Over a decade after its first publication, Fodor's book is still a classic. At a reasonable 162 pages, it is slim, crammed with compelling argumentation, and leavened with irreverent wit. Its lessons remain as important today as they did a decade ago, and the book repays many re-readings. It is necessary reading for anyone with an interest in concepts. I can't recommend it enough.
However, two caveats are in order.
First, this is not a good introduction for the uninitiated. Although Fodor's book is thankfully light on formalism, which makes it easier to read than a lot of work on concepts, the book's argumentation is dense and presumes a lot of understanding of theories of concepts. Do you understand the difference between theories that take mental particulars as basic and those that take mental dispositions as basic? Do you know what the semantic features of lexical items are (e.g. [+/-telos])? If you don't, then this book will be heavy slogging, because an understanding of these ideas is presumed, not explained, in Fodor's book. I first read this book as part of an undergraduate course on concepts, and it was very difficult reading for me. It wasn't until a few years later when I had read the book three or four more times and had done a lot of background reading that I finally started to understand Fodor's arguments and their implications.
Second, as the other review by "daorcboy" suggests, this book does not do a good job of explaining what is compelling about Fodor's favored hypothesis about concepts, except that this hypothesis is supposedly better than its competitors. Nor does the book do a good job of explaining what is compelling about Fodor's competitors in the first place. This feature of Fodor's presentation not only makes it difficult for neophytes to understand what is at stake in Fodor's discussion, but it also leads to some serious distortions in the presentations of the issues. Here's an example that might illuminate this feature of Fodor's presentation, but it is only an illustration and can't represent the range of issues at stake (or do them full justice): consider the N-word and 'black person" (I'll be treating these words as stand-ins for the concepts they express; nothing important turns on this simplifying assumption). I hope that none of us would use the N-word, but all of us would find 'black person' acceptable to use, presumably because we think these terms have different meanings. According to Fodor's account, however, the N-word and 'black person' are literally *synonymous* because they have the same reference. Most people would find this an uncomfortable conclusion, but Fodor doesn't even acknowledge how uncomfortable this implication is, let alone that it is an uncompelling feature of his view, or that it is a prime motivation for alternative views.
No, instead, Fodor just assures us that his view must be true because all the alternatives are worse, primarily due to abstract considerations involving semantic productivity and systematicity, not to mention the horrors of semantic holism that follow in the wake of alternative theories. We are also given assurances that whatever differences there are between the N-word and 'black person' are due either to the pragmatics of speech (a la Grice, or more recently, Tim Williamson), or are incidental insofar as they are only features of these terms' different modes of presentation, which, Fodor tells us in his previous book, "The Elm and the Expert," are only associated with their semantic/referential properties by a miraculous pre-established harmony (a la Leibniz). If you get the sneaking suspicion from the characterization I'm peddling here that important explanatory burdens are being shirked by Fodor's theory, then you aren't alone among cognitive scientists and linguists: as linguist James Pustejovsky puts it with reference to Fodor's manner of defense, "anything can be explained by appealing to a general enough mechanism, with the subsequent lack of theoretical interest or scientific merit." And so if this chacterization of Fodor's theory doesn't sound that compelling, then you should suspect that there is more to the story than what makes it into Fodor's book.
After reading a few of Fodor's articles and books, any reader will start to discern a pattern in Fodor's style of argumentation. Fodor thinks concepts are unstructured mental particulars whose semantic content is directly determined by their reference; anyone who disagrees with Fodor's thesis is, to put it crudely, a closet behaviorist such that no matter how many epicycles they add to behaviorist theory, they will just be repeating the same old behaviorist mistakes. In many ways, Fodor's analytical pattern is accurate and yields important insights about the shortfalls of some theories, but it is also too simplistic. An analogy might help. Imagine a free-market economist from the University of Chicago who argues that anyone who doesn't agree with ultra-libertarian free-market philosophy is a closet communist; anything short of orthodox purity is apocryphal. Keynesians? Communists. Advocates of Prospect Theory? Communists. Behavioral economists? Communists. Whatever ounce of truth there is in such allegations obscures a lot more than it reveals, and the same is true for Fodor's repeated insinuation that other researchers (virtually all of cognitive science, apparently!) are closeted behaviorists who just haven't given up the ghost of behaviorism. This argumentative strategy gets old. It's not just that its predictable hole-poking manifests an unscientific spirit (because scientific theories are always interested in expanding their explanatory power or precision), but it is somewhat self-defeating, since it gives Fodor's theory nowhere to go, no ways to be improved or deepened. No doubt this is why "daorcboy" was dissatisfied that Fodor didn't have more to say about his own theory.
The foregoing caveats shouldn't put anyone off to reading Fodor's book, which has taught me a tremendous amount. But, like any book, it should be read with a grain of salt and only after you've read a more balanced introduction.