Spread of Agriculture Westward Across the Mediterranean

views updated



Arene Candide . . . . . . . . . . . . . . . . . 253

Caldeirão Cave . . . . . . . . . . . . . . . . 255

The beginnings of agriculture in the western part of the Mediterranean basin (which includes Italy, southern France, and Spain as well as major islands, such as Sardinia, Corsica, the Balearics, and, traditionally, Portugal) are associated archaeologically with the start of the Neolithic period. The earliest agricultural sites are known to have appeared across the western Mediterranean around 5500 b.c. The environment at this time was not significantly different from today's, the Ice Age having ended c. 12,000 b.c., and the climate having gradually warmed to roughly its present state. It was a typical Mediterranean climate, characterized by hot, dry summers and wet winters. The western Mediterranean was populated largely by mixed forests of oak, lime, and elm in the lower altitudes, changing to fir at higher elevations. The current open, brushy landscape characteristic of the modern Mediterranean is the result of erosion from millennia of agriculture, which did not really take hold for a thousand years after its initial appearance. Deer, ibex, and boar roamed the region and were hunted widely.

The western Mediterranean is replete with limestone caves, as a result of the uplift of Jurassic and Cretaceous geological beds. It also has an abundance of clay deposits (important for making pottery), with rich alluvial plains good for raising crops, separated by relatively dry uplands for hunting. With the melting of the Pleistocene glaciers, the sea levels in the western Mediterranean gradually began to rise. Flooded river valleys became rich coastal estuaries brimming with shellfish, fish, and waterfowl. Many Mesolithic (the last prehistoric period before the arrival of agriculture) archaeological sites are from estuarine shell middens. Middens are large piles of shells that sometimes contain artifacts and even burials left by prehistoric populations. Some of the largest in Europe are the Portuguese middens near the town of Muge in the Tagus valley. The rise in sea level had one additional consequence—it submerged sites. The sea level during the Early Neolithic was about 10 meters lower than it is now. A very few early coastal agricultural sites, such as Leucate-Corrège, just north of Perpignan in France, were discovered by dredging operations, but most undoubtedly have been lost.

After a century of excavation and study, we have a basic framework of knowledge regarding the spread of agriculture in the western Mediterranean. Identified sites tend to be primarily caves, although that probably stems from the fact that caves are visible features in the landscape (unlike open sites, which are evident mostly by their artifact scatters). Pre-agricultural Mesolithic sites other than middens are mainly caves or rock shelters, where flaked stone tools typically are found with butchered remains of deer and boar. The tools are quite small, sometimes less than a centimeter in length, and geometrically shaped, with transverse arrowheads (the business end being a blade edge that cuts into the target). It is thought that the Mesolithic people who created these sites lived in nuclear family units. They tended to disperse to the highlands to hunt in the summer and then aggregate along the coasts for the winter.

These flaked stone tool forms did not disappear during the Neolithic, and it seems that many Mesolithic sites continued to be inhabited into the Early Neolithic at this critical transition phase. The Early Neolithic material complex most notably contains pottery, along with the earliest groundstone axes and other groundstone objects, such as bracelets. There is direct evidence of domesticated forms of wheat and barley as well as domesticated sheep. The appearance of sheep is confirmed by the presence in site deposits of their bones, which can be distinguished from remains of wild species, such as ibex. Evidence of domesticated cereals has been recovered as carbonized remains in ancient fire pits as well as the occasional grain impression in the wall of an Early Neolithic pot.

Excavation of numerous classic cave sites has helped archaeologists date and define the material culture of these first agricultural populations. The ceramics are the most distinctive and informative. They were fired at low temperatures, without a kiln, and have distinct patterns of manufacture and decoration. A particular type of stamp-impressed early pottery, termed Impresso, was first identified from Arene Candide (5800–5300 b.c.) in Ligurian Italy. Other key cave sites with the more classic Cardial pottery are in the west, at Châteauneuf-les-Martigues near Marseilles (5750–5500 b.c.), La Grotte Gazel on the southern flank of the French Montagne Noir (4900–4830 b.c.), Cova de l'Or in Valencian Spain (5900–5300 b.c.), and Caldeir~ao in central Portugal (5900–5600 b.c.).

The earliest of early pottery, sometimes termed Le Vrai Cardial (true Cardial) is well made and highly burnished, and it is distributed broadly across the western Mediterranean; it may have been a trade good. Although only simple pyrotechnic methods were available to fire this pottery, which would have affected its hardness and durability, a great deal of labor went into manufacturing and decorating each vessel. The decorative style of these vessels is uniform, with a standard technological "recipe" of manufacture, showing a shared manufacturing tradition. Most important, many of these vessels have been carried over long distances. Later Cardial vessels have more regional styles, are less well made and decorated, and appear not to have been carried over short distances or used locally.

In addition to cave sites, some open-air sites have been excavated, indicating a diversity of settlement, typically in small villages. In Italy, Passo di Corvo, a walled settlement, and Piana di Curinga, a village of wattle-and-daub huts, represent two such sites. In France the site of Peiro Signado sits on a hilltop near the coast, and Leucate-Corrège was a coastal site now submerged. The site of La Draga in Spain has evidence of wooden walkways similar to the classic Neolithic lake dwellings of Switzerland.


Agriculture appears to have spread through this area fairly quickly during the sixth millennium b.c. With the help of radiocarbon dating, it is possible to trace the rate of agricultural dispersal in the western Mediterranean. In 1971 Albert Ammerman and Luigi Cavalli-Sforza proposed a "wave of advance" of early agriculture of approximately 5 kilometers a year for this area, much faster than the rate for other parts of Europe. Jo~ao Zilh~ao, a Portuguese prehistorian, later reexamined the dates for the western Mediterranean area and proposed that the rate was closer to 10 kilometers a year. This would mean that agricultural societies spread from Italy to Portugal in just one hundred years.

There is good information about the possible mechanisms for the spread of agriculture in the Mediterranean basin. Boats have been discovered at Mesolithic sites, so people at this time could have traveled faster and over longer distances than by foot on land. Moreover, the landscape was not empty. The spread of agriculture could have been either helped or hindered by the existence of the pre-agricultural populations that already inhabited the Mediterranean estuaries and caves. Finally, the earliest agricultural communities in this area may not have been sedentary farming villages but rather communities using a combination of domesticated and wild foods. The arrival of agriculture in the western Mediterranean area certainly was accompanied by a period of rapid cultural and economic change that formed the foundation for subsequent prehistoric developments.

Theories of the spread of agriculture across the western Mediterranean have evolved from "ages" to "revolutions" and then to models of human economic and social behavior. For the most part, preagriculturalists are thought to have had an economy based on the hunting of animals and fishes and the gathering of shellfish, berries, seeds, roots, and other edible plants. This lifestyle is the basis for the term "hunter-gatherers." Also known as "foragers," they depended on the collection of foods from the natural environment. Early agriculturalists are seen as subsistence farmers or pastoralists, not highly productive but able to guard against difficult times by storing surplus or keeping herds. Still, the basic question concerning the adoption of farming by gathering societies or the migration of agricultural villagers persists.

The most promising of the models of the transition to agriculture build upon the concept of frontiers—zones that lie between groups with different economies or ethnic territories, across which people, goods, ideas, innovations, and conflicts pass. Frontiers where the economic strategies are mismatched, for example, between agriculturalists and hunter-gatherers, are the ones likely to move as one strategy replaces the other. Thus, frontier models seem to explain the agricultural transition in the western Mediterranean and elsewhere.

The Ammerman and Cavalli-Sforza "wave of advance" model mentioned earlier was one of the first frontier models, and the most basic. As a population-diffusion (migration) model, it proposed that agricultural settlements spread outward at a regular rate in a wave, similarly to a ripple moving across the surface of water. It was easy to model mathematically, yet could not accommodate important variables. The types of agriculture or hunting and gathering being practiced on either side of the agricultural frontier could easily affect the rate of agricultural spread as well as possible cultural resistance or receptivity. Social factors would have determined whether these groups interacted peacefully or through conflict. Geography, environment, climate, and transportation (such as boats) also could have influenced the rate of spread. Obviously, some of these factors contributed to the five- to tenfold inaccuracy of this model's proposed dispersal rate in the Mediterranean.

Given the few data we have about these transitional agricultural societies, it is unlikely that any mathematical model, no matter how complex, can ever be tested. Archaeologists may never excavate even 1 percent of all the sites inhabited during this period. Most of the materials made and used were organic and have long since decayed and disappeared. Moreover, it is difficult to reconstruct the ways in which they might have been used by prehistoric peoples. Radiocarbon dating, the best technique we have for identifying contemporaneous sites, provides a statistical estimation accurate to about one hundred years at 64 percent likelihood. That time span represents roughly five generations of habitation, which makes it very difficult to relate to real people and the activities that produced particular sites.

There are alternate ways to improve our understanding of these sites. Researchers have used frontier descriptions as models to understand how agriculture might have spread across the western Mediterranean. This method has allowed archaeologists to incorporate more variables or even to lay predictive patterns over an actual rather than a theoretical landscape. By looking at real situations and the large-scale impacts of small-scale societies, we can gain a better idea of the potential underlying forces.

These ethnographic models derive mostly from studies of present-day small scale societies (band- or village-level societies), where it is assumed that such societies are more traditional and therefore somewhat like our prehistoric ancestors. These models have been used extensively to better understand how foragers and farmers might have interacted in the past, as agricultural economies were established. Exemplars of prestige exchange are based on studies of potlatching among populations along the northwest coast of America or on Polynesian prestige trade rings. Ideas about simple hunting-andgathering groups come from studies of migratory groups in the Kalahari Desert of southern Africa and from interior Australia. Examples of aggressive agricultural populations similarly come from the study of twentieth-century Pacific societies or village agriculturalists of South America and Africa, whereas notions of simple farmers are taken from studies of "subsistence" farmers of Africa and elsewhere.

The nature of the societies on either side of the agricultural frontier greatly influenced the rapid transition in the western Mediterranean. Some scholars have viewed pre-agricultural populations as relatively simple family-sized groups, whereas others have seen them as more complex societies with exchange networks, driven by competition for prestige. Similarly, on the agricultural side, cultures can be viewed either as very simple farmers not much better off than their forager neighbors or as true village agriculturalists. None of these four options ideally captures what it must have been like in the western Mediterranean six thousand years ago, but it is a good start and one based on ethnographic studies of real peoples.

The notion that simple subsistence farmers made contact with simple gathering groups cannot account for the rapid agricultural spread across the Mediterranean, as there is no predominant inequality, such as prestige or settled agriculture, to move the frontier. Complex foragers or complex agriculturalists could have effected this change. Brian Hayden has developed a model for the emergence of agriculture through prestige competition. His theory is that among some foraging societies there was competition for status. Agricultural products represent storable wealth as well as a potentially greater food supply. With prestige competition among foragers, power was accumulated through prestigious objects or through obligations. Agricultural items, such as domesticated animals, would have been an innovative and therefore prestigious object in such a setting. They could have been rapidly dispersed through such a culture, with agricultural dependence as an unintended consequence of the desire to accumulate prestige.

In 1986 Marek Zvelebil, an archaeologist at the University of Sheffield, presented the concept of adoption as the availability model for the transition from foraging to farming. In this model there are three zones across a frontier between agriculturalists and foragers. First, in an availability zone, where domestic plants and animals as well as pottery and other new items became available to foragers by trade, but these items did not really affect the overall economy of the group. A desire for prestige or power would have driven individuals to invest in these innovations. In subsequent phases, farming developed as an alternative economic strategy alongside foraging. Last, settled farming consolidated as the principal economic strategy.

How does the evidence in the western Mediterranean support this theory? The availability model proposes stratified Mesolithic populations with a network of trade in prestige goods. It predicts the initial appearance of domesticated plants and animals within a predominantly foraging economy. It does not require early settled agricultural villages.

Evidence of pre-agricultural trade is slim. There are few signs of the movement of material goods during the Mesolithic, but there also is little recoverable and traceable material. The stone sources used to make flaked tools in the latest Mesolithic times come from more local regions than in earlier periods, so they probably were not exchanged. It is possible that there were valued exchange items made of organic materials that do not survive. From the earliest Neolithic period, however, there is evidence of trade. It has been shown that the earliest pottery, Cardial, was transported long distances, as were the earliest groundstone axes. The exchange of obsidian, a volcanic glass used to make very distinct flaked tools, also began across the Mediterranean at this time.

In terms of economy, there is evidence from many Early Neolithic sites that wild species initially predominated among the deposits. Bones of wild deer and boar at first outnumbered those of domesticated sheep at many Early Neolithic cave sites. The proportions of wild to domesticated animals gradually changed over a period of hundreds of years, until domesticated animals came to dominate the archaeological assemblages. The animal bones recovered from Early Neolithic coastal sites, such as Leucate-Corrège, include the remains of species hunted at different times of the year, indicating that there were permanent settlements at this time that were not necessarily agricultural villages.

The results are mixed in terms of making the case for adoption. There is no good evidence of trade before the first agricultural sites, but trade of objects clearly is present at these Early Neolithic sites and is quite uniform across the western Mediterranean. Still, there might have been trade in organic materials, such as furs or meat, or in social obligations, such as labor. The availability model for agricultural adoption correctly predicts the proportions of wild to domesticated animals. With boats, it would have been possible for these innovations to spread rapidly enough to get from Italy to Portugal in one hundred years.

Equally, village agriculturalists also might have migrated rapidly across the Mediterranean, even in a landscape already occupied by Mesolithic foragers. Zilh~ao has proposed a model of enclave colonization. Enclave colonization involves resettlement by small seafaring groups of agriculturalists across the western Mediterranean. It is described as the budding off of small groups to found new agricultural colonies. This budding off might have been driven by offspring required to homestead enclaves as part of their entry into adulthood or through inheritance or as part of planned enterprises. Upon arrival, they would have assimilated or displaced local foragers, despite their smaller numbers, due to the inherent superiority of a production economy. This agricultural economy would have provided a stable surplus of food that would have allowed their populations to grow and to trade successfully with local foragers.

How do the data support enclave colonization? This model predicts rapid spread through the establishment of far-flung settled villages. Because these societies initially were smaller and more isolated than the resident Mesolithic societies, they might have had to settle in areas that were undesirable or relatively unused by resident foragers or to have been readily adopted by the foraging groups. This theory also requires that a "package" of domesticated plants and animals and new technologies, such as pottery and groundstone axes, spread as a uniform and interdependent economic strategy.

Certainly, it seems as if agricultural items spread across the western Mediterranean at almost the same time, insofar as the relative inaccuracies of radiocarbon dating permit us to say. This is in contrast to a more piecemeal adoption of innovations that might have taken place had they been brought across the Mediterranean as unlinked prestige trade items. Zilh~ao also has identified upland areas in regions of the western Mediterranean, particularly in the Estremadura of central Portugal, where there appears to have been little or no Mesolithic habitation, yet a strong Early Neolithic occupation. Just to the south, in the Tagus estuary, are the remains of some of the largest Mesolithic shell middens in Europe. This area could have been a region leaped over and not colonized by agricultural enclaves.

In terms of economy, there is no strong evidence of a dramatic shift to full-scale agriculture. As mentioned earlier, most Early Neolithic sites are dominated by the bones of wild, not domesticated, animals. It is possible, of course, that these initial agriculturalists ate mostly wild animals because they traded locally with foragers or that they added to their food supplies with regular hunting, as do many modern peoples.


How can we distinguish what really went on when agriculture first spread across the western Mediterranean? Neither adoption nor migration models seem to single-handedly match up with the thin amount of information we have collected. Mesolithic foragers could not have traded for agricultural goods without moving, and early agriculturalists could not have moved without coming in contact with whomever was there already. There are tantalizing bits of evidence, such as the rapid spread of domesticated plants and animals and new technologies like pottery. There also is evidence of the birth or rapid expansion of trade routes at this time. Still, there is the persistence of flaked stone tool traditions, habitation, and economy across the transition. The models we are using, based on modern examples, will have to expand beyond the simple one of migration versus adoption.

Zvelebil has described more flexible options that will provide a good testing ground. These include individual frontier mobility (spread through kinship-based exchanges of individuals or small groups), leapfrog colonization (highly selective colonization by seafaring peoples), and infiltration (gradual penetration by groups that assume a subordinate political position and perform specialized tasks). As we broaden our perspectives and find ways to evaluate these models, we will come closer to understanding what it was like at the moment when the first inklings of the foundations of European civilization spread across the western Mediterranean.

See alsoMuge Shell Middens (vol. 1, part 2). Caldeirão Cave (vol. 1, part 3).


Ammerman, Albert J., and Luigi L. Cavalli-Sforza. The Neolithic Transition and the Genetics of Populations in Europe. Princeton, N.J.: Princeton University Press, 1984.

Barnett, William K., and John W. Hoopes, eds. The Emergence of Pottery: Technology and Innovation in Ancient Societies. Washington, D.C.: Smithsonian Institution Press, 1995.

Price, T. Douglas, ed. Europe's First Farmers. Cambridge, U.K.: Cambridge University Press, 2000.

Renfrew, Colin. Before Civilization: The Radiocarbon Revolution and Prehistoric Europe. London: Jonathan Cape, 1973.

Zilh~ao, Jo~ao. "Radiocarbon Evidence for Maritime Pioneer Colonization at the Origins of Farming in West Mediterranean Europe." Proceedings of the National Academy of Sciences 98, no. 24 (2001): 14180–14185.

Zvelebil, Marek, ed. Hunters in Transition: Mesolithic Societies of Temperate Eurasia and their Transition to Farming. Cambridge, U.K.: Cambridge University Press, 1986.

William K. Barnett