摘要：It is widely agreed that, the foundationalist program being exhausted, it is time for new direction~ in the philosophy of mathematics. Unfortunately, the plural in "directions" has been too rarely justified. We have seen a series of sociological case studies of the discovery of proofs in such areas as number theory, logic and the foundations of calculus. These have been not without interest, but are their choices of subject-matter perhaps too narrow? Sociology, like such sciences as psychology, history and anthropology, is a two-edged sword (or, perhaps more accurately, a boomerang). For these sciences are empirical (at least potentially), and subject to certain norms and standards of procedure and evidence. When Joseph de Maistre remarked, "You might as well say sheep are bor.n carnivorous, and everywhere they eat grass", he exposed the fact that Rousseau was not an actual anthropologist, and that the Noble Savage was an inhabitant of the European mind's stock of paradigms, thought experiments, rational reconstructions, or myths, not of Tahiti
