This is a dappled book. The chapters seem to have been separately written because they really do not form a unified and focused whole. The argument in the first few chapters is clear (though it didn't convince me), but after that the author bounces from one topic to another without much order or reason.

Nancy Cartwright certainly has fashioned a unique place for herself in the philosophy of science, as a mathematically and economically literate writer prepared to write books with titles like How the Laws of Physics Lie - not exactly from the Carl Sagan playbook, after all. However, despite certain allegations to the contrary, this is not wooly headed postmodernism but technical, analytical philosophy and as such suffers less, not more, than usual from allegations of academic irrelevance: Cartwright knows her maths and her economics, and she can talk turkey. Boy can she talk turkey.

Which (certainly to the extent this book purports to be aimed at the popular market, and probably even where it doesn't) is a large part of the problem. Perhaps in feeling the need to prove her credentials, Cartwright not only chooses highly arcane, technical and therefore, to readers like me, obscure examples, but then expounds them in mind-numbing, greek-alphabet fetishising, detail. The level of assumed knowledge to follow the worked examples in physics and econometrics is too high certainly for the mass market, but also I suspect for many professional philosophers. While I'm not one of those, I'm read enough professional philosophy in this field to know that I ought to be able to keep up with most of it, and that a better job might have been done in keeping me along for the ride than was actually done here.

Nor is Cartwright a particularly elegant writer. The concepts she is asking the reader to accept are radical, and whilst I thought they were pretty clever and - for the part where I could keep up - compelling, they're not especially well expounded, assuming as they do a familiarity with Cartwright's earlier work which it really isn't safe to assume. A greater faculty for expounding difficult concepts - such as that possessed by a Daniel Dennett - would have been an advantage here. Cartwright's is pretty leaden prose.

To the extent I understood it, Cartwright's programme really interested me: to invert the usual wisdom that scientific laws drive and explain physical events in the universe, and observe that physical regularities precede and therefore drive the composition of scientific laws - the laws are convenient models for making sense of pre-existing regularities, and not vice versa - but that even this is a step too far; that in order to even observe the regularities we need to devise "nomological machines" - a pretty phrase, I'm sure you'll agree - which prescribe the conditions in which regularities will be observed. We should talk in terms of capacities rather than regularities, though I couldn't really derive much more insight than that, despite repeated attempts.

The early chapters are just about manageable for the lay reader; after about half-way through I hit a brick wall when talk moved to the technical details of quantum theory, and never seemed to re-emerge.

It is certainly true that this book is beyond my grasp and almost certainly wasn't targeted at people like me, so those with the requisite background should disregard my vote and look into this book, but those more used to browsing the popular science section might want to steer clear. ( )