The second edition of Judea Pearl’s book, Causality, remains the definitive text for the use of graphical models in reasoning about causal relations. Since the publication of the first edition in 2000, the use of graphical causal models has increased considerably in epidemiology and has found its way into the social sciences (Morgan and Winship, 2007). Like the first edition, what the book sets out to do, it does well: it provides a thorough exposition of the power and flexibility of a structural approach to causality by using causal diagrams. The book has been updated with a few recent developments and additional commentary has been added throughout the book. The content of the first 10 chapters of the book very closely matches the first edition with Chapter 1 providing an introduction to causality, Chapter 2 on causal discovery, Chapters 3 and 4 on identification of the effects of single and multiple interventions respectively, Chapters 5 and 6 on relations with the social science and epidemiologic literatures, Chapter 7 on an axiomatization of the author’s framework, Chapter 8 on bounds, Chapter 9 on necessity and sufficiency in causal reasoning and Chapter 10 on the problem of actual causation. The table of contents closely corresponds to that of the first edition and the reader is referred for further detail to the many reviews of the previous edition that can be found on Pearl’s Web site (http://bayes.cs.ucla.edu/BOOK2K/book__review.html). In the second edition, these chapters include minor updates and in some cases a new conclusion or postscript section that offers further reflections and occasionally interacts with some of the related literature on the topic that was not included in the first edition. What is principally new in the second edition is Chapter 11, entitled ‘Reflections, elaborations, and discussions with readers’. The chapter amounts to approximately 70 pages and consists largely (but not exclusively) of responses to and discussion of comments and questions that have been made by readers of the first edition.
No takes yet. Share an insight, caveat, or question.
Tyler J. VanderWeele (2011) studied this question.