Daniel Milo. The unusual but edifying encounter of Culture and Numbers. It is often assumed that learned culture is résistent to statistics yet, series of the best quality, uninterrupted, in large quantity, sometimes exhaustive and first hand, are often available in this field and nowhere else. The historian therefore finds himself virtually compelled to deal with numbers. But on the one condition that he admits the fundamentally external character of this approach. He must therefore not seek at all costs to anchor his series, his grids his juxtaposition of variables into reality. For, as in all experimental methods, quantification submits the object of research to proofs that « nature » avoids providing. One need not ply to reality, parado...