were patchy rather than fully assimilated, and details, rather than the whole approach, were adopted. Excavation techniques were used when necessary, as was the criticism of ethnographic analogy when trying to establish a different analogy (Leroi-Gourhan 1985). Nevertheless by the end of the 1990s and at the beginning of the third millennium, his overall view of the process of hominid development—which raises the question of what it is to be human—was at last put into perspective, and his analysis of post-Neanderthal epiphylogenesis became part of the cognitive sciences. The visionary postulate of his book Le Geste et la parole (1964, 1965) and the logic underlying Mécanique vivante (1983), written in 1954, appear now to be remarkably relevant and epistemologically inspiring, compared with the debates on the origins of modern humans that stir up and cause futile clashes between researchers from all parts of the world.

Stumbling Blocks on the Way Out of the Wilderness

Etudes archéologiques, the wide-ranging volume by several authors edited by Paul Courbin in 1963, was uninspiring and made no impact outside professional circles. But some of the papers (such as the one by Courbin himself on stratigraphy) were critical of French archaeology abroad, which had long been accustomed to clearing the most renowned sites around the Mediterranean in a somewhat cavalier fashion. Other papers, such as those of André Leroi-Gourhan and Jean-Claude Gardin, set out the program of work for the future. For this reason, these two authors were to be among the few French role models for a whole generation.

Almost all archaeological attention in the 1960s and 1970s concentrated on the technical aspects of observation and recording. As prehistoric archaeology emerged from the wilderness of a lack of public and government support, essential improvements in excavation methods quickly became an ideological stumbling block: for classical archaeologists, stratigraphic excavation was an end in itself, whereas prehistoric archaeologists always and everywhere excavated using the delicate tools of dentist’s spatula and fine brush. Consequently, when in 1971 Bohumil Soudsky started excavating at Cuiry-les-Chaudardes and other sites in the Aisne valley by machine-stripping, which he had perfected twenty years earlier in Bohemia, this use of this innovation—which was generally accepted everywhere else in Europe—caused outrage. And when, also in the Aisne valley, the first report appeared in 1973 listing all the sites threatened with destruction and suggested the idea of “selecting” which should be given priority on the basis of urgency, available funding, and scientific interest, the same violent reaction was unleashed. At that time in France there was no middle way between excavating inch by inch and total destruction. Methods based on physics and chemistry (“archaeometry”) that had begun to be used by archaeologists were another stumbling block. The ability to obtain data and generate results using scientific apparatus and statistical techniques opened up the possibility of dispensing with philosophy, or the under-picture, which gave physicists—themselves just as much novices in this new collaboration—an unwarranted feeling of absolute power. Disillusionment was, however, painful when the first assessments were made in the early 1980s and these substitutes for clear thinking were found lacking. Documentation, data-processing, and interpretation encountered the same problems. Leroi-Gourhan’s seminar at the Collège de France was devoted to developing a “temporary terminology” that was not just descriptive but also provisional. The creation of archaeological databases and expert systems as proposed by Gardin (1970) suffered from the same generalizing and self-defeating tendencies. In fact, the difficulty of creating a list of standardized generally applicable terms, required in order to computerize data, was an obstacle to creating a completely formal scientific discipline in the manner prophesied by Gardin (1979). Instead, two disparate kinds of study developed: nonspecialist documentary studies using a highly simplified descriptive vocabulary and individual, narrowly specialized research projects using incompatible complex descriptive terms. In studies of prehistory, the “analytical typology” of Laplace was unlikely to be adopted because its