by Chris Fowler
Even optimally recovered samples must be interpreted—these data do not ‘speak for themselves’. We must always remember that Neolithic farmers did not deliberately create an archaeological record for our benefit. The animal bones are the discarded waste products from successful human utilization of animal products. Plant remains may survive accidentally—the accidental burning of a cereal store—or incidentally—the discard of waste products (weeds, chaff, nutshell) into a fire. Samples of 1,000 animal bones and 1,000 charred plant fragments tell very different kinds of story: the animal bones probably accumulated over a period of time, as individual animals were slaughtered and consumed, whilst the plant fragments may have been burnt in an episode taking only moments. The animal bones from a site therefore present an averaged picture of the economy, whilst numerous plant samples from the same site each present a snapshot of economic activity.
The way this affects interpretation is shown by the British Neolithic samples in Fig. 22.1. The animal assemblages are remarkably consistent, suggesting that the animal economy was cattle-based in southern England at least. The predominance of pig on late Neolithic sites is a major change resulting from a different mode of animal consumption (Serjeantson 2011; Rowley-Conwy and Owen 2011). The plant remains are however much more variable. The late Neolithic samples from Scord of Brouster comprise a ‘house floor’ sample, almost entirely edible barley, perhaps an accident during food preparation, and a ‘hearth’ sample, predominantly inedible weeds and chaff. Cereals need processing to separate them from these waste products (Hillman 1981). The final fine sieving removes small weed seeds and chaff fragments. The ‘hearth’ sample is probably this kind of waste, disposed of into the fire by which the barley processor sat.
FIG. 22.1. Economic remains from selected British Neolithic sites. (a) Histogram showing relative frequency of the main food mammals at three early Neolithic sites. (b) Pie charts showing relative frequency of charred plant remains. Bones: Windmill Hill calculated from Grigson (1999, Tables 145.1–3), summing pre-bank and early Neolithic contents, excluding ribs and antler. Hambledon Hill main enclosure ditch from Legge (2008, Tables 8.6 and 8.7), assuming 3.3% of cattle are wild (cf. 1 of the 30 metacarpals plotted in fig. 8.3) and 5.3% of pigs were wild (cf. 1 of the 19 measurements listed in Tables 8.33 and 8.34). Etton summed from Armour-Chelu (1998, fiche tables 75 and 58–60). Plant remains: Scord of Brouster from Milles (1986, fiche tables 27–28); ‘house floor’ is total of samples 79-82, ‘hearth’ is total of samples 56–58, Yarnton from Robinson (2000, Table 8.1), Lismore Fields 1 from Jones (in Jones and Rowley-Conwy 2007, Table 23.1), Tankardstown from Monk (1988, 186–187).
Wild pigs and cattle were present in Neolithic Europe, and separating wild from domestic can be problematic. In both species, wild were larger than domestic, but sexual dimorphism in cattle means that Neolithic wild females are about the same size as domestic males, which can pose a challenge. Pigs are not sexually dimorphic but have other problems: they are often supposed to have been loosely herded by Neolithic farmers, so bones of feral and mixed specimens may occur. The size range in a single population is however quite well understood (Albarella and Payne 2005), and a recent survey has shown that most European Neolithic pigs were in fact close-herded and fully domestic (Rowley-Conwy et al. 2012).
Dairy products were important. Before Neolithic times, humans would seldom consume milk after weaning. To digest milk, young mammals produce the digestive enzyme lactase until they are weaned. Some humans however continue to produce lactase into adult life, particularly in northern Europe, which allows them to consume milk throughout their lives. Some zooarchaeologists have argued that this was post-Neolithic adaptation. However, the gene for continued lactase production occurs in Neolithic humans from Sweden (Malmström et al. 2008), and lipids from milk have been found in Neolithic pottery (Evershed et al. 2008). The cattle bones themselves have long been argued to show dairying (Legge 1981). The clearest zooarchaeological example is the Bronze Age site of Grime’s Graves (Fig. 22.2). In dairy herds, most male cattle are killed shortly after birth, to free the milk for human use. The adult herd thus consists largely of the breeding and lactating females. At Grime’s Graves many animals were killed very young, seen in the steep reduction in ‘per cent survival’ next to the vertical axis in Fig. 22.2a. Jaws cannot be sexed, so sex ratios among older animals are established through limb bones. Distal humerus and distal metacarpal, which fuse at 12–18 months and 24–30 months respectively, show that by these ages the herd already consisted mainly of females (Fig. 22.2b, c). The high kill in the first couple of months (Fig. 22.2a) must therefore comprise the missing males.
FIG. 22.2. Age and sex of the cattle from Grime’s Graves, showing the evidence for dairy production. Top: mortality curve based on the mandibles (data from Legge 1992, Table 6d). Middle and bottom: dimensions of fused distal humeri and distal metacarpals, with tentative sexual division (data from Legge 1992, Appendix 1). Measurements as defined by von den Driesch (1976).
THE NORTHERN AND WESTERN LBK
The rapidity of the LBK spread across the loess soils of central Europe is no longer put down to shifting cultivation. European soils do not lose their fertility after a couple of years, so there would be no need to move to a new location (Lüning 1980; Rowley-Conwy 1981). LBK people lived in permanent and quite widely dispersed settlements. Thick forest faced the colonists, inhibiting their ventures, so the LBK may have spread by boat, along the major rivers where the earliest settlements are found (Rowley-Conwy 2011). There would have been few natural clearings, although some areas along watercourses may have been clear of forest (Kreuz 2008).
Although LBK settlements were surrounded by forest, farmers made remarkably little use of its products. The animal bones are dominated by domestic animals. Figure 22.3 presents three examples spanning the LBK’s northern edge from Paris to Poland. Cattle predominate, because they cope relatively well with forest grazing. This was a new development in the European Neolithic: farmers to the south-east relied on sheep and goats as their main domesticate. LBK farmers reconfigured the animal economy for the central European forest environment. Pigs were also suited to this forest, but they were not common, often being outnumbered by sheep. This suggests that the animal economy was centred on the farming clearings themselves: the wild and domestic pigs remained of different sizes, suggesting there was little interbreeding between the two. Domestic pigs were evidently close-herded near the settlements (Rowley-Conwy et al. 2012).
FIG. 22.3. Economic remains from a few LBK sites. (a) Histogram showing relative frequency of the main food mammals at three sites in France, Germany, and Poland. (b) Pie charts showing relative frequency of charred plant remains. Bones: Armeau from Poplin (1975), Eilsleben from Döhle (1994, table 6), Zalecino from Sobocinski (1984, table 2). Plant remains: Hienheim from Bakels (1978, table 15), Bedburg-Garsdorf from Knörzer (1974, Table 1), Oldenburg-Dannau LA191 from Kroll (1981).
LBK farmers arrived with the full suite of domestic animals. Cattle were not independently domesticated in central Europe, and their DNA shows that the breed had its origins in modern Anatolia (Legge 2010). Figure 22.4 presents distal metacarpal measurements from various populations. The Danish sample of aurochs forms the metrical benchmark. The German LBK animals are considerably smaller, and at no sites are there animals intermediate in size between aurochs and domestic cattle. Pigs are more complex. The first domestic animals in Europe carried Near-Eastern mtDNA, indicating importation from that region. Later in the LBK, morphologically domestic pigs appear that carried the mtDNA lineage of European wild boar. This suggests that farmers incorporated female wild boar into their domestic pig population (Larson et al. 2007). There is zooarchaeological evidence that this happened in the Paris Basin (Tresset and Vigne 2007, Fig. 22.2).
FIG. 22.4. Histograms of cattle metacarpal distal breadth, BD as defined by von den Driesch (1976). Danish aurochs from Degerbøl and Fredskild (1970, table 11), German LBK from Müller (1964, 154) and D
öhle (1994, 186–187), Troldebjerg from Higham and Message (1968 table C), Etton from Armour-Chelu (1998, appendix 1), Hambledon Hill from Legge (2008, table 8.28), Windmill Hill from Grigson (1965, table VII, xv, 1999, appendix 1.1).
Cultivation was also modified to suit the temperate forest environment. The wide range of crops cultivated in the Balkan Neolithic dwindled to just five: einkorn wheat, emmer wheat, pea, lentil, and linseed (Kreuz 2007). It is remarkable that barley, now tolerant of poorer conditions than the wheat species, was not part of this crop suite. Two botanical assemblages are shown in Fig. 22.3. As discussed, these fully agricultural economies are represented mainly by their waste products: chaff and weed seeds. The weed seeds are characteristic of crops grown in small intensively cultivated ‘garden’ plots rather than more extensive fields (Bogaard 2004), so cultivation was probably carried out with hoe and digging stick rather than the ard. The pig may also have been useful as a quasi-tool, preparing and manuring land in advance of sowing (Rowley-Conwy 1981). It is not clear whether the sowing season continued to be the autumn, as in the Balkans (Bogaard 2004), or was changed to the spring (Kreuz and Schäfer 2011)
The LBK thus appears as a scatter of specialist ‘loess farming units’, operating a system of agriculture that was considerably modified from that of its parent cultures in the Balkans. It is remarkable how little attention the LBK farmers paid to the resources of the forests that surrounded them: their economic efforts were focused almost entirely on farming the clearings they occupied.
NEOLITHIC FARMING NORTH OF THE LBK
Along its northern border, the LBK spread stopped as abruptly as it had begun. It had reached the northern edge of the loess to which it was adapted. Well drained and fertile morainic soils lay further north in southern Scandinavia, but the loess was separated from this by a zone of poor glacial outwash sands, which formed a zone of ‘badlands’ not well suited to agriculture.
The farming standstill lasted for some 1,500 years. Mesolithic hunter-gatherer populations were extensive well to the north—the Ertebølle of southern Scandinavia—and also closer to the farmers further west—the Swifterbant of the Rhine–Waal–Maas estuarine complex and the Ijsselmeer. Fewer appear to have occupied the area between the LBK and the Ertebølle. Whether these hunter-gatherers acquired any domestic animals from the farmers has been a subject for discussion for several decades.
Some specific instances of claimed domestic animals may be mentioned. A Bos metacarpal from Rosenhof in northern Germany was originally claimed to be domestic (Nobis 1975), although there were grounds for believing it was from an aurochs (Rowley-Conwy 2003). Recently, molecular biological work has suggested that the animal was indeed wild: its diet was similar to aurochs (Noe-Nygaard et al. 2005), and its mtDNA lineage was of Near-Eastern and not native European origin (Scheu et al. 2008). Some other cases are claimed, including a few goat bones in the Netherlands (reviewed in Rowley-Conwy 2013). The goats are undated by radiocarbon, but if they really are Mesolithic they must have been acquired from, or have escaped from, the nearby farmers.
Farming finally spread northwards around 4000 BC. The Trichterbecher or trægtbæger (TRB) extended north to the latitude of Oslo and Stockholm. Its farming regime was very different from that of the LBK (Fig. 22.5). The cultivation regime was further modified: peas and lentils, fairly common in the LBK, disappear; einkorn diminishes nearly to vanishing point; and barley reappears and dominates at some sites. In the late Neolithic, cereal agriculture spread into south-western Norway, and the sample from Voll has mostly barley among the identified cereal grains (Fig. 22.5). In Denmark, the ard was present at Hanstedgård at the start of the middle Neolithic around 3300 BC (Eriksen and Madsen 1984), and at the late Neolithic site of Hjelle in south-western Norway at the time of the first evidence for cultivation in this region (Soltvedt 2000).
FIG. 22.5. Economic remains from selected continental Neolithic sites in the area north of the LBK. (a) Histogram showing relative frequency of the main food mammals from Schipluiden (Netherlands), Hüde I TRB level (Germany), Spodsbjerg and Muldbjerg I (Denmark), and Alvastra (Sweden). (b) Pie charts showing relative frequency of charred plant remains from Schipluiden (Netherlands), Oldenburg-Dannau LA 191 (Germany), Spodsbjerg (Denmark), Alvastra (Sweden), and Voll (Norway). Bones: Schipluiden from Zeiler (2006, table 22.2), Hüde I from Hübner et al. (1988, table 16), Spodsbjerg from Nyegaard (1985, table 1), Muldbjerg I from Noe-Nygaard (1995, table 6), Alvastra from During (1986, table 6). Plant remains: Schipluiden from Kubiak-Martens (2006, Appendix 19.2), Oldenburg-Dannau from Kroll (1981), Spodsbjerg from Robinson (1998, table 1), Alvastra from Göransson (1995, tables 3, 4, and 5), Voll from Soltvedt (2000, table 1).
The animal economy was also different: the LBK uniformity disappeared. Some sites have a dominance of domestic animals (Fig. 22.5), and even at Skumparberget, at the northern extremity of the TRB near Stockholm, the bones comprised only domestic animals with cattle at 82% (Bäckström 1996). But at other sites wild mammals predominate. In Denmark, Muldbjerg I and others have a large proportion of wild mammals in the early Neolithic, before c. 3300 BC. Thereafter domestic species predominate everywhere. Isotopes in human bone mostly suggest a terrestrial diet, based presumably on agriculture, from the start of the Neolithic (Richards and Koch 2001). Further south, in northern Germany, wild resources continued to be significant until much later. Hüde I (TRB layer) is dominated by wild mammals (Fig. 22.5), and Parchim continues this pattern into the late Neolithic. Some of the sites in this region have remarkably high frequencies of wild horse (Sommer et al. 2011), and many also have substantial numbers of beaver and otter, birds, and fish (reviewed in Rowley-Conwy 2013). All in all the evidence for agriculture in the ‘no man’s land’ region between the northern edge of the former LBK area and the Baltic hinterland is rather weak throughout the Neolithic.
At the end of the early Neolithic around 3300 BC, agriculture went into retreat: the northern edge of farming withdrew from near Stockholm and the island of Gotland, to the southernmost part of Sweden in the middle Neolithic. So-called ‘Pitted Ware’ hunter-fishers replaced TRB farmers in much of the central Baltic. Ancient DNA suggests that Pitted Ware people may have been of a different ethnicity to the farmers (Malmström et al. 2009). On Gotland, wild boar replaced the domestic animals of earlier times (Rowley-Conwy et al. 2012). Animal bones from Pitted Ware sites are overwhelmingly from wild animals (reviewed in Rowley-Conwy 2013), and stable isotopes in human bones reveal that Pitted Ware people throughout eastern Sweden consumed a diet very high in marine foods (Fornander et al. 2008).
BRITAIN AND IRELAND
The earliest sign of any domestic species in this region (apart from the dog) is from Ferriter’s Cove in south-western Ireland. Two domestic cattle bones have been directly dated to c. 4300 BC (Woodman et al. 1999). How they arrived there is unclear (Tresset, this volume); no continental Neolithic influence can be seen on Britain at this time, which remained firmly Mesolithic, so direct contact with possible areas of origin such as Brittany is most likely. Perhaps the bones arrived in joints of meat rather than live animals.
The Neolithic appears in Britain just before 4000 BC, in Ireland perhaps a century or two later (Whittle et al. 2011). The British Neolithic is the scene of a unique debate about the onset of farming: was the adoption gradual, with a mobile ‘Mesolithic’ lifestyle continuing through much of the period (e.g. Thomas 2008), or rapid, involving a considerable amount of immigration (e.g. Rowley-Conwy 2004)? Or is a middle position, stressing local variability, the best option (e.g. Cummings and Harris 2011)?
Animal bones from almost all Neolithic sites have an overwhelming predominance of domestic animals (Fig. 22.1). There is now general agreement that aurochs, wild boar, and deer are surprisingly rare (Legge 2010; Serjeantson 2011). The only exception is the isolated early Neolithic pit at Coneybury (the ‘Coneybury anomaly’), which contains the remains of ten domestic cattle and seven roe deer, perhaps the remains of a single feasting event (Maltby
1990). This is the only site of this kind, and the roe deer in any case produce less than 10% of the meat weight (Rowley-Conwy 2003b, fig. 5). It would certainly not be safe to generalize from it to the whole Neolithic economy. It is true that animal bone assemblages from the first couple of centuries of the Neolithic remain elusive, and those plotted in Fig. 22.1 are from ritual monuments. Assemblages from settlements, dated to the crucial couple of centuries, will be of the greatest interest. What can be said at this stage is that by the time the early Neolithic was fully established, wild species played only a peripheral role in the economy.
Apart from some finds of linseed, cultivated plants are dominated by just two species, emmer and barley, with bread wheat appearing at a few sites; barley apparently becomes more common in the late Neolithic. The importance or otherwise of the cereals has been very contentious. We saw in the methodology section that we cannot simply sum the plant remains from any site and expect that the result will present an ‘averaged’ picture of the economy in the way that the animal bones do. The contrasting samples from Scord of Brouster make this clear (Fig. 22.1). Workers elsewhere in Europe all assume that cultivated cereals were overwhelmingly predominant despite the high frequency of weeds at some sites (e.g. Bedburg-Garsdorf in Fig. 22.3). Hazelnut shell is common at many British Neolithic sites. Samples containing many weeds and hazelnut shells, interpreted at face value, have been used to suggest that wild species provided most of the plant food (e.g. Thomas 2003). But nut shell, like weed seeds, is a waste product, which may even have been used for kindling and thus commonly be charred (Jones 2000). Recognition of the complexities of plant samples suggests that cereals were predominant in most places: what we see are their waste products, discarded as useless and burnt by chance along with the nut shell (e.g. Legge 1989; Jones and Rowley-Conwy 2007).
