Comments

Abstract

To one who had recently read Theory of Experimental Inference (TEI) and has considerable interest in statistical inference and scientific method generally, it came as a disappointment that the first three chapters of this book received no mention whatsoever in Churchman’s review after 21 years. In these early chapters, Churchman had examined the possibility that the then still fairly new statistical theory of Neyman and Pearson provided a general basis for all inductive reasoning. He concluded, correctly, that on this view every inductive inference rests on empirical presuppositions that could not all be tested by the same methods without making further presuppositions, and so on indefinitely. The rest of the book is really an examination of possible ways of dealing with this result. From the standpoint of the last 21 years of philosophical discussion about inductive reasoning, it is unfortunate that Churchman’s subsequent work follows certain trends from the latter part of TEI rather than the trend of the first three chapters. Since no one else developed this viewpoint, few philosophers today know much about the statistical methods which became standard throughout the scientific world. But perhaps the loss to philosophy has been offset by the gain to systems science, which is where Churchman is today.