I posted yesterday on my blog (though the posting was mysteriously dated 12/28 by WordPress) about what looks like a whimsical proscription from Ambrose Bierce, who in 1909 instructed his readers not to use
Because for For. “I knew it was night, because it was dark.” “He will not go, because he is ill.”
Jan Freeman pointed me to this "rule" in Write It Right. She and I have been unable to find it elsewhere (in the 19th or 20th centuries); it seems to have been an invention of Bierce's, a concocted usage rule — like the ones that Freeman discussed in an entertaining recent column entitled "Rule by whim".
Meanwhile, in a comment on Freeman's latest column (on usage advice that hasn't aged well) I came across another candidate for whimsical proscription, against complement clauses missing the complementizer that.
As I noted in a posting to the American Dialect Society mailing list earlier today, when you write about usage advice, readers almost always take it as an invitation to unload their pet peeves in public. So it was in this case. Commenter "HughMann" had among his gripes:
I know he is a good man … No, I know that he is a good man ….
Now, many usage handbooks caution against "omission of (complementizer) that" where a temporary ambiguity could result (because the subject of the complement clause might be interpreted as the object of complement-taking verb), as in
I know Kim will come.
(In actual practice, that is very often omitted even in the face of such a temporary potential ambiguity; the intended reading is clear from context, real-world plausibility, discourse structure, etc.) Some handbooks also note that that isn't omissible with certain complement-taking verbs, like add:
Kim added that they would come. but ??Kim added they would come.
But the commenter's example is a case where no temporary ambiguity is possible and where the verb (know) is one that generally allows omission. It looks like the commenter believes that that omission is never acceptable — a preposterous proscription I've never encountered. And a claim that is flagrantly at odds with actual practice. (In fact, Mark Davies noted on ADS-L that in the Time magazine corpus at BYU the frequency of complementizer that (versus zero) has been steadily and significantly declining since the 1960s. Similar declines have been observed in other corpora.)