sionally seems to look upon Verstehen as a fruitful source of hypotheses about behaviour hypotheses that must then be subjected to empirical scrutiny and validation. And in doing this it is quite in order to bring into play forensic skills and quantitative methods in the classic Durkheimian fashion. Unlike some of his intellectual heirs, Weber by no means regarded the use of statistical techniques as an exercise in mystification or as a distortion of the subtle realities of social life. Statistical probability was an important check upon the general validity of any proposition At the same time, caution was called for in attaching explanatory significance to numerical correlations.