The Problem
The antiquity of dairying is a problem which has received scant archaeological attention, yet one which is crucial to the understanding of prehistoric animal husbandry in the Old World. The keeping of animals for their milk leads to an entirely different set of economic relationships from those conditioned by the carnivorous exploitation of domestic stock. Hither to it has been difficult to establish that the mill( from domestic animals was exploited earlier than the 4th millennium B.C. in the Near East and the 3rd millennium B.C. in Europe. Most of the available evidence for the earliest dairying in the Near East and Europe has been iconographic and from the interpretation of certain container forms as having been associated with milk handling (Sherratt 1981:275-82). Recently, however, data from earlier European Neolithic sites indicate an even greater antiquity for dairy production.
The modern populations of temperate Europe are unusual among the world’s peoples in that the majority of the adults can ingest raw milk without adverse side-effects. Most of the world’s adult population does not produce a sufficient amount of the enzyme lactase which is required to metabolize lactose or milk sugar. As a result, they suffer from cramps, diarrhea, and vomiting when they drink milk. The development of a tolerance for lactose through the continuation of lactase production into adulthood is thought to have been a relatively late evolutionary process, occurring sometime in the last few millennia among the peoples of northwest Europe (who subsequently spread throughout the world in the past 500 years), and Asian and African pastoralists (Simoons 1979). An alternative theory has been advanced by Cook that argues that the gene for the persistence of intestinal lactase into adult life originated in the Arabian peninsula and spread outward from there (1978).
Despite the fact that the issue of lactase production has been the focus of considerable argument, the ability of the adult inhabitants of temperate Europe to produce intestinal lactase prior to 3000 B.C. may be irrelevant to the question of prehistoric dairying. There are numerous ways to process raw milk to obviate the effects of lactose mal-absorption, generally involving the separation of milk into its components and permitting various bacterial agents to alter chemically the structure of the sugars in it. Products such as cheese and yogurt are the results of this activity. Archaeologically the problem is twofold: to demonstrate the exploitation of domestic stock for their milk as opposed to their meat, and to demonstrate the existence of artifacts used in milk-processing. There is considerable evidence for both from later Neolithic temperate Europe. I argue that dairying began somewhat earlier, based on the evidence of both the faunal remains from Early Neolithic sites in temperate Europe, and the presence of ceramic strainers or sieves on these sites, artifacts arguably associated with dairy production.
Early Neolithic Animal Bones
Tlie Linear Pottery culture (Linearbandkeramik) is an Early Neolithic culture, distributed across a wide area of temperate Europe, from the Ukraine to France and from Hungary almost to the Baltic Sea (Fig. 2). This culture appears to have represented one of the few documentable demographic expansions of a unified cultural entity in prehistoric Europe, for its house forms, stone tools,pottery, and settlement locations differ completely from those of the indigenous hunter-gatherer populations in this area (Tringhamn 1968; Hamond 1981). The earliest radio-carbon dates for Linear Pottery from Hungary faIl around 4600 b.c. (unrecalibrated—ca. 5400 B.C. recalibrated), and those from Poland and Germany begin only a century or two later (Quitta 1967). Since Linear Pottery settlements are generally found on or near the fertile loess soils of central Europe, it has been generally assumed that the cultivation of grain crops constituted the primary subsistence base of this culture and that domestic animals played a decidedly subsidiary role. Moreover, it has also been believed that the cattle, sheep, and goats kept by Linear Pottery communities were used solely for their meat (and other slaughter products such as hide and bone), and that the possible intolerance of lactose by the Neolithic peoples of Europe precluded the use of their milk (e.g. Milisauskas 1978:71; Sherratt 1981:276-77).
Linear Pottery culture faunal assemblages are almost always composed primarily of the bones of domestic cattle, with sheep/goat and pig represented in decidedly smaller proportions. Only on some East German sites do the frequencies of sheep/goat bones exceed those of cattle ( Müller 1964). Bones of wild mammals are generally rare, suggesting a relatively low degree of hunting. Traces of other wild species, such as fish, birds, and molluscs, are also rare, although their absence may be due more to the recovery methods used in the excavations than to the economic patterns of the inhabitants.
Linear Pottery faunal samples are generally small, and it is difficult to assess whether the cattle were used primarily for either meat or dairy production solely on the basis of their age distribution. In order to make a reliable judgement, the samples would have to be large enough to permit an evaluation not only of their aggregate age profiles but also of the relative proportions of males, castrates, females, and juveniles (see Stein, this issue). My very subjective assessment of the metrical data from a number of Linear Pottery sites suggests that, in most instances, the assemblages are dominated by mature females. Legge has argued that a female bias in the adult cull and a high neonatal cull, presumably of male calves, can be taken as an indication of a dairy economy (1981a).
In some areas, there does appear to have been a relatively high degree of calf-killing at Linear Pottery settlements. Willer, in his study of Linear Pottery faunal remains from 71 East German sites, found an age distribution among the cattle bones of 60.5 percent adult, 11.5 percent subadult, and 28 percent juvenile individuals. Although it is unclear what Müller’s upper limit for his juvenile category is, it appears that many individuals in this group were under six months of age, with most probably under a year old. Müller takes this to represent ‘autumn slaughter’ due to shortages in winter fodder (1964:64). Legge has pointed out that the killing of calves to free milk for human consumption is functionally similar to such ‘autumn slaughter’ in that it involves a reduction in the fodder requirements of the herd while increasing the food output for humans (1981b:180). In his words, “cattle could hardly be kept for the annual production of one little carcass,” and instead, the killing of infantile and juvenile animals is the result of their keepers’ desire to have undivided access to the mothers’ milk production.
The killing of calves may not necessarily be the wisest approach in a dairy economy. Amoroso and Jewell (1963) have pointed out that it is generally difficult to persuade lactating cattle to let down their milk without their young around (save for modern breeds of ‘improved’ livestock). In the African cultures which they studied, the practice was to keep the calf near the mother, allow it to suckle, and then remove it from the udder before exhausting the mother’s milk supply. Among cattle-herding cultures such as the Karimojong of Uganda, the calves are generally not killed, since their economic usefulness as adults outweighs the amount of milk they might consume as infants (Dyson-Hudson and Dyson-Hudson 1970:122). Legge is of the opinion that although cattle in Africa seldom surrender their young, European cattle commonly do so, and that African husbandry practices cannot be generalized to include those of prehistoric Europe (1981b:221).
The Linear Pottery Culture
The LBK culture (Linearbaudkeramik; Linear Pottery) is the earliest known Neolithic farming culture of temperate Europe, and is named after its characteristic pottery (Fig. 1). LBK sites are widely distributed from the western Ukraine to northeast France (Fig. 2), showing a marked association with areas having loess subsoil. Loess, a fine-grained yellowish material, carries a fertile and well-drained topsoil that, presumably, was the main attraction for these Neolithic farmers. Despite the wide geographical range of LBK sites, their material culture and chronology are markedly coherent. The earliest radiocarbon dates for LBK are in Hungary, around 4600 b.c. (uncalibrated; ca. 5400 B.C. recalibrated), while those from Germany and Poland are only about 100 years later (Quitta 1967). LBK material culture and settlement locations both differ markedly from those of the indigenous Mesolithic hunter-gatherers of the area (Triugham 1968; Hamond 1981).
Many features of LBK culture can be paralleled in the Neolithic period of the Balkans, which began some 1000 years earlier. For example, some of the pottery forms, the characteristic Dissection ‘shoe-last’ axe or adze, and the use of spondylus shell (imported from the Mediterranean) all strongly indicate cultural connections. Moreover, the LB farmers’ cultivated wheat and barley and their domesticated sheep and goat must have derived from the southeast, since these species are not known to have been native to temperate Europe. A fair case can be made that LBK farmers were colonizers, moving northwestward out of the Balkans into the cooler and more heavily forested regions of central Europe (though, needless to say, there are arguments over this).
One notable feature of LBK, however, is not found in the Balkan Neolithic—the very large timber ‘long-house,’ often 30 meters (about 100 feet) or so long. These are known only from their groundplans, of course: the pattern of wall-slots and post-holes dug into the subsoil to carry the timber uprights of the structure (Fig. 5). The houses occur in groups, almost all on loess soils, and are known from numerous excavations. These groups are usually called ‘villages,’ although the occasional overlap of house plans shows clearly that not all houses were contemporary. The often wide spaces between houses, and the millennia of plowing that has destroyed most stratigraphy an the sites, make it difficult or impossible to establish clearly the chronological relationships of houses in a ‘village.’
It may well be that, in many cases, neighboring houses were successive, not contemporary, and that the typical settlement was a ‘hamlet,’ or even a more or less isolated individual farmstead.
LBK sites without these famous long-houses’ are also known, however. Some of these are on the loess lands, and are often the very earliest LBK sites in a given region, predating the establishment of ‘long-house’ settlements. Others are found right off the loess, in areas where ‘long-house’ settlements never occurred at all—for example on the north European plain, where the terrain was considerably less well-drained than the loess lands, and where the soils and vegetation were substantially different. It is at such sites, far less eye-catching than the ‘long-house’ settlements, that most of the pottery sieves or strainers have been found; these sites, moreover, appear to have been occupied only seasonally, during the summer months. As this article proposes, therefore, there is a good case for interpreting them as summer camps, around which cattle were grazed and at which milk was processed. In the fall, presumably, the herders would drive their cattle back to the permanent settlements with their long-houses, carrying cheeses and perhaps other products of the summer’s activities.
Economics and Ecology
If the age and sex data from Linear Pottery faunal assemblages are still ambiguous about the extent of dairying during this period, they do provide some hints of culling patterns consistent with dairy production. There are also sound economic arguments as to why the livestock kept by Linear Pottery communities should not have been used primarily or solely for meat production. It must be remembered that Linear Pottery communities were dealing with an environment which was largely unfamiliar to them. Each new tract of loess or lowland glacial soils presented a new set of environmental advantages and disadvantages which took time to sort out (Bogucki 1979).
In a situation in which decisions were tempered by risk and uncertainty, a concentration on cattle primarily as a meat source would seem to be poor economic strategy. Since cattle require 42 to 48 months to reach their optimal meat weight, a great deal of labor and energy must be invested in each head in return for its meat yield. Not only would the stock have to be assured adequate supplies of forage and water, but they would also have to be maintained through the winter on cut fodder. If no dairy products were to be obtained from the herd, all this investment of labor, time, and energy would far outweigh the 300-400 kilograms of usable meat available from each head. Moreover, if meat was the only return expected from the cattle economy, it would have been impossible for a self-sufficient Neolithic community to increase the output from its herd quickly in response to temporary shortfalls in other subsistence resources without either maintaining an enormous reserve of surplus animals or seriously affecting the viability of the herd as a reproductive population.
Given the uniparous nature of cattle, and the potential for the loss of animals to predators and disease, Neolithic communities would have needed to be very selective in their slaughter of cattle in order to assure that sufficient breeding stock remained. If they had only regarded cattle as a “meat bank” that would offset crop failures and shortfalls, it would have been necessary to predict crop yields three or four years in advance in order to receive the maximum return on their investment.
There is an ecological rationale for Neolithic dairying as well. The process by which plants are converted to milk and meat involves a net loss of energy at each step. When humans milk their lactating stock, they place themselves at an earlier point in the conversion chain than when they slaughter the animals for their meat. As a result, a much higher proportion of the original energy input can be taken back in the form of milk, thus permitting the maintenance of a larger human population than can be returned in the form of meat. Mangold argues that “Pilch pastoralism” is the most efficient use of uncultivatible land, while “carnivorous pastoralism,” in which meat is the only desired product, is no more efficient than hunting and probably less so in the long run (1980:176).
Hand meat been the only return desired from their domestic stock, it would have made much more sense for Linear Pottery communities to concentrate on pig husbandry. Pigs are multiparous and reach a high meat weight and sexual maturity within a year of birth (Grigson 1982:298). The forested environments in which most Linear Pottery sites are found, both the floodplain forests of the loess belt or the woods of the North European Plain, would have been excellent sources of pannage. Yet pigs are consistently the rarest domesticated taxon in Linear Pottery faunal assemblages. At Linear Pottery sites in the Polish lowlands, pigs are virtually absent, and it is not until several centuries later that they appear prominently in faunal assemblages (Bogucki 1982, 1984a).
The relative proportions of cattle and pigs on Linear Pottery sites, along with the generally low degree of exploitation of wild herbivores such as red deer and roe deer, make it appear reasonable to conclude that meat was not the sole reason why these communities kept domestic cattle (and sheep and goats as well). Rather, it would seem that cattle served a variety of purposes in the economy and were slaughtered only when they were no longer economically useful. For most males, this would be either as calves or when they had reached their maximum meat weight, and for females, when they had ceased to produce milk or calves. In any case, Linear Pottery communities clearly had access to milk and to ignore such a resource would negate the economic and ecological rationale for keeping domestic cattle in the central European forests.
Ceramic Sieves
It is in this context that it would be worthwhile to examine a eglected Linear Pottery artifact type, the ceramic sieve. Archaeologists working at Linear Pottery settlements in central Europe have often found sherds which are perforated by many small holes (Fig. 3). Usually, one or two such sherds are encountered at any single site, if they are found at all. Sieve sherds are not ubiquitous: many large sites with much pottery have produced no such artifacts. On the other hand, smaller sites mentioned only in short notices in regional journals often include sieve sherds in their meager assemblages. They frequently are found at sites of this culture on the lowlands of the North European Plain in East Germany and Poland.
These lowland sites are generally small and do not have the longhouses that are found at the larger Linear Pottery sites elsewhere. Nonetheless, they have yielded relatively dense concentrations of ceramics, among which are one or two sieve sherds per site, almost without fail. Ceramic sieves may be a poorly known artifact category, but when viewed in light of the increasing body of empirical data on Linear Pottery subsistence, they could be significant for the understanding of the Early Neolithic economy in central Europe.
The hypothesis that the Linear Pottery ceramic sieves played a role in the production of dairy products finds considerable support in both the European archaeological and ethnographic records. The best-known examples of ceramic sieves from later European prehistory are found at the Bronze Age sites of central Italy. On the basis of their similarity to the metal vessels used by modern Italian shepherds for the separation of curds from whey in the production of sheep cheese, these vessels have been interpreted as cheese strainers (Barker 1981). Modern counterparts to the prehistoric sieves are found not only in central Italy, but in many other parts of Europe as well, particularly among the pastoral societies of the Balkans (Novak 1969; Dunâre 1969).
In all cases, these sieves are associated with dairy production, serving to strain curds from whey in the manufacture of cheese. In central Europe, ceramic sieves were also used for cheese manufacture into the first part of this century and formed an important product of the small-scale ceramic industries in many areas. Such perforated vessels for cottage cheese production are not confined to European peasant culture. I have seen 19th-century American examples from Vermont, an area that still has a significant dairy component in its economy.
The presence of clay sieves, presumably for cheese production, on Linear Pottery sites indicates that the people of this culture had at their disposal a means for countering whatever degree of lactose intolerance they may have had. In milk products such as cheese and yogurt, most of the lactose is removed with the whey in their production, and what little remains in cheese becomes hydrolyzed into lactic acid. Aged mature cheese contains no lactose. Without dairying, the herding of cattle in the Neolithic forests of temperate Europe would have been of questionable value, and the presence of the sieves counters any argument that the Neolithic inhabitants of temperate Europe did not milk their cattle due to a possible intolerance of lactose.
The Social Context of Neolithic Dairying
Them evidence of the clay sieves and the faunal remains associated with them indicates that there is a very high probability that milking and the use of dairy products such as cheese (and probably yogurt) were known by the earliest Neolithic inhabitants of temperate Europe. Dairy products which could be transported and stored may have played a significant role in the Early Neolithic subsistence system. As I noted above, the establishment of a successful agricultural economy in the uncharted forests of temperate Europe was an uncertain undertaking for the small Linear Pottery communities. Grain crops would have been subject to the predations of wild herbivores and plant diseases, as well as having to adapt to a shorter growing season than in southeast Europe. In addition, there is some evidence to indicate that fields were generally quite small and that the crops were often contaminated with a variety of weeds (Knörzer 1971). Based on these factors and other constraints, Gregg has argued that it would have been impractical for grain crops to have met more than 80 percent of a Linear Pottery community’s nutritional requirements and still permit adequate surpluses for emergency stores and possible exchange (1985).
In light of these constraints, it was necessary to have reliable secondary resources to enable the communities to meet basic dietary minima and to tide them through years with poor harvests. One such secondary resource was probably wild plants for which there is little archaeological evidence, hut whose use is probable given the great natural productivity of the primeval temperate European forest (Clarke 1976), and the fact that they would have started to be available in late spring and summer before the grain could be harvested. Dairy products were probably another such supplementary resource in the Linear Pottery economy. Not only would these have been available early in the growing season when the grain supplies from the previous harvest would have run low, but they could also be stored into the winter, and thus used to supplement the grain when wild plant foods were scarce.
Dairy products, then, would have been a part of the diversification of the Linear Pottery economy. These additional resources should not be considered `minor’ or ‘marginal’. Rather, they would have played an important role since they would have provided a substantial degree of reliability in subsistence. The more diversified their economic strategy, the more secure the Neolithic communities were.
The establishment of dairying as an economic pursuit during the Early Neolithic in temperate Europe fits in nicely with the model of the organization of these communities which has emerged in recent years. Linear Pottery and other Early Neolithic settlements have traditionally been viewed as “villages,” with the level of integration of activities that this term implies. Recently, however, the argument has been made that the settlements really represent loose concentrations of individual households or farmsteads, each an independent productive unit (Luning 1982; Crygiel 1986). The fine-grained chronological analysis of these settlements shows that not all houses were contemporaneous and that the actual density of population was not as great as the site plans suggest. The model of independent households as the fundamental social and productive unit of Neolithic society is supported by ethnographic data which indicate similar patterns in virtually all small-scale agrarian societies world wide (e.g. Netting, Wilk, and Arnold 1984).
Two crucial resources are of fundamental concern to any agrarian household: land and labor. In Neolithic Europe, potential agricultural land was effectively unlimited, particularly in the lowlands colonized by the Linear Pottery culture. Instead, a more critical problem for a Neolithic household was the amount of labor which it could mobilize, either internally or through external cooperative arrangements with other households. Agricultural communities, then, are inclined to have high reproduclive rates in order to maintain such labor pools. Much of this labor, however, is required for only a limited range of activities, such as field clearance, planting, and harvest, and would possibly be underused during much of the growing season. Livestock management, however, would have provided an additional productive activity which would have been most crucial at precisely the time of year when major physical activity would not have been required in the fields.
With this parallel source of nutrition, which could be managed without markedly detracting from the labor requirements for field agriculture, Neolithic households could afford to maintain a certain degree of economic self-sufficiency. This would have been possible only if dairying were the primary economic purpose of keeping cattle. If meat production were the sole goal, it would have been necessary to maintain much larger herds which in turn might have exceeded the ability of the household labor pool to maintain them. Moreover, in a household economy, the exclusive use of cattle for meat would have posed a problem, for only a fraction of the carcass could be consumed while still fresh, and the need to preserve and store the remainder would have increased the potential for spoilage. Again, pigs would have been better for this purpose, but they are quite rare in Linear Pottery contexts.
Self-sufficient Neolithic house-holds would have occasionally needed to enter into exchange relationships with other households, both local and distant. Such relationships would have been necessary to do things such as mobilize cooperative work groups or make up for tem-porary food shortages. Although exotic materials such as Spondylus shell and obsidian are found in some Linear Pottery sites, it seems likely that the major commodities in such exchanges would have been mundane subsistence products. Dairy products, especially cheese, would have been especially well-suited to these sorts of transactions, for they do not have the awkward bulk of grain or the indivisibility of living animals. Moreover, they can be produced by a household without cutting into its capital, whereas trading away living animals, particularly cows, would have reduced the productive capacity of the house-hold until replacement stock could be acquired. Gregg has proposed that Linear Pottery communities also entered into relationships with indigenous hunter-gatherer groups in parts of central Europe, and that dairy products may have been among the commodities exchanged (1985).
Conclusion
The presence of ceramic sieves on a number of Linear Pottery sites argues that the milking of domestic cattle was practiced by the Early Neolithic peoples of temperate Europe around 4500 b. c. (5400 B.C.). When these data are combined with the zoo-archaeological evidence from Linear Pottery sites, it appears that a system of dairy husbandry developed in Neolithic temperate Europe to supplement the cereal cultivation which formed the mainstay of the subsistence system. Such a diversified economy would have been crucial to the successful establishment of agrarian communities in the forests of temperate Europe. In some areas, such as the lowlands of the North European Plain, dairy husbandry appears to have been the predominant subsistence practice during this period.
The recognition of such an antiquity for dairy production in temperate Europe does not contradict the notion that towards the end of the Neolithic there was a shift towards the maximum utilization of animal resources that Sherratt has termed the “Secondary Products Revolution” (1981, 1983). At this time, particularly in eastern Europe, subsistence systems appear to have emerged which had a primary emphasis on animal husbandry. Legge notes that although any economy which includes cattle will have access to both meat and milk, the exploitation of one or the other will be most efficient if it is developed in a specialized way, as appears to have been the case in this area during the Late Neolithic and Early Bronze Age (1981a:89). The roots of these systems, however, lie several millennia earlier, during the colonization of temperate Europe by the Linear Pottery culture.
Peter Bogucki received his B,A. from the University of Pennsylvania and his Ph.D. (1981) from Harvard University in anthropology, Since 1976, he has taken part in research at Brzesc Kuawski, Poland, on Early Neolithic settlements. Currently, he is the director of studies at Forbes College, one of the residential colleges of Princeton University. (Author photo by Robert P. Matthews, Princeton University.)
Acknowledgments
An earlier article first articulated some of the arguments expanded here and presented more of the basic data on faunal remains and ceramic sieves (Bogucki 1984b). Dr. Andrew Sheraton (Oxford) and Dr. A.J. Legge (London) made useful comments on that article which were equally helpful in this one. Dr. Bernard Wailes and an anonymous reviewer made additional comments on this article, Responsibility for errors of fact or logic rests with the author, of course. Dr. Ryszard Crygiel (Lódz, Poland) provided the photographs of Linear Pottery sieves from Brzesc Kujawski. Since the 1984 article on Linear Pottery sieves, several European colleagues have called my attention to additional sieve sherds from Early Neolithic sites in central Europe, Peggy Hoffman patiently entered the text of this article into the word processor.