1The question addressed in this paper is, to what extent can the development of agriculture and of pea-sant societies in the west Mediterranean be seen as « inevitable », both in general terms and, specifically, in the 5th and 4th millennia b.c. ?

2Recent agriculture in the west Mediterranean has been based on the triad of wheat, vines and olives, but only the use of cereals, wheat and barley, will be dis-cussed here, since effective cultivation of grapes and olives post-dates the Neolithic in the west Mediterranean. The earliest sites with radiocarbon dates and evidence for cereal cultivation - apart from Châteauneuf-les-Martigues and possibly two South Italian sites - lie in the 5th millennium b.c (figure 1). Other sites probably dating from the same millennium b.c. contain pots with grain impressions, sickle blades, axes and querns, and -more rarely - are accompanied by pollen spectra including cereal grains.

3Recent discussions of the development of agriculture have emphasized the apparent inevitability of the spread of agricultural techniques, though emphases have varied. Jarman, Bailey and Jarman (1982) say,

’It can reasonably be argued that the change from Pleistocene to Holocene climatic conditions may have occasioned a signi-ficant decline in the biomass accessible to human exploitation in many environments, thus encouraging more efficient and productive patterns of utilisation... (Changes) could have placed human exploiters adjusted to Late Pleistocene ecosys-tems in positions of great economic stress. It seems more than likely that this was a most significant factor in the widespread development of modem agricultural economies’... The appa-rently intractable question of why these did not arise earlier résolves itself ultimately as a matter of population levels’. (p. 23).

4Butzer is equally emphatic about the inexorable grip of agriculture on the societies that practised it, both populations settled in a particular region and others colonizing new environments (1982, p. 311) :

’… the train of end-Pleistocene to late prehistorical adap-tive transformations exemplified in the Near East, Europe, Asia, Meso-America, and Andean South America comprised an apparently inexorable series of strong positive-feedback loops, generally involving greater human control over the environment. Each feedback sequence involved socio-economic, demographic, and ecological changes of a fundamental kind, equilibrium thresholds that required redefinition of the interrelationships between the key variables.’

5A particularly extreme point of view is taken by Rin-dos, who argues that agriculture is inherently instable, its low variety of crops being at risk from even minor environmental changes, animal predators, weed invasion and other problems (1980) ; according to Rin-dos, this instability, combined with population growth during favourable years, assured the inevitable movement of the agricultural economy from the east Mediterranean to both western and northern Europe.

6How does the evidence from the west Mediterranean fit this picture ? As suggested from Figure 1, and as illustrated in Figure 2 (taken from Jarman, Bailey and Jarman 1982 : fig. 48) 5th millennium b.c. sites in the west Mediterranean do not always lie on high arable potential soils (some of which, in any case, were too heavy for Neolithic agriculture). On the other hand, in apparently unsuitable coastal Portugal, for instance, site territory analyses have identified small patches of arable near some sites (Morais-Arnaud 1982). In addition, pollen analyses from the vicinity of archaeological sites have suggested human interference with the landscape from at least the mid-5th millennium b.c. : Nacimiento, Jaen (Rodriguez 1982) ; Llobregat valley (Guilaine et al. 1982) ; Le Baratin, Vaucluse (Triat-Laval 1978) ; Châteauneuf-les-Martigues (Renault-Miskovsky 1971). By the 4th millennium b.c. such indications are on the increase (see Phillips 1982 for the Midi).

7Settlement locations in the 5th millennium b.c. in the west Mediterranean include both caves and open sites, but the cave type of settlement is probably over-represented numerically. Cave assemblages commonly contain fewer evidences of agriculture than open sites, and may have been accumulated by farmers during seasonal special activities - such as animal transhumance -or by societies not involved in settled agriculture. Open sites have probably been lost by a combination of sea-level rise near the coast (Leucate is suggested to have been occupied when the sea-level was a minimum of 5 m. below its present - Freises and Montjardin 1982) and by erosion and valley infill inland (Butzer 1982 ch. 8, Frenzel 1979). Where open sites are known, and have been excavated to a sufficient extent, they contain a range of features such as hearths, pits, daub (presumably from huts) stone ’floors’ and ditches (e.g. Salemas, Portugal ; Le Baratin, Vaucluse ; Caucade, Alpes Maritimes ; Molino Casarotto, Veneto ; Leopardi village, Abruzzo ; and Piana di Curinga, Calabria). Such sites, which combine location near suitable soils for agriculture, evidence for relatively permanent settlement, and - in some cases - actual grain remains - seem to represent the ’inexorable’ march of the agricultural economy. Figure 3 illustrates schematically the rise in evidence for axes, harvesting knives (stone knives with sickle gloss) and querns at this time ; however, the innovation of such features, as opposed to their development, lay much earlier. Spratt (1982) has detailed the many possible barriers - both social and economic - to the adoption of innovations. In the case of agriculture in the west Mediterranean, it can be suggested that :

Either : 1. cereals were introduced and cultivated at the same time as the introduction of sheep and goat from ultimately east Mediterranean sources, and of the development of pottery : but that we have not found the relevant early sites.

Or : 2. the ‘cash flow’ curves in Figure 3 reflect an accurate view of the past, with cereal agriculture post-dating sheep-herding and pottery usage.

8Even if the first hypothesis is correst - and there is littly evidence to support it at present - the majority of 6th millennium b.c. sites were occupied by non-farming groups herding sheep or goat, among other subsistence strategies. Jarman, Bailey and Jarman (1982) have emphasized that the coastlands of the west Mediterranean would have been good for animal herbivore grazing, but suggest that some transhumance would have taken place during the summers (evidently not sharing the opinion of local pollen analysts who suggest that rainfall was much more evenly spread throughout the year at this period than today). An as-yet untested means of determining the diet of these 6th millennium b.c. people would be to subject parts of the skeletons (the Muge skeletons could be compared, for instance, with the La Sarsa burial, and the Avignon burial) to C¹³ analysis, which, as perfected by the Danes, can elucidate the diet (Tauber 1981).

9The relative rarety of burials from either Epipalaeolithic or Neolithic sites in the west Mediterranean may indicate that those buried were in some way ’special’, with a particular social standing (see Bartel 1982 for a review of analyses of mortuary practises).

10The role of elaborately decorated vases, bracelets, and the rare axes of 6th millennium b.c. coastal societies can best be explained as elements in the web of exchange goods reflecting ties of kinship and social obligations (as Lewthwaite 1981 suggests). Perhaps there was some improvement in boat-building or sailing technology at this period, as has been suggested for the expansion period in the Pacific (Bellwood 1978). The integration of cereal-growing into the economies of these societies would occur when and where settlement had become more permanent and populations had increased - both overall and probably in group size. (Such increases might have arisen as greater emphasis was placed on herding, as opposed to a broad-spectrum economy, and evidence from stratified sites should be tested to identify changes in group size and seasonality to check this point - for instance in West Languedoc, Geddes 1981).

‘Cash-flow’ diagrams for 3 Neolithic innovations in the west Mediterranean

11The influence of social obligations in leading to the development of greater sedentism (and ranking) has been clearly expounded by Bender (1978 ; see diagram in Phillips 1982, fig. 25). It is tentatively argued that the adoption of cereal agriculture in the west Mediterranean was delayed by a combination of relatively widely dispersed good arable soils, and of small local populations determined to pursue traditional mobile life-styles, in a ’negative cash flow’ attitude towards agriculture. Even more hypothetically, the removal of the bar to agriculture may have been occasioned by local population rise and social pressures towards sedentism.

12In sum, the development of agriculture and peasant societies were inevitable processes in the west Mediterranean, though social and environmental factors may have slowed the process at different stages.