In my reading experience, John McPhee is incapable of writing a bad paragraph -- although I admit I didn't read him during that decade or so when he wrote about nothing but geology -- so this collection of pieces about his own writing process was, as I expected, interesting and elegant, full of comical anecdotes about his experiences with interview subjects and with New Yorker editors and fact checkers. [Two examples of the latter: (1) One of his earliest New Yorker pieces, later turned into a book (the first of his I ever read), was entirely about oranges, and he called it "Oranges." Duh. Legendary New Yorker editor William Shawn, however, had a general rule that the title of a piece should not be the name of the subject, so he changed the title to "Golden Lamps in a Green Night." After McPhee threw a fit, it was changed back to "Oranges." McPhee uses this example to illustrate his dictum that the title of a piece "ought not to be written by anyone other than the writer of what follows the title. Editors' habit of replacing the author's title of one of their own is a like a photo of a tourist's head on the cardboard body of Mao Zedong." (2) At the end of a six-page shaggy dog story about the fact checking for a piece he wrote about the Swiss Army, detailing the difficult task faced by his fact checker (Richard Sacks), who spent weeks telephoning obscure Swiss authorities and questioning them in English, French, and/or German as necessary, detouring into an anecdote about how it was Orson Welles, not Graham Greene, who thought up his character Harry Lime's famous Ferris wheel speech in the movie The Third Man (". . . and what did [they] produce? The cuckoo clock."), McPhee concludes by noting that after the piece was published, first in the magazine and later in a book collection of some of his pieces, it was more widely read in Switzerland than he had anticipated, and he braced himself for a swarm of letters from Swiss readers pointing out errors he had made. But, he says, he never heard of a single error in the English edition of the piece. Consequently, when McPhee recently spoke to Mr. Sacks, who is now retired from fact checking, he told him how impressed he was that in over 30 years no Swiss had ever sent a corrective letter about that story. Punch line: "This fact did not check out with Richard. 'Oh, but there was one letter,' he said. 'Something about a German word, but the reader was wrong.'"]
If you like McPhee (and aren't looking for a geology fix) this is well worth reading, although some of the writing advice he offers here is less than universal. (The chapter about "Structure," for example, contains a number of his own hand-drawn diagrams outlining the structure of some of his pieces; they are weird and elegant and intriguing, particularly if you remember the piece he's describing, but also somewhat inscrutable and maybe not that useful as guidance for most writers, I'd suspect. The chapter also contains a longish -- and to me, rather tedious -- section about the editing software he's used for decades but is now heading towards obsolescence.) But that's a quibble; there's a lot of great advice in here too. Ultimately, though, I think the reason I only liked this book and didn't love it as much as some of McPhee's other work is that writing just isn't an obscure or unknown topic to me; McPhee's advice and anecdotes are interesting, but I've read others that are equally so. Whereas in his other work, what McPhee can do brilliantly is take some obscure or arcane subject -- a place, a craft, a profession, even ORANGES! -- and somehow, magically, make it seem like the most fascinating topic imaginable.