What usage arguments are persuasive? [closed]

Given the recent blog post that some kinds of subjective questions are OK on Stack Exchange sites, I thought I would try to formulate a good subjective question about usage to see what kinds of arguments people find persuasive and why.

So, for questions about a usage dispute:

1. What basis do you use for deciding what is correct or grammatical?
2. Why do you find some arguments persuasive and others not?
3. Does the pervasiveness of a usage affect your opinion of it?
4. What kinds of authorities on usage do you respect and who do you find to be not authoritative?

Solution 1:

This is a great question.

1. Combination of ear, what I've gleaned by formal study and by learning from colleagues (I'm an editor), opinions by informed people (see next points), and -- importantly -- context. Your question doesn't explicitly ask about this, but the appropriateness of usage will depend on whether we're talking about conversation, emails, cover letters for a job, or Supreme Court rulings. :)

2. Arguments are persuasive if (as in any context) they reflect knowledge and experience and provide specific and reasoned evidence for their POV. Perhaps the best way to answer is to address the opposite: arguments are not persuasive if a) they obviously reflect just a personal opinion ("I really hate it when people say _"); b) they provide no particular evidence (historical and/or practical); c) the person opining has no particular expertise or experience in thinking about usage; d) it's easy to find counterevidence among native speakers. For me personally, any argument that reflects a belief that English is in decline or is being "ruined" has no credibility, since that is a kind of litmus test for all of the previous.

3. Sure. How else would you establish usage guidelines? But again, it's about context -- maybe everyone says "I was just laying there". Arguing with your friends about it in conversation is dumb, but that doesn't mean one shouldn't respect the distinction between lie and lay in formal written communucations. The trick is taking into account the pervasiveness of usage in the context you're interested in.

4. Good: Garner's Modern American Usage [http://astore.amazon.com/theslotaspotforc/detail/0195382757]. Garner meets all the criteria listed above for persuasiveness. Good: the usage notes in AHD and M-W, ditto. Good: The style guide(s) from my work, which are written by professional editors and ditto. Bad: The Elements of Style, which is vague, is full of personal opinion, and effectively represents two people's opinions about "rules" that are over 100 years old and some of which were never real rules to begin with. Bad: Any reference that lays out "rules" without explanations based on actual usage, including examples. Bad: Any reference that purports to be anything other than the author's opinions about usage or that does not acknowledge that usage is contextual and to some extent arbitrary.

Solution 2:

I'm not a linguist but I do read Language Log and speak a couple languages. So here are my guidelines:

1. If it sounds natural to me (vague), if I can understand it, if I can parse it, or if I recognize it as an established usage then I consider it correct and/or grammatical. Some things are correct but don't appear grammatical, and some things are grammatical but don't appear correct.
2. If I don't immediately recognize it as grammatical, a reasoned logical explanation might convince me, in which case I'd revise my opinion, but otherwise an argument based on the actual usage would convince me. Other arguments I find unpersuasive because I've concluded that language arises from how people use it, and thus if you see it being used it's part of the language.
3. I'm not sure what is meant by "opinion" in this question. But the pervasiveness of a usage affects my acceptance of it as part of the language. There may still be problems with it; for example lots of people say "nucular" when they mean "nuclear". I consider that pronunciation wrong, yet accept that it's part of the language. Is that a contradiction? :)
4. I'm not informed enough to answer this question reliably. I trust Language Log because the writers there seem to be honest and they do not seem to be prescriptivists. But for all I know they could be lying to me. Language Log has conditioned me to reject Strunk & White and by extension be suspicious about any book which tells me how to write.