ABSTRACT
Every historian, as he approaches a historical problem, forms in his mind a picture of the evidence that he would like to have to enable him to solve his problem. We can consider this ideal as a set of data, the possession of all of which would provide a complete answer to whatever question the historian wishes to ask. Each historical problem will have, in this schematic sense, its own ideal data set, although clearly one data set will overlap with another, just as one historical problem overlaps with another. We can consider each ideal data set as consisting of a set of cases; each case is composed of a set of variable characteristics which distinguish one case from another, and the whole makes up an ideal data matrix.
