special characteristics that are not seen elsewhere on earth. These special characteristics were, first, a sharp juxtaposition of hills, desert and alluvium (deposits of sand or mud formed by flowing water) and, second, narrow strips of land which caused a build-up of population so that the isthmus could not support traditional hunter-gathering.4 These 'hot spots' therefore became 'nuclear areas' where the prevailing conditions made it more urgent for early man in those regions to develop a different mode of subsistence. Whatever the truth of this attractively simple theory, or in regard to the number of times agriculture was 'invented', there is little doubt that the very first time, chronologically speaking, that plants and animals were domesticated, was in the 'fertile crescent' of south-west Asia. To understand fully what we are talking about we need to grasp the nature of the evidence about domestication, which means in the first instance understanding the relatively new science of palynology, or pollen analysis. Plants-especially the wind-pollinated tree species-each produce thousands of pollen grains every year, the outer skins of which are very tough, and very resistant to decay. Pollen varies in shape and size and, being organic, can be carbon-dated. Its age and genus, if not its species, can therefore often be determined and this has enabled archaeo-botanists (a relatively new specialism) to reconstruct the surface vegetation of the earth at different periods in the past. Plant remains (i.e., not just pollen) have now been identified and radio-carbon dated from hundreds of sites in the Middle East and, according to the Israeli geneticist Daniel Zohary, the picture is more or less clear. First, there were three cereals which formed the principal 'founder crops' of Neolithic agriculture. In order of importance, these were: emmer wheat ( Triticum turgidum , subspecies dicoccum ), barley ( Hordeum vulgare ) and einkorn wheat ( Triticum monococcum ). They first appeared in the tenth and ninth millennia BP. Second, the domestication of these cereals was accompanied by the cultivation of several 'companion plants', in particular the pea ( Pisum sativum ), the lentil ( Lens culaniris ), the chickpea ( Cicer arietinum ), bitter vetch ( Vicia ervilia ) and flax ( Linum usitatissimum ).5 In each case, the original wild variety, from which the domestic crop evolved, has now been identified; this enables us to see what advantages the domestic variants had over their wild cousins. In the case of einkorn wheat, for example, the main distinguishing trait between wild and cultivated varieties lies in the biology of seed dispersal. Wild einkorn has brittle ears, and the individual spikelets break up at maturity to disperse the seed. In the cultivated wheat, on the other hand, the mature ear is less brittle, stays intact, and will break only when threshed. In other words, to survive it needs to be reaped, and then sown. The same is true for the other crops: the domesticated varieties were less brittle than the wild types, so that the seeds are spread only once the plant has been reaped, thereby putting it under man's control. Comparison of the DNA of the various wheats all over the fertile crescent shows that they are fundamentally identical, much less varied than the DNA of wild wheats. This suggests that in each case domestication occurred only once. 'The plants with which food production started in the South West Asia "nuclear area" were transported (already as domesticated crops) to initiate agriculture all over these vast territories.'6 A number of specific sites have been identified where domestication may have first occurred. Among these are Tell Abu Hureyra and Tell Aswad in Syria, which date back to 10,000 years ago, Karacada(gcirck) in Turkey, Netiv Hagdud, Gilgal and Jericho, in the Jordan valley, and Aswan in the Damascus basin, also in Syria, which date back even further, to 12,000-10,500 BP. An alternative theory- still speculative-is that man's increasing control of fire enabled him to burn huge tracts of forest, and that the tender grasses and shoots that would have grown up amid the burnt remains would themselves have been, in effect, domesticated plants and would have attracted herbivorous game.7 This would have needed a knowledge of 'slash and burn' technology and tools sufficient to cut down large trees-to create fire-breaks. It is by no means certain that early humans had such tools. In the case of animal domestication the type of evidence is somewhat different. In the first place we should note that the general history of the earth helped somewhat: after the last ice age most species of mammal were smaller than hitherto.8 One or more of three criteria are generally taken as evidence of domestication: a change in species abundance-a sudden increase in the proportion of a species within the sequence of one site; a change in size-most wild species are larger than their domestic relatives, because humans found it easier to control smaller animals; and a change in population structure-in a domestic herd or flock, the age and sex structure is manipulated by its owners to maximise outputs, usually by the
conservation of females and the selection of sub-adult males. Using these criteria, the chronology of animal domestication appears to begin shortly after 9000 BP-that is, about 1,000 years after plant domestication. The sites where these processes occurred are all in the Middle East, indeed in the fertile crescent, at locations which are not identical to, but overlap with those for plant domestication. They include Abu Hureyra, at 9400 BP, Ganj Dareh in Iran, at 9000-8450 BP, Gritille in Turkey, at 8600-7770 BP, and Tell Aswad, Jericho, Ramad, 'Ain Ghazal, Beida and Basta, all just post-dating 9000 BP. In most cases, the sequence of domestication is generally taken to be: goats then sheep, to be closely followed by pigs and cattle. 'The transformation from a hunting and collecting economy, perhaps beginning with the cultivation of wild cereals, to the establishment of permanent villages and a mixed agricultural economy with fully domesticated races of plants and animals, took place over at least 3,000 years.' There was no radical break; for many years people simply tended 'wild gardens' rather than neat smallholdings or farms as we would recognise them. There was a transition period where hunter-gatherers culled smaller animals. Pigs do not adjust to the nomadic way of life, so their domestication implies sedentism.9 So far as animal domestication is concerned, it first took place in the hilly/mountainous region where modern-day Iran, Iraq and Turkey meet, the most likely reason for this being that, in a situation where most wild species were not naturally domesticable, hilly regions (with a variety of altitudes and therefore of vegetation) would have produced the greatest range of animal species, and the greatest variation of individuals within species. Such an environment would have been the most likely to have produced smaller types, more amenable to control. For the Old World, then, the location and timing of agriculture is understood, as are the plants and animals on which it was based. Further, there is a general agreement among palaeobiologists that domestication was invented only once and then spread to western Europe and India. Whether it also spread as far afield as south-east Asia and central Africa is still a moot point, and the most recent genetic evidence of farmers (as opposed to their plants) is not as conclusive as it might be. It shows that modern-day Greeks share 85-100 per cent of their (relevant) genes with Middle Easterners (from Baghdad, Ankara and Damascus), whereas Parisians share only 15-30 per cent. Some archaeologists have suggested that this means that it wasn't the idea that spread, but people practising the idea, but not everyone accepts this.10 Much more controversial, however, are the reasons for why agriculture developed, why it developed then, and why it developed where it did. This is clearly of major importance in understanding mankind's mental development. It is also an even more interesting question than it looks when you consider the fact that the hunter-gathering mode is actually quite an efficient way of leading one's life. Ethnographic evidence among hunter-gatherer tribes still in existence shows that they typically need to 'work' only three or four or five hours a day in order to provide for themselves and their kin. Skeletal remains of Stone Age farmers reveal more signs of malnutrition, infectious diseases and dental decay than those of their hunter-gatherer predecessors. Why, therefore, would one change such a set of circumstances for something different where one has to work far harder? In addition, reliance on grain imposed a far more monotonous diet on early humans than they had been used to in the time of hunting and gathering. In any case, when people first domesticated crops, these remained a minor part of the diet for centuries, possibly more than a thousand years. Again, why the change? One theory is that the switch to agriculture was made for ritualistic or social reasons, because the new foods were rare luxuries, which gradually spread, the way designer goods do in our own day. Lentils, for example, grow just two per wild plant and would hardly have staunched the hunger of a Stone Age family. Yet lentils are among the first crops of the Near East. Some palaeontologists feel beer was the most important end-product of these grains, the importance of alcohol in a ritual feast being obvious. But the most basic of the economic arguments stems from the fact that, as has already been mentioned, some time between 14,000 and 10,000 BP, the world suffered a major climatic change. This was partly a result of the end of the Ice Age which had the twin effects of raising sea levels and, in the warmer climate, encouraging the spread of forests. These two factors ensured that the amount of open land shrank