An excellent book ...
To write history, or even to read it, is to be endlessly engaged in a process of selection. No part of the job is more difficult or more important, and yet no part has been studied with less system, or practiced with less method. Many facts are called, but few are consciously chosen, on explicit and rational criteria of factual significance.
Historians have, I think, deliberately resisted the refinement and rationalization of this aspect of their task. For some of them, the idea of a chosen fact is hateful in itself. Others think that the process of selection is not merely an unknown but a mystery. Both of these beliefs are utterly mistaken. The process of selection can and must be clarified if history is to develop beyond its present condition. Criteria of factual significance can and must be specified, or else historians will be running through records like rats in a maze, without even a rudimentary notion of the nature of their predicament. In every historical inquiry, the process of factual selection operates in several separate ways. One species of selection, commonly called "sampling," consists in the selection of representative facts of a certain predetermined kind, within a closed universe of investigation. This undertaking, difficult as it often is in practice, presents comparatively few conceptual problems. We shall consider it in a later chapter. A much tougher problem is presented by the complex selection process that precedes sampling-namely, the selection of the kinds of facts to be sampled. This involves the determination of fundamental criteria of
factual significance.
Significance is the sort of word which sends shudders down the spine of every working historian. But if he turns his back upon the problem, he will operate upon covert criteria of factual significance, which may be incompatible with his own objectives. "He who is deficient in the art of selection," Macaulay wrote, "may, by showing nothing but the truth, produce the effect of the grossest falsehood." And he who is deficient in his ideas of significance is sure to be equally so in the art of selection.
There are a few common fallacies in this respect, a few criteria of factual significance which are inconsistent with empiricism in history. In each of the following instances, the fallacy consists not in the criteria themselves but rather in an attempt to combine them with the methods and objects of empirical inquiry. Every historian, of course, has the inalienable right to do any kind of history that pleases him-a constitutional right to go wrong in his own way. But no historian is liberated from the logical consequences of his own assumptions. The object of this chapter is not to condemn certain deviations from narrow academic norms. It is rather to demonstrate that all historians operate upon certain criteria of factual significance in their work, and that those criteria must be aligned with their own purposes, and methods.
