Maggie's Farm

If you're just saying that starting a sentence with "and" or "but" looks ugly to you, I can't argue about that. But you seem to be saying something more -- that it is "not" "acceptable" "in formal English." Could I ask what basis there is for such an assertion? Are you appealing to some supposed linguistic authorities, and, if so, which ones? I ask because the authorities that I have seen say that there's nothing wrong with this usage.

My question is whether the assertion that "a conjunction joins two related thoughts in a sentence" is indeed an authoritatively recognized rule of grammar (as opposed to a rule that your teacher might have thought is so recognized). Webster's Dictionary of English Usage (1989) asserts that this is not such a rule, but rather "[p]art of the folklore of usage," which "[e]verybody who mentions this question" rejects. (I assume "everybody" here means all the usage authorities consulted by the Dictionary's authors.) The Oxford English Grammar (1996) doesn't mention this controversy, but lists -- without any criticism -- a sentence that starts with "But" as one of its examples. Harper's Dictionary of Contemporary Usage (1985) does report that starting a sentence with "but" was "a practice deplored by Victorian grammarians," but doesn't agree with those grammarians, and doesn't suggest that any other respected modern authorities agree with them. Garner's Dictionary of Modern American Usage (1998) calls the claim that a sentence may not start with a "but" "a gross canard," and indeed says that starting a sentence this way "is highly desirable in any number of contexts," citing several other authorities.

And of course my original post also cited the New York Times and the Supreme Court, which commonly use the sentence-initial "but," and which are usually seen as at least relatively formal publications. So my question remains: As between the teachers who taught you the rule you describe, and the authorities that say that it is not a rule, why should the word of the teachers be treated as the defining standard of what is "acceptable"? As a descriptivist, I think that what's acceptable is determined by actual educated English usage. A prescriptivist might think that what's acceptable is determined by the authorities. But what is the basis for rejecting both actual usage and the authorities that I cite?

I agree that this probably doesn't merit much more time. Still, let me note that your argument seems to assume the conclusion -- that a conjunction must connect related thoughts in a sentence. But it seems to me, from the usage examples and from the authorities, that a conjunction connects related thoughts, including related thoughts in two separate sentences. To quote the Oxford English Grammar, "co-ordinating conjunctions," such as "and," "or," and "but," "link units of equal status"; and the examples show that these units may be sentences, and not just words or clauses.

So why should we decide what is "acceptable" based on the definition you assert, as opposed to the definition that seems to be assumed by the authorities and by actual usage?

Weaver: I haven't yet had a chance to check out the Azar works. But even if Azar does say that it's improper to begin a sentence with a conjunction, I've cited many considerably more prominent authorities -- including the Oxford English Grammar and usage dictionaries -- that says that it is proper, without any limitation to informal use. Why should we judge correctness based on the one source you cited, rather than on the several sources I cited?

I ask this not so much because of a concern about the conjunction-at-the-start-of-a-sentence question, but because I'm curious about how people form opinions and make arguments on these sorts of questions more generally.