1. Scott Thornbury (2004) Natural Grammar: The key words of English and how they work. Oxford: OUP.
This book claims to look at grammar in a fresh new way. The book selects 100 key words and examines the structures and collocates commonly found with them. The alphabetical list starts with a, includes back, can, did, for, get, if, and through, and finishes with you. Scott uses the results of concordances done on big corpora to identify high-frequency words in English and their collocates, and offers explanations and practice exercises for each one. He argues that “by learning these high-frequency words and their high-frequency patterns, the learner is getting traditional grammar ‘for free’ as it were.”
There is nothing new in this “lexical approach”, as Scott acknowledges; Michael Lewis made a dog’s dinner of it, and the more reasonable Willis team both tried it and failed miserably. But I fear that Scott’s book is the worst of them all. It manages to combine boring, unhelpful detail with more boring, unhelpful detail, with nothing to relieve the boredom whatsoever.
I opened the book at random and I got the page dealing with the key word or. Five “Grammar patterns” are listed, one of which is:
“verb I + whether [ … ] I + I or + [ … ]”, as in Do you know whether the plane has arrived or not?. We learn that the verbs which most frequently precede whether are know, decide, see, and wonder, and that “set phrases” include an hour or so, two or three, and Is she coming or not? Wow!
Why do we need a grammar of the word or? In fact, why do we need any information about the word or? There’s an argument for substituting an explanation of the comparative for an explanation of more, which is the point that Biber, Johns, and others made 20 years ago, but there’s no justification for this terrible mixture of boring words like or and words like have, has and had, Scott’s treatment of which can hardly be said to provide an explanation of the English perfect tenses.
Most words at the top of frequency lists depend on their collocates for interest, i.e., they themselves are boring. To use them to get some “free” grammar seems totally perverse to me. I’d rather read Murphy’s dead prose than read this stuff. Scott’s treatment of back doesn’t include the “set phrase” back to the drawing board. He should have gone there and thought of something else to boost his publishing tally.
Actually, what Scott should have done is pointed people to the enormous fun and knowledge to be got from concordancing, which he fails to do. Why didn’t he suggest that you can use a concordancer to search a corpus for the phoneme, word or phrase you might be interested in? If you do a concordance on any you’ll see that it’s most frequently used in the affirmative; if you do one on set you’ll see that it’s got over 200 uses, and so on. Go to the British National Corpus site http://sara.natcorp.ox.ac.uk/lookup.html and try it out, but stay away from this book, unless you have trouble sleeping.