Geoff Nunberg has an interesting take on why, despite the proliferation of books explaining to the public at large how language works (I was happy to see that he name-checks my man Robert A. Hall), people persist in believing all manner of nonsense, “from their conviction that African American Vernacular English is slovenly and without rules to their certainty that Elizabethan English persists in Appalachian hollows.” He uses Amazon.com’s “customers who bought this book also bought” feature to suggest that people who buy books by real linguists buy other books by real linguists, while those who buy books on “proper usage” and “better English” stick to that sort of book and thus are not exposed to more enlightening material. Sad if true, and I have to admit it sounds plausible.
While I’m on the Language Log Geoff beat, I should mention that Geoff Pullum has called a halt to his quest for “attested cases of coordinate structures… with large numbers of coordinates,” having been sent a reference to a coordinative listing of 433 different kinds of cod (“ball-bag” in the Penguin translation) in Book 3, Chapter 28 of Gargantua and Pantagruel. His conclusion:
The class of all English sentences may be regarded as indefinitely huge though ultimately finite (like the set of all pine needles) or actually infinite (like the set of all integers), but either way, one thing is clear: no grammar that names a specific number as the maximim number of coordinates permitted in a coordinate structure is a correct grammar.
I regret to say that an examination of both the original French text of the chapter and the Urquhart translation does not show a coordinate structure with a conjunction, or even a sentence at all, but merely a list of names, each treated as a separate entity, so unless the Penguin translation (which I do not have) renders it as a long sentence with “and,” the example should be disqualified. The conclusion, however, is irrefutable.
Update. I lied: it turns out I do have the Penguin translation, and it too has a list in columns with no coordinating conjunction. I rest my case.