More on Upper Paleolithic Engraving

by C. J. J. VERMEULEN and A. DE Rui1rTER Universiteit van Amsterdam, Antropologisch-Sociologisch Centrum, Keizersgracht 397, Amsterdam C, the Netherlands; Instituut voor culturele antropologie, Rijksuniversiteit Utrecht, Heidelberglaan 2, Utrecht, the Netherlands. 4 x 74 Flower's problem is, according to us, an important one, because it raises the question whether and how one can count cultural units. In our discussion of classification, we raised the same question in, we think, a more general and fundamental way by asking what the epistemological status of these (classified) units is. We nowhere stated that Murdock does not use theories in his Social Structure. We agree with Barnes that he does. We believe, however, that Murdock's Social Structure is largely taken up with definition, classification, and correlation-as Barnes himself does (1971:22)-and that when Murdock uses theories he only tests them indirectly (see our general remarks on this point in the summary and conclusions of our article). Barnes makes the same points when he states that "Murdock's empirical generalizations impinge only tangentially on his chosen theory." It seems to us that our criticism of Murdock is similar to Barnes's.