I know this isn't like French that has a somewhat recognize but debatable effective governing body.
We do have the OED, Websters dictionary, the various style guides like the MLA, Chicago Manual of Style, etc.
If we are all just going by ear, how are we to avoid a lot of "to-may-toh"/"to-mah-toh" discussions?
I thought about this quite a bit, and formulated my thoughts long before we went into beta. Posts should be encouraged which use one or more of the following lines of evidence:
Not every question will be addressable in one of these ways, but questions of "correctness" should preferably be backed up by something off of this list.
Conversely, the following lines of argument should be discouraged and downvoted:
This information should be put in the FAQ.
My preferred authorities are posts on Language Log and citations in the Corpus of Contemporary American English. Following that, the Oxford English Dictionary and the Merriam-Webster Collegiate Dictionary.
Each authority can have weight in a discussion, but if someone appeals to e.g. Garner's Modern American Usage or Fiske's Dictionary of Disagreeable English as some kind of decisive authority rendering opposing opinions (based on evidence of actual usage) moot, they shouldn't be surprised to get downvoted by me.
The problem is that all the dictionaries mentioned change every couple of years as more evidence comes in from corpus studies. And the beauty of English is that there is no single authority. In fact many academics now speak of "Englishes" rather than "English", because there are so many varieties, and they are all correct in their own context.
nohat has listed lots of appropriate references. I'm a fan of Language Log and OED myself. I also think Google is a great tool, at least for comparison purposes.
I don't think a reference to one of these is necessarily important, though it lends credibility. More than anything, I think an appeal to one of these sources should provide the ability to find more information on the topic.
I also think we will find at least some situations haven't yet been addressed by these credible authorities. In these cases, do you reference academic papers if possible? How far should you attempt to answer a question on your own—can tweaking and reviewing a home-grown answer be a community responsibility? I would say yes, especially so in the case of grammaticality judgements, where native speaker intuition is infallible.
Subjectivity in areas of debate will be unavoidable, and permissible to some extent. E.g., "How do I pronounce x" would be fine, but "is this better than x" is obviously something to be closed. I think StackOverflow's rules would work fine: "Avoid asking questions that are subjective, argumentative, or require extended discussion."
