Review of “David Lewis”

on books
David Lewis
history of analytic

University of Michigan


January 1, 2007


A review of David Lewis by Daniel Nolan.

David Lewis left us with a rich and highly integrated body of work. Its richness means that any serious student of philosophy should study it closely. And at first this looks like it should be something that any student should be able to tackle. Lewis is one of the great stylists of his generation, and his views on most topics are expressed with admirable clarity. But the integration makes it difficult for the beginner to get a foothold. Any place you look it seems you will have to master five other topics before you really understand what Lewis says on this topic. So a systematic introduction to Lewis’s views is needed, and Daniel Nolan’s new book provides one.

There are two types of reader who will most benefit from Nolan’s book.

The book will obviously be very valuable for students, especially undergraduates. It would make an excellent textbook for upper level classes on metaphysics or philosophy of mind where Lewis’s views were an important part of the course.

Although it is mostly not pitched at experts, the book should also have value for professional philosophers because of how it draws out the connections between Lewis’s views. For instance, someone working on the rule-following paradoxes who was interested in learning more about the notion of semantic eligibility that does so much work in Lewis’s solution to these paradoxes could learn here how this notion is related to notions from the theory of properties and the analysis of physical law. But this is not just a textbook, and the critical element of Nolan’s exposition should be helpful at points even to experts.

Nolan’s summary starts with Lewis’s commitment to realism, both scientific and metaphysical. He then spends three chapters setting out the building blocks we need for Lewis’s metaphysics (properties, other times and other worlds) and showing how these can get used to rich picture of reality, one replete with causation, laws, dispositions and chances. Chapter five is on Lewis’s distinctive metaphysics of mind, and discusses how it relates to functionalist theories and the identity theory. Chapters six and seven are on content, mental and linguistic respectively. Chapter eight surveys Lewis’s views on ethics and value theory, and the last chapter is on Lewis’s methodology, especially his use of Ramsey sentences as a way of defining theoretical terms.

There is much to like through all of this. Although all the parts of the metaphysical picture are set out throughout Lewis’s writings, this is the best systematic exposition of the picture in a single place. The treatment of causation, which cuts through a lot of complicated discussion to get to the essence of Lewis’s theory, is especially useful. The discussion of Lewis’s ethical views, focussing on his complicated relationship to consequentialism and virtue ethics, does an excellent job of drawing a relatively systematic theory out of scattered remarks from several obscure sources.

Obviously there is a lot that could not be covered in this kind of book. So there is very little on perception, nothing on philosophy of mathematics, next to nothing on formal philosophy save a small discussion of the semantics of counterfactuals, nothing on Lewis’s arguments that desire and belief are separate existences and so on. These must have been hard cuts, but given the target audience I think they were the right ones. If anything I would have been tempted to cut even more to allow a little more space to the topics covered. I doubt that typical readers will get much out of the discussion of indeterminate probabilities towards the end of chapter 6 for example, as interesting as that should be to experts.

Although the book is primarily expository there is a good amount of critical discussion interspersed throughout. Lewis’s single strangest view, that dispositions must be grounded in intrinsic properties of the bearer of the disposition, comes in for extended and well-targeted criticism. And Nolan raises some interesting cases that suggest the cases not covered by Lewis’s story in “Mad Pain and Martian Pain” could be closer to home than Lewis wants. It might be worried that this much critical engagement will undermine the effectiveness of the book as a text, but I think it is all beneficial. For one thing, the criticisms often help highlight the contours of the theory. But there is a deeper reason too. A student learning Lewisian philosophy shouldn’t just be learning a bunch of Lewis’s theories. They should be learning something about how to do philosophy, which means putting forward theories and criticisms of theories. The criticisms Nolan makes, all of them the kind of criticism that Lewis would have taken seriously and even have made in other circumstances, help teach the student how progress is made within the paradigm Lewis established.

Lewis had a weakness for fantastic examples. His work is littered with stories of Martians and gods and wizards and infallible predictors. He thought, as I do, that these were perfectly appropriate in the contexts he used them. But they create difficulties because students, and professionals, don’t see the relevance of these fantasies to real world analysis. Nolan does the student, and the instructor, a real service by replacing these examples with down-to-earth ones. The sections on causation and causal decision theory in particular are clarified by these changed examples.

There are a few things that may have been done differently given the target audience. The bibliography only includes those works by Lewis that are cited in the text; it should have been a complete bibliography of Lewis’s work. Nolan from time to time refers to things that Lewis’s critics say without referring to those critics by name, let alone citing a reference. The book would be a more useful resource if it pointed explicitly to where the reader might see these criticisms set out in more depth. And on one or two occasions the book presupposes much more knowledge than its primary target reader will have. In the discussion of the view that all belief is de se belief, for example, Nolan suddenly presupposes familiarity with causal descriptivism about names without so much as introducing descriptivism.

But the virtues of the book outweigh these possible imperfections. Lewis’s philosophical work should be taught to as many of the next generation of philosophers as possible. (Not to mention the present generation.) Those of us engaged in this task would find our job easier if we had a clear and systematic presentation of Lewis’s philosophy. Now we do.