Peter Bogucki

School of Engineering and Applied Science
Princeton University

The introduction of agriculture and the successful establishment of farming communities on the lowlands of north-central Europe between 5000 and 3500 B.C. (recalibrated dating) marked one of the most significant transformations of prehistoric society in this region. Many difficulties in the discussion of the establishment of agriculture in north-central Europe stem from an overemphasis on the distinction between "Mesolithic" and "Neolithic" as adaptive patterns. Such a distinction brings about the notion of a boundary between communities practicing these two strategies. It is clear that there was a "frontier" of sorts between these Neolithic groups and the local foraging peoples. Yet it was a permeable frontier, and once domesticated plants and animals became available on the lowlands of north-central Europe, a well-defined boundary between distinct social entities effectively ceased to exist. Moore (1985: 94) has characterized frontiers between sedentary farmers and mobile foragers as "a cultural mosaic of interspersed communities with varying subsistence and settlement requirements." The North European Plain between 5000 and 3500 b.c. (perhaps a bit earlier and perhaps a bit later) can be described in such terms, as a mosaic cultural landscape.

Postglacial Foraging Groups
By the early fourth millennium bc, some foraging communities in north-central Europe appear to have approached "low mobility", to use the term proposed by Bocek (1985). Examples of "low mobility" adaptations are known from many parts of the world, primarily in lacustrine, estuarine, and riverine environments, and there are many such habitats on the North European Plain with their attendant productivity and diversity. Wetland environments such as the Satruper Moor, the Dümmer basin, the Rhine/Maas delta, and perhaps the lake belts of north-central Poland, probably supported growing low-mobility hunter-gatherer populations by the middle of the sixth millennium B.C., although this hypothesis still needs further testing against the archaeological record.

First Stockherders and Farmers -- 5400-4800 B.C.

The earliest food-producing communities to appear on the North European Plain were those of the Linear Pottery culture, or Linearbandkeramik), which had also colonized the loess belt across central Europe between 5400 and 5000 B.C. (recalibrated dating). There are three main clusters of Linear Pottery settlement on the North European Plain: the Kujavy region west of Poznan and south of Torun, the area along the lower Vistula north of Torun, and along the lower Oder river south and west of Szczecin. There are vast areas in which Linear Pottery settlements have not (yet?) been found, including the Pomeranian moraine belt, the Baltic coastal plain, and the glacial outwash areas west of the Elbe, but there is the potential for considerable change in this picture. For instance, prior to 1980, only a handful of Linear Pottery sites were known from the area along the lower Vistula north of Torun. Today, close to 200 have been discovered, thanks to the interest taken in them by a local university.
The ceramics, flint tools, and ground stone tools found on the Linear Pottery sites of the North European Plain are essentially similar to those found elsewhere in east-central Europe. Unlike the large Linear Pottery sites of the loess belt, substantial longhouses have hitherto not been found in the lowlands. Instead, most of the lowland Linear Pottery sites are relatively small collections of shallow pits. The pits often have dense concentrations of refuse, however, with relatively large sherds and many reconstructable vessels. All the same, they do not appear to represent the same level of commitment to particular settlement locations as do the longhouse settlements of the loess belt.

The Neolithic Mosaic -- 4500-3900 B.C.
During the period between 4500 and 3900 B.C., it is very difficult to speak of general developments that happened uniformly across the area between Holland and Poland. Rather, a number of specific local adaptations are found which defy categorization. The most archaeologically "visible" of these are the ones with economies that involve food-production to some degree, found in areas like the Rhine-Maas delta, the Dümmer basin in NW Germany, the lakes and moors of northern Germany, and the lowlands of north-central Poland.

Lengyel Farming Communities
Settlements of the Brzesc Kujawski Group of the Lengyel culture appeared in north-central Poland around 4500 b.c. The most characteristic aspect of these settlements is their trapezoidal-plan longhouses with bedding trenches. At Brzesc Kujawski, nearly 50 houses have been identified, many of which are associated with storage pits, workshop features, and graves (Grygiel 1986). At Oslonki, over 20 houses have been excavated between 1989 and 1994. Other sites have fewer houses but also represent permanent settlements of agrarian households.
The Lengyel subsistence strategy in the Polish lowlands was based primarily on the cultivation of grain and the keeping of domestic livestock. A broad spectrum of wild resources was utilized by the inhabitants of Brzesc Kujawski, including waterfowl, fish, turtles, small mammals, and deer. The domestic pig, well-suited to the forest environment, increased markedly in its economic importance, although cattle were still predominant. The Lengyel settlement system was more complex than the Linear Pottery pattern, and again the best data are from Brzesc Kujawski and environs. In addition to the large residential base, there are a number of outlying satellite sites.

Foragers Adopt Farming in the Polish Lowlands
The Lengyel communities in the Polish lowlands disappeared about 3900 B.C., succeeded by sites of the Funnel Beaker culture. Funnel Beaker sites are spread across much of northern Poland, with a much broader distribution than the Lengyel settlements and a significantly different character. Unlike the Lengyel residential bases, which display unbroken intensive occupations of particular locations for very long periods of time, the Funnel Beaker sites are often smaller and shallower. Funnel Beaker sites in northern Poland are found in a wide range of habitats. They can be clustered or dispersed, stratified or single-occupation. It is difficult to reconstruct systems of regional exploitation, such as has been done for Brzesc Kujawski, since the identification of central residential bases and functionally differentiated outlying sites is elusive.
The Funnel Beaker occupations have a distinctly "Mesolithic" character to them in that they are often on sandy soils and have relatively few sub-surface features. A major depositional characteristic of a site like Brzesc Kujawski is the large, deep borrow pits, from which clay was extracted for wall daub. Features of this size are very rare on the Funnel Beaker sites of the Polish lowlands. Instead, there are often small pits, postholes, and sheets of midden containing numerous fragmented potsherds. Rather than the concentrated rubbish deposits that occur in the Lengyel borrow pits, the Funnel Beaker sites often have expansive, low-density distributions of refuse.

The evidence suggests that Funnel Beaker communities developed out of the local Mesolithic foraging societies, while the Lengyel inhabitants of sites like Brzesc Kujawski played a crucial, but indirect, role. The earliest Funnel Beaker sites of northern Poland appear to reflect communities with one foot in the Mesolithic past and the other in the Neolithic future.

Hüde I on the Dümmersee
The glacial lakes on the northwest German outwas plain may have crucial significance for the transition from foraging to farming in this area. One lake basin, the Dümmer, appears to have been of particular importance. Here, the Hüde I site has three occupation phases between 4900 and 3600 B.C. (recalibrated). The first phase provides no evidence of cultivation or stockherding, while the second, with ceramics similar to those of the Rössen culture, and the third, with Funnel Beaker ceramics, provide traces of an agrarian economy.
Despite the presence of domestic taxa in the faunal assemblage, it is clear that their representation is much less than that of wild species. Domesticated plants are known from both carbonized grain and imprints on vessels. Kampffmeyer (1983: 127) has interpreted Hüde I as a site devoted to the acquisition of a broad spectrum of wild animal resources to supplement an otherwise agricultural economy. The sites around the Dümersee are perhaps the tip of an iceberg, and the prehistoric settlement system in this area between 5000 and 3500 B.C. may shed considerable light on the transition from foraging to farming in northwest Germany if it is fully investigated.

Early Funnel Beaker Sites in Northern Germany
Around 3900 B.C. (recalibrated), Funnel Beaker communities represent the earliest populations with domesticated livestock and plants in northern Germany. These communities are tagged with the abbreviation "EN", for "Early Neolithic", although they have little in common with the Early Neolithic settlements of the loess belt. One group of EN settlements is found along the Baltic coast of East Germany (Nilius 1973, 1975; Nilius and Warnke 1984) and inland in the lake belts of Mecklenburg (Schuldt 1974, Nagel 1980). These sites appear to be very similar in structure to the earlest Funnel Beaker sites of the Polish lowlands. At Ralswiek, an irregularly-shaped depression 2.5 meters in each direction contained hearths and suggested a habitation location. Gristow had similar dark discolorations which contained intrusive stones and a large amount of highly-fragmented sherds. At Basedow, on a sandy island, a Neolithic settlement layer with burnt wall daub was found over an area about 20 by 20 meters. There is apparently evidence for a Mesolithic component at this site as well, although the stratification is not well defined.

The Rhine-Maas delta
The Rhine-Maas delta presents a special type of estuarine environment on the North European Plain, in which there were extensive marshes, tidal flats, and peat bogs. Sites of the Initial Delta Neolithic, located primarily in the peat zone, have been the focus of considerable research in the last 15 years (Louwe Kooijmans 1987). Of particular importance are settlements at Hazendonk and Bergschenhoek, and the complex of sites at Swifterbant.
These sites have yielded an extraordinary range of subsistence data. The faunal remains include both wild and domestic mammals, estuarine and anadromous fish, and numerous birds. At Hazendonk, large amounts of carbonized grain, chaff, and internodes have been found, while Bergschenhoek has yielded only wild fruits and nuts. The best botanical data come from Swifterbant S3, with many wild plants, barley, and emmer wheat. A discussion is taking place as to whether the cereals were grown locally or brought to the sites on the ear.

The interpretation of these sites, according to Louwe Kooijmans, suggests that Hazendonk and Bergschenhoek are short-term camps for fishing and fowling, while the Swifterbant sites appear to reflect longer-term occupations. There is caution about calling these settlements fully agricultural, but evidence for continuity from the local Mesolithic has led many to conclude that the Initial Delta Neolithic reflects the adoption of pottery, cultivation, and animal husbandry by indigenous foraging communities without, at first, major changes in other aspects of the settlement system.

Motivations for Farming
Why should the Mesolithic inhabitants of north-central Europe have adopted farming at all? One impetus to adopt certain agricultural practices and domestic animals on the part of the native foragers of the North European Plain may have come from the imbalance between growing populations on one hand and the patchily distributed resources of the lowland forests on the other. Agriculture has the net effect of creating an additional patch in the environment, one which is predictable in space and time. The ability to create such high-yield patches could have been attractive to growing local Mesolithic populations. There was also the storability of grain, which would have addressed the need for resources to sustain populations through the winter months on the North European Plain. Given what might have been a widespread pattern of local subsistence intensification among the Mesolithic populations of the North European Plain, agriculture may have presented itself at an opportune moment. Yet, if we think in terms of a rigid Mesolithic/Neolithic frontier, it is difficult to see how food production would have taken hold among the Mesolithic hunter-gatherers. Instead, we need to envision ways in which the techniques, technology, and attractiveness of food production could have passed from farmers to foragers.
Models for Interaction
Spielmann points out that "intersocietal subsistence exchange can become a viable strategy for low density hunter-gatherer societies given two conditions: (1) the presence of horticultural societies supplying regularly available carbohydrate resources, and (2) the availability of (a) small but regularly harvestable supplies of protein ... or (b) predictable, periodic large concentrations of protein resources..." (1986: 297). The complementarity between the two subsistence systems enables two separate populations to exist within a fairly small area and gives each group the best of both worlds. Hunter-gatherers have access to storable carbohydrates whose production requires a sedentary existence, while farmers have access to wild resources whose pursuit requires mobility.
It is possible that a major limitation of the subsistence options of hunter-gatherer populations on the North European Plain was the lack of storable wild carbohydrate resources. With the arrival of agriculturalists on the southern fringe of this zone, including the Lengyel settlements at Brzesc Kujawski and environs, this problem was alleviated. The foragers near these settlements would have had sources of carbohydrate which did not result from the expenditure of energy in the pursuit of low-volume wild plant foods. Moreover, the opening of the forest for agriculture would have had the effect of improving hunting, the implications of which are discussed further below. Agriculturalists would have had sources of protein which would not be derived from their herds of livestock, thus reducing the amount of energy expended on hunting and providing them with needed resources in crucial times of the year. There is thus good reason to hypothesize the existence of such mutualistic relationships in the vicinity of the Lengyel sites in north-central Poland and elsewhere on the North European Plain. Future archaeological research must test for such patterns of interaction in subsistence remains and artifact inventories.

Garden Hunting and Feral Stock

The incorporation of domestic plants into the intensively exploited habitats of the North European Plain may have had the ancillary effect of concentrating animal resources. In the faunal assemblages from some sites, the number of bones of red deer and wild pigs is quite striking. Red deer in particular would have flourished on the forest edges and in cultivated fields, and it is possible that an exploitation pattern on the model of the "garden hunting" proposed by Linares (1976) for the American tropics was practiced. In garden hunting, animals are hunted in cultivated fields and gardens as they are attracted to the growing crops. Some degree of crop loss to animal pests is considered acceptable, since the net effect is to concentrate the animals in the gardens and fields and thus reduce the expenditure in time and energy in hunting them. Moreover, the biomass of certain species increases when permitted access to cultivated crops. In garden hunting, protein from wild animals becomes a by-product of farming. One way of testing the possibility that a similar strategy was followed by some Neolithic communities would be to examine the seasonal indicators such as tooth eruption and dental annuli in the pig and deer bones to see if the animals were in fact killed during the growing season.
Davidson (1988), in writing about the introduction of food production to Spain, has drawn on the example of the colonization of Australia and the contact between foragers and fishers on one hand and farmers and herders on the other. The first things to penetrate the frontier between these two populations were animals, which had either escaped from the protection of food producing populations or had been stolen. It seems quite possible that there was some escape of domestic stock from the loose herding arrangements that probably prevailed in Neolithic communities in north-central Europe as well. We must recognize the possibility, at least, of the Mesolithic hunting of feral domestic stock. An important consideration is the type of ecosystem into which the feral stock would have escaped. In the Australian case, the environment was characterized by grasslands and savannahs, which had been further modified by fires set by the aboriginal population, and feral livestock found an inviting habitat. It is possible, however, that escaped domestic stock in Neolithic Europe found the browse in the forests to be perfectly adequate, for temperate woodlands can support grazing and browsing at low stocking rates (Adams 1975). Moreover, if some Mesolithic groups of northern Europe had already created artificial glade habitats along lakes and streams, then some escaped domestic stock may have found a quite favorable environment in which they could multiply.

Sliding Along the Continuum
Ingold (1984: 5) has noted that anthropologists have the tendency to remove people from the category of foragers if they have any attributes of agriculture or pastoralism. There are really, however, four basic categories of human groups:
1. those who subsist on uncultivated plants and wild fauna (foragers);
2. those who have a mixed subsistence economy, based partly on domestic and partly on wild resources. These can be subdivided further into two sub-categories: a.) foragers who farm (yet still closer to category 1), and b.) farmers who hunt (yet still closer to category 3 below);
3. those who gain no significant subsistence from uncultivated plants or wild fauna (agriculturalists and pastoralists).
In general, anthropologists have tended to limit the universe of human subsistence economies to those in category 1 and those in category 3. The root of the problem may be that colonization of Africa, the Americas, Australia, and the Pacific in the 18th and 19th centuries essentially "froze" many societies as either foragers or food producers depending on their circumstances at the time of contact. The point of this frozen moment was then assumed to be fixed on the unilinear scale of progress from hunting and gathering to farming and stockherding. Early anthropological studies that were based on short-term contact tended to reinforce this viewpoint. Yet all does not seem to be so simple. Long-term studies of a number of groups reveal fluctuations between foraging and farming with an annual or even longer periodicity.

There is also a prevailing belief among anthropologists that once a society heads down the road to agriculture it cannot return. In general, it is widely believed that once population growth occurs as a result of the "improved" food supply brought about by food production, the society comes to depend more and more on its crops (e.g. Cassiday 1980). Actually, there is little evidence to suggest that population growth is an automatic effect of the adoption of agriculture, and it appears that even among agricultural societies there are mechanisms that restrain population growth and fertility (see, for example, Englebrecht 1987 for a discussion of this among the Iroquois). It is entirely possible for societies at this boundary between foraging and farming to slide back and forth from one strategy to another, following one for a few years, then reverting to the other, and back again. For instance, the Agta in the Philippines, long thought to be prototypical hunter-gatherers (e.g. Peterson 1978), actually are opportunists who make use of the subsistence strategy that best suits the conditions of the moment (Griffin 1984). These conditions can be determined by the natural environment but also by the sort of interactions that a group is having with agricultural neighbors at any given moment. A similar pattern characterizes the Semang of Malaysia, described by Rambo (1985: 43) as being "opportunistic foragers" who take up agriculture "when the terms of trade are unfavorable for wild forest products and drop it when other opportunities appear more rewarding."

I believe that it is possible to view the period between about 4500 and about 3500 B.C. (recalibrated) across the North European Plain in this fashion. Different local groups of indigenous populations did different things -- and did not necessarily progress along a steady trajectory from foraging to farming. It was a patchwork, not only in space but also in time. Some groups may have been quick to try agriculture, others more resistant. Although in the long run the communities living on the North European Plain may have been caught in the web of food production, this change did not occur overnight. Rather, it is more probable that as more local populations embraced agriculture the limitations on hunter-gatherer foraging areas and information exchange modeled by Moore (1985) came into play, and only then was the feedback loop triggered, resulting in the widespread use of agriculture and animal husbandry by about 3500 B.C.