Boole's logic and probability. A critical exposition from the standpoint of contemporary algebra, logic and probability theory. 2nd ed., rev. and enlarged (Q1087535): Difference between revisions

In passing from a first edition (1976; Zbl 0352.02002) into a second edition, this work has grown from 243 pages (not counting the index) to 428 pages (counting the index). The new pages are welcome pages. In particular, there is a whole new chapter, Chapter Six titled, ''Applications. Probability Logic'', that contains welcome and useful material concerning the relation between the assignment of probabilities to sets of related statements and the linear programming problem of maximizing (or minimizing) a function subject to linear constraints. Much of this material will be of interest to those concerned with the propagation of ''uncertainty'' in networks representing statements - a problem of central concern in artificial intelligence. But let me begin at the beginning. Chapter 0, providing background, is essentially unchanged. It provides an elegant and clear summary of the rather large variety of mathematical disciplines that are relevant to a discussion of Boole's work. Chapter 1, largely historical in character, has been significantly expanded. Chapter 2 has changed terminologically and it has an expanded discussion of Boole's notion of ''uninterpretability''. Chapter 3 has been expanded by the incorporation of new historical material. The greatest substantive changes occur in Chapters 5 and 6. Chapter 5, ''Boole's probability made rigorous'' has been extensively rewritten, and is now of considerable relevance to modern probabilistic methods. Chapter 6, as remarked, is entirely new. It contains an extensive treatment, not of historical interest only, of the modern pressing problems of inverse probability, the probability of causes, the combining of testimony (read also: the combining of evidence). This was an important and useful book in its first edition. It has now become even more important, since it now addresses a large class of contemporary problems in the handling and propagation of uncertainty.