I was at first a bit unsure as to whether to give this book a full 5 stars. Sure, I liked it a lot, but it does need some negative comments here. But I decided I liked it enough to give it 5.
At first, I felt that the book was not as well integrated as I'd like. With further reading it grew on me. McWhorter covers a lot of topics in comparative/historical linguistics, and does so in a very readable style, but the breadth of his coverage seemed to me at first to be just too great. As I said, as I read further, I changed my mind somewhat. I still think that it could be somewhat better integrated, but I think it's not as badly integrated as I first thought.
One thing that I can fault McWhorter on is that if you have read his other book, "The Word on the Street," you'll find him repeating some of his ideas. They aren't at the same level of prominence: major theses he advances in the earlier book become minor points here. But it does seem he could at least change his examples. He gives the SAME example of a sentence that is traditionally considered ungrammatical which everyone naturally uses, and the SAME example of a passage in Shakespeare that is universally misunderstood because of semantic change over time, that he used in the other book. But he's not as bad as Keith Devlin, who has published popularizations of math with whole chapters taken verbatim from earlier books. So again, I can't fault McWhorter that much.
Other than these two comments, however, I have only positive things to say. I think that this book is a good treatment of historical and comparative linguistics, of dialect variation and pidgin/creole structure, and such at a level accessible to the interested general public. And so I recommend it to all interested readers.