1) That the '-ing' present progressive and 'do' constructions in English come from Celtic languages. I had twigged to the former while dabbling in Irish last year; but it seems that both Welsh and Cornish also use the verb 'do' in the equivalent of questions and negatives - Do you agree? I don't agree. McWhorter argues that English should be regarded as having a strong element of Celtic grammar as well as Germanic at its root. I basically agreed with this already but he goes on a bit too much about it, almost accusing the likes of Crystal and McCrum of deliberately ignoring it. A side argument is that written Anglo-Saxon is a poor guide to how it was spoken.
2) That there is no harm in linguistic innovation and no absolutely right way to write or speak English. That's his privilege as a linguist, though in fact denies the interesting social process by which some innovations do become acceptable.
3) That the drop-off in inflection of English nouns was caused by mixing with Old Norse from the Viking invasions. Not a lot to say on this, though he says it at length, and I'm pretty sure I had heard this before.
4) That the Sapir-Whorf theory is rubbish. He doesn't spend as much time on this as Steven Pinker, but it is after all stating the bleeding obvious.
5) That the differences between Proto-Germanic, as reconstructed, and the other Indo-European languages might be explained by its speakers around 500 BC having been strongly influenced by a Semitic language. This was a new and fascinating idea for me. McWhorter says he is quoting a German scholar, Theo Vennemann (who also argues for hidden influence from Basque on most Northern European languages including the Germanic ones), and that there are three key observations here:
i) the shifts from stops to fricatives of p -> f (pater/father), t -> th (tres/three) and k -> h (canis/hound) would fit with contact with a language which had a lot more sibilants, as Semitic languages do. On the surface this is the weakest of the points, but it's slightly stronger than McWhorter realises - he notes without further comment that the only other Indo-European language group to have undergone that kind of shift is Armenian, but in fact Armenian is demonstrably geographically close to various Semitic languages and for all I know has more evident signs of contact;
ii) vowel shifts to mark tenses - sing/sang/sung, ride/rode/ridden - a mutation found universally in Germanic languages, not at all in other Indo-European languages, and universally in the Semitic languages (at least between present and past tenses)
iii) an area where again I think the evidence is stronger than McWhorter allows, the etymology of some of the words that are found in Germanic and not in other languages is not a bad match for Semitic: "fear"; being /furkhtaz/ in Proto-Germanic and /p-r-kh/ in Proto-Semitic; the word for a group of soldiers that became our "folk" and Hebrew פלוגה ("detachment") from Proto-Germanic /fulka/, Proto-Semitic /p-l-kh/; the Germanic word from which we get "over" and German and Dutch get "Ufer" and "oever" meaning "shore", linked to the Biblical Hebrew ʔeƀer/עבר, a root with various travelling and crossing-over connotations which is itself the source of the word "Hebrew".
McWhorter doesn't even mention the British Israelites, who will love this theory once they find out about it, but instead postulates contact through Phoenician maritime exploring, for which there is some archaeological evidence, though more such evidence may be drowned on the floor of the North Sea. Anyway, I love this theory, and someone should write a novel about Phoenicians getting lost among the swamps and bogs of Scandinavia. (Harry Harrison, if memory serves me right, already did one suggesting such a lost mariner getting involved in the building of Stonehenge, but that would have been a bit earlier.)
Anyway, a book that is slightly uneven in style but provocative and well-sourced. Great for those of us who are interested in that sort of thing. There's a lot of metaphorical reference to other languages beating English up, reminiscent of james_nicoll's famous statement, but different in that McWhorter is discussing grammar rather than vocabulary and portrays English as the victim rather than the perpetrator.