I meant to write a post a couple of weeks ago on this story about accent bias. It reports that Essex accents were judged 'less intelligent' than others in a survey - though it seems the study only looked at London and home counties accents and didn't include other accents, such as Brummie or various northern accents.
Accent bias can not only unfairly affect an individual's job prospects, but even democratic politics.
I didn't get round to a post at the time, largely because much of my time has been spent on marking lately, but I was reminded of that study by this column on mispronunciation. The author is professor of phonetics and has lots of interesting things to say about why certain mistakes come about (e.g. 'expresso' instead of 'espresso').
However, the last section of the column moves from how language changes to making normative claims about prejudice and discrimination. "Correcting pronunciation can actually be an act of linguistic prejudice." I suppose it could be, but it isn't clear that it has to be or even that this is the normal case. Though she acknowledges that correcting a language-learner may be different, it seems that how and why someone corrects someone else makes a big difference here.
She ends by remarking on 'accent prejudice' which is indeed alive and well, as demonstrated by the BBC story I mentioned earlier, but it's something of a non-sequitor. There's a difference between who someone pronounces a word differently, because of their accent (which they will share with a wider group), and idiosyncratic mispronunciations. You might think that there's something objectionably elitist about correcting even the latter, but that doesn't follow from the fact that it's wrong to correct the former.
(For the record, I was born and raised in Essex, but I'm not sure I really have an Essex accent any more. On the other hand, I'm regularly unsure how to pronounce words that I'm more used to seeing in print.)