The degree of agreement between the results of a comparative analysis of seven cross-cultural samples is dependent upon both the nature of the samples and the nature of the variables. All three of the smaller samples examined are demonstrated to be subject to a fatal amount of sampling error. The study incorporates the Tylor-Galton problem and makes an epistemic correlation between the concepts by postulation, historical-diffusional and functional or causal aspects, and the empirical findings of statistical correlation. It speaks in terms of realms of order in sociocultural phenomena and of different degrees of invariant relationship and contingency. Many philosophical puzzles and many problems in theoretical anthropology have, at least in part, empirical solutions. Any theoretical inquiry into questions such as “the nature of culture” should be closely allied with an examination of empirical data. The present approach is an attempt to integrate rational analysis with empirical data.