N. M. J. Woodhouse's comparatively short Special Relativity is a pleasure to read and therefore qualifies right off as a good source to use for learning about special relativity on your own. A lot of very nice material is touched on in its pages, presented in a natural sequence consonant with history, and is not improperly belabored. It's also rather informal in style. One gets the sense of breezing along pretty fast while, in actuality, a lot of material is being dealt with. So, an autodidactic reader had better be prepared to cover the book with marginal notes, to fill a thickish note-book with more carefully worked-out material, and to do most if not all of the exercises. The latter come in several flavors, by the way, including, by the author's own admission, quasi-standard "borrowed" problems (from standard older texts) and Oxford examination problems. It's meaty stuff.
The book does presuppose a certain amount of mathematical maturity in its audience, but not unduly so: a strong background in vector calculus is certainly called for and (very properly, I think) no apologies are made for matrices. Indeed, Woodhouse states in his preface to the book that "it is written for students who are more familiar with linear algebra than with electromagnetism." And this says a little about the reader or student for whom Woodhouse intends the book. Qua audience, the real question is specifically one of distinguishing between two ultimately different kinds of animals, the physicist and the mathematician who wants to learn some modern physics. Admittedly, this is a somewhat controversial business, but I propose that one tell-tale feature in this distinction has to do with the kind of prose one is comfortable with: mathematicians, I claim, hunger almost irrationally for theorems and proofs, very explicitly. Since Woodhouse's book has lots of theorems and proofs, it's clearly meant for mathematics students.
As I already indicated, the selection of topics in the book is very nice indeed, and is historically sound and will therefore reward the reader with an element of culture, to boot: he'll learn some history of modern physics. Classical mechanics and Maxwell give way to the properties of light, Einstein and Lorentz. The latter's famous transformations are dealt with in a general way so as to precipitate a marvellous treatment of Minkowski space, setting the stage for a discussion of the geometry of space-time. Then relative motion, relativistic collisions, and finally relativistic electrodynamics (!) are discussed, and the book ends with a chapter on tensors and isometries done very tersely but also very well.
I wish this book had been around when I was a student.
Michael Berg is Professor of Mathematics at Loyola Marymount University in California.