It may seem untimely to start worrying about the philosphical basis of survey research at a time when the main difficulty facing any of us is most probably getting cash to do the research at all. However, perhaps for that very reason, arguments, which are declaring that survey research is after all perhaps not on a very sound epistemological footing are gaining currency; I was forced to reflect hard on the process of decision-making in large-scale organisations when the S.S.R.C. suddenly discovered in 1975, when funds were beginning to dry up, that it had changed its mind on the importance of survey research and decided to close the S.S.R.C. Survey Unit. Many of those arguments and arguments since have made vague references to unease about survey research as a method applicable to producing sociological theory, and some of them I think touch chords in all of us when we consider what contributions to sociological theory have actually been made by survey research.