated pottery, polished flint axes, complex depositional practices, and ceremonial structures, and sometimes also by elaborate funerary observances (Larsson 2006, 50). Yet subsistence activities were dominated by the exploitation of marine species (seals and fish), with only limited evidence for domesticated cattle and pig (see Schulting, this volume). So although these people were ‘culturally’ Neolithic, some might prefer to define them as Mesolithic on the basis of their limited reliance on farming. In the Netherlands, the change from hunting and gathering to agriculture was a rather protracted process, with wild and domesticated resources used together in various ‘broad spectrum’ combinations over a long period (Raemaekers 1999, 25). As late as the middle of the fourth millennium BC, freshwater fish and wildfowl remained important elements of subsistence (Louwe Kooijmans 2011, 128; Schulting, this volume). If economic practice alone were to be used as a criterion for identifying the Neolithic, the point at which Dutch societies ‘became Neolithic’ would clearly be quite difficult to define. Should it be the first appearance of domesticates, or the stage at which they constituted the majority of protein consumed, or the moment when wild foods disappeared? These developments need not necessarily coincide.

  Similar problems of definition are evident in Britain. Here, cereals appear in the archaeological record shortly after the start of the Neolithic, around 3900 BC (Brown 2007, 1048). But following an initial period of enthusiasm, the representation of carbonized grain became scarcer after the thirty-seventh century BC, and (on the British mainland, if not the Scottish offshore islands) negligible after around 3300 BC (Stevens and Fuller 2012, 714). It is also possible that this decline was paralleled by a relaxation of the intensity of manuring, which was substantial at the primary Neolithic timber building at Lismore Fields, Derbyshire, but less so at the causewayed enclosure of Hambledon Hill in Dorset, a few hundred years later (Bogaard et al. 2013, 12591) (although note that the authors of this work attribute the contrast to the possibility that ‘the cereals deposited at ceremonial aggregation sites [ … ] derived from less-intensive land-use regimes than cereals at settlement sites’). Both in Britain (Stevens and Fuller 2012) and in Ireland (Whitehouse et al. 2013) the apparent slump in cultivation has been attributed to a ‘boom and bust’ on the part of pioneer agriculture. Whilst it is perfectly possible that a decline in yields after an initial period of great fertility (Dark and Gent 2001) did contribute to a shift away from horticulture, it seems unlikely that this was sufficiently catastrophic to cause a generalized collapse of population and significant social disruption. For in both regions, the period between 3700 and 3200 BC actually saw an appreciable increase in the scale of monument building: first causewayed enclosures and then cursus monuments in Britain (Whittle et al. 2011), whilst Ireland experienced the floruit of passage tombs (Bergh and Hensy 2013). An alternative argument would suggest that the gradual abandonment of cereal cultivation corresponded to the growing importance of livestock, as manifested in the evidence for large herds of cattle at causewayed enclosures such as Hambledon Hill, and Windmill Hill in Wiltshire (Rowley-Conwy and Legge, this volume; Whittle et al. 1999). As the centre of gravity of diet and economic practice shifted toward animals, the growing of domesticated crops was perhaps treated more casually, and increasingly supplemented by the gathering of wild foods, a pattern that continued into the later Neolithic (contra Rowley-Conwy and Legge, this volume). Yet there is no sense in which we would wish to argue that Britain became ‘less Neolithic’ over this period, simply on the grounds that some aspects of economic practice had become less intensive. The implication of this is that communities either were or were not Neolithic: although economic regimes were variable, there was no intermediate social identity between Mesolithic and Neolithic.

  VARIATION IN SUBSISTENCE PRACTICE

  All of these examples point to the complicated reality that although the geographical spread of cultivated plants and herded animals was in many ways the most significant development of the European Neolithic, there was no standardized pattern of ‘mixed farming’ held in common by all Neolithic societies across the continent. Nor did agriculture form a temporally changeless infrastructure from which all of the other innovations of the period could be derived. As Brown, Bailey, and Passmore (this volume) point out, the adoption of farming across Europe took the form of a patchwork of contingent developments, in which subsistence activity was repeatedly reconfigured in relation to local ecological and social conditions. But despite this, the pattern of economic change in the European Neolithic was not entirely random, and there were some directional trends that are worthy of note, as the process of Neolithization progressed from south-east to north-west. In their contribution to this volume, Bogaard and Halstead present a vivid and coherent account of the kind of Neolithic that became established in the southern Balkans in the seventh millennium BC. Here, long-lived settlements, many of which eventually developed into tell mounds, were composed of closely packed households engaged in intensive cultivation. Plant foods including cereals and pulses were the staples of the human diet, and only small numbers of animals (principally sheep and goats, but also cattle and pig) were kept. The meat of these animals was eaten infrequently, often in the form of shared meals involving a number of households, and there is little evidence for specialized dairying. Livestock were often grazed on the stubble in the small garden plots after harvest, or allowed to nibble the new shoots of sprouting plants to prevent them from growing too fast and lodging. Garden horticulture with small numbers of stock therefore represented an integrated and coherent economic system (Halstead 2011, 134). Halstead (1992, 32) presents the Greek and Bulgarian Neolithic as an instance of Marshall Sahlins’ (1972) ‘Domestic Mode of Production’, in which all of the labour and equipment necessary for production are contained within the household. Under these conditions, there is a tension between the tendency toward social fission and economic underproduction, and the need for relationships beyond the household in order to acquire marriage partners and recruit extra labour at critical junctures in the agricultural year. This perhaps explains the role of feasting and complex material culture in maintaining relations beyond the house, and provides a context for the use of anthropomorphic figurines in the negotiation and definition of domestic roles and identities (Robb, this volume; Bailey 2005).

  If the economic pattern of the first Neolithic communities in south-east Europe was based on horticulture, to which livestock were subsidiary (if tightly integrated into the strategy), animals subsequently appear to have become progressively more important. As Rowley-Conwy and Legge (this volume) note, the dominance of domestic species in faunal assemblages now increased, and cattle began to proliferate at the expense of sheep and goats. Indeed, large herds of bovids are now indicated in areas like north-west France (Tresset 2000). Although Rowley-Conwy and Legge attribute the change in the representation of species to the forested conditions of central and western Europe, cattle appear to have emerged as a pre-eminent animal for a complex set of reasons. First, whilst pigs had been locally domesticated during the Mesolithic (Krause-Kyora et al. 2013), and the pigs introduced from Asia at the start of the Neolithic appear to have gradually interbred with, and eventually been replaced by European stock, cattle of Near-Eastern origin rarely interbred with European aurochs (Tresset, this volume). In other words, the two species were treated in rather different ways, and the unadulterated bloodlines of cattle appear to have been something of a preoccupation. This reflects both their economic and social pre-eminence. At the same time, dairying seems to have gradually become much more important (Copley et al. 2003), whilst populations in northern and western Europe had probably by this time become lactose tolerant (Beja-Perreira et al. 2003; Shennan, this volume). This meant that animals, and particularly cattle, were now capable of providing a larger proportion of a community’s protein intake, in the form of milk produc
ts. But over and above this, the change of emphasis from plants toward animals was also a change from a subsistence economy to a wealth economy. For as well as providing everyday sustenance, cattle represented a form of mobile capital, which could be converted into power and influence through feasting, could be given as gifts to secure alliances, or could be lent out to create clients (Russell 2012, 322). Nonetheless, it is probable that Neolithic wealth was neither as alienable nor as personalized as that of the Bronze Age (see Kristiansen, this volume). It is arguable that whilst the horticultural economy of the Balkan Neolithic presented few attractions for Mesolithic hunter-gatherers beyond a more stable and predictable source of rather unexciting food, a wealth economy based on animal capital offered greater inducements for indigenous people to either attach themselves to existing Neolithic groups, or for whole communities to ‘become Neolithic’. A growing tension might thus have emerged between these temptations and the traditional focus of hunting societies on sharing, mutual aid, and social equality. Whilst the latter would have promoted temporary resistance to the adoption of Neolithic innovations, Mesolithic communities arguably played a greater part in the Neolithization of the Baltic and Atlantic zones than that of south-east Europe. In these northerly areas, stable, long-lived villages were rather scarcer than in southern and central Europe, and large funerary monuments seemingly supplemented their role as communal foci.

  This evidence that Neolithic subsistence practices were both spatially and temporally diverse does not entirely chime with Rowley-Conwy and Legge’s suggestion (this volume) that plant cultivation took much the same form everywhere. Quite rightly, they emphasize the taphonomic factors that affect the representation of botanical remains on archaeological sites, but they underplay one important aspect of the formation of the archaeological record: contextual variation. It has long been recognized that the composition of both carbonized and waterlogged assemblages depends upon the location from which they were retrieved (Dennell 1978), and this may be a consequence of the stages of crop processing and storage represented at a particular site, the character of deposition, and the type of site being investigated. In the case of the large timber halls that characterize the earliest phase of the British Neolithic, Rowley-Conwy and Legge mention that the structure at Yarnton produced an assemblage dominated by hazelnut shells. But in addition, the large collections of cereal grain from the halls at Claish and Warren Field in Scotland were recovered from contexts associated with the decommissioning and destruction of the buildings (Thomas 2013, 307). These contexts also contained large numbers of pottery sherds, which had apparently been deliberately placed, whilst samples taken from elsewhere within the structures contained appreciably less grain. Clearly, these results have profound implications as far as the representativeness of these assemblages is concerned, and regarding what they tell us about Neolithic economic activities. Similarly, Andersen (this volume) notes that the large cereal assemblages from the Danish causewayed enclosure of Sarup had been deliberately deposited in two ceramic jars, again raising the question of how far they can be identified as a random sample of the products of everyday farming practice.

  NEOLITHIC DISPERSALS

  Several of the chapters in this book reflect on the diversity of processes by which Neolithic entities and practices spread across the European continent. Some authorities note that aspects of the Neolithic way of life may have been integrated by hunter-gatherer societies, having been transmitted in advance of Neolithic settlement. However, it has often been observed that the acquisition of a few cattle or pots does not necessarily render a community Neolithic (e.g. Tresset and Vigne 2007). It follows from this that the Neolithic cannot be defined purely on the basis of the representation of a checklist of traits, and must be understood instead in organizational or structural terms. Müller (this volume) makes the important observation that networks circulating obsidian in the Aegean actually predated the Neolithic. Similarly, the expansion of the Bandkeramik into central Europe was prefigured by the exchange of mollusc shells and lithic raw materials over long distances into areas occupied by hunter-gatherers (Gronenborn 1998, 1998). This evidence implicitly challenges the received idea of Neolithic colonists intrepidly advancing into unknown territories. In fact, the colonization of unfamiliar landscapes is especially dangerous for agriculturalists, who would normally need to take many years to properly familiarize themselves with local soils and climate. Landscapes need to be ‘learned’ before they can be occupied, and knowledge is therefore one of the most important constraining factors on the expansion of agriculture (Meltzer 2003). It has been argued that the successful translocation of farming populations by ‘leapfrog’ colonization would have required the acquiescence and support of indigenous people, achieved through protracted contact and negotiation (Fiedel and Anthony 2003). It follows that the affairs of Mesolithic and Neolithic communities may often have been more entangled than we are accustomed to thinking (Borić 2005). Even where the expansion of the Neolithic involved the colonization of uninhabited islands, advance understanding would have needed to be acquired through scouting and visiting (Broodbank and Strasser 1991).

  Whilst some areas including Crete and the Thessalian plain have produced evidence for planned, large-scale migration by agricultural communities, this was by no means a universal pattern (see Guilaine, this volume). Malone (this volume) notes that in several parts of the Mediterranean there are indications of a contrast between developments in the coastal zone and more inland areas. In the Adriatic, southern Italy, and southern France, maritime colonists associated with the use of Impressed Wares may have established sedentary sites on the coast (Broodbank 2006; Forenbaher and Miracle 2005; Guilaine and Manen 2007). Further away from the sea, often in upland areas, caves and rockshelters sometimes suggest continuity of occupation from Mesolithic to Neolithic (Mlekuz 2005). These sites may contain pottery, and the remains of either domestic or wild animals, or both. Two possible interpretations present themselves: either coastal Neolithic groups may have been practising transhumance and hunting away from their permanent settlements, or indigenous Mesolithic communities may have acquired pottery, sheep, and goats from them. If the latter, the adoption of these new elements will have had a transformative effect on the inland groups, gradually drawing them into Neolithic social networks. In contrast, in Atlantic Portugal it has been postulated that Neolithic colonists and Mesolithic groups using the shell middens of the Alentejo, Tagus, and Sado existed in close proximity to each other for some hundreds of years. Here there is no evidence for the piecemeal adoption of Neolithic innovations, although some contact between farmers and hunters must have taken place (Zilhão 2001). In this case, long-term resistance to the changes associated with the Neolithic seems more probable.

  There is, perhaps, a disjuncture between these sorts of complex micro-scale processes, increasingly invoked in archaeological accounts of the period and involving combinations of population movement of different kinds, acculturation, intermarriage, and economic transformation, and the larger-scale and more homogeneous developments sometimes implied from population genetics and ancient DNA studies (Pluciennik 1996). One recent investigation, for instance, suggests that the modern population of Europe is made up of the descendants of ‘brown-eyed farmers’ and ‘blue-eyed hunter-gatherers’, on the basis of a limited number of ancient DNA analyses (Callaway 2014). The uncritical equivalences that are sometimes set up between genetic groupings, categories of economic practice, and linguistic entities are potentially quite problematic. ‘Farmers’ and ‘hunter-gatherers’ were probably not fundamentally different kinds of people: they simply organized themselves and acted in rather different ways. Neither kind of community is likely to have been genetically homogeneous; their diversity would probably have only increased over time.

  However, the large-scale analyses sometimes employed by archaeologists can have problems of their own. For instance, the summed probability analyses of radiocarbon dates discussed by Shennan (this volume; see a
lso Rick 1987; Williams 2012) indicate massive population increases at the start of the Neolithic in Belgium, the Netherlands and Germany. These developments are connected with the notion of a ‘Neolithic demographic transition’: the argument that new subsistence practices lead causally to sedentism and population rise (Bocquet-Appel 2008). The method uses the number of radiocarbon determinations within a given chronological interval as a proxy for the size of population. Where the behaviour of prehistoric people remained constant over time, it is reasonable to argue that an increasing number of dated samples reflect a growing number of people creating a greater number of dateable traces in the archaeological record. However, at the transition between the Mesolithic and the Neolithic, two other variables are introduced into the equation. Firstly, the nature of the archaeological record and its formation changes radically, in that many Mesolithic groups were mobile hunters who left a minimal trace on the landscape, whilst Neolithic societies dug pits, ditches, and postholes, some associated with dwellings or occupation sites and others with monumental structures, and placed their dead in graves and monuments, rather than dispersing their remains across the landscape (Conneller 2006). All of these new contexts serve as ‘traps’ for organic materials suitable for radiocarbon dating, and much less likely to survive in the ephemeral occupation sites of the Mesolithic. The consequence of this is that as much as a change in population size at the start of the Neolithic, what is being measured is an increase in archaeological visibility. Secondly, one is also measuring the behaviour of archaeologists in the present, whose criteria for selecting dating samples vary between the accepted ‘periods’ into which the past is divided. It seems possible that these methods can in some cases result in misleading assessments of the date, scale, and location of population change (see, for example, Collard et al. 2010).