At the end of a recent post ("Opening up the Google-China mailbag", 1/14/2010), James Fallows adds this "language note":
Usually when quoting reader responses, I leave them just as they are, warts and all. But if I am sure that the note is from a non-native speaker of English, I will sometimes correct small mistakes of spelling, grammar, or usage — "have" for "has," "hypocracy" for "hypocrisy," — that would unduly draw attention to themselves. In this note I made three or four of these tiny copy clean-ups while leaving the rest of the phrasing and word choice unchanged. Writing in a second or third language is one of the harder intellectual challenges that exist. (Hey, writing in a first language is not always that easy!) Even though English has a larger share of non-native speakers and writers than any other language and therefore a greater tolerance for "diversity," I think it's justified to remove minor brambles from the writer's path.
I admit that this practice leaves a logical gray zone. If somebody seems to be a native speaker who just writes sloppily, I don't bother trying to save that person from himself. But if I quickly get the sense that this is not a native speaker — and within a sentence or two I think I can always tell — I may do a little cleanup. The gray zone is when the command of grammar is shaky enough to raise questions, but not unusual enough to suggest that the writer grew up with a different language and therefore deserves affirmative-action help. This is all part of the endless saga of language being one of the most absorbing aspects of dealing with different cultures.
This is at the other end of the spectrum from the use of "eye dialect" to undermine the credibility of a quotation, or at least to put the author in his or her social place.
[As someone who frequently commits typos and brainos in writing, I silently correct such errors in comments when I notice them, regardless of the author's presumed linguistic history.]