Note on the Construction of Mortality Tables by Means of Compound Frequency Curves [Discussion]

The general results achieved by Mr. Fisher in this paper are certainly as fascinating as any which have come from the laboratories of the modern analytic school in the physical and social sciences. Mr . Fisher's paper is in one sense a challenge. It calls for a statement of what sort of facts social statisticians are trying to discover, what limitations govern the use of such facts, and by what proper means we may bring data thus defined and circumscribed to bear upon the practical problems which face insurance science. Whether we work with the sense data of either the physical or the social sciences, we may agree, before we take up Mr. Fisher's paper in detail that there are four distinct categories of knowledge.