That sounds very pompous as a title. But I’ve been thinking about this recently because of teaching intro linguistics. We have been talking about IPA and English transcription, and in particular we’ve had a lot of discussion about English back vowels.
Now, the students learn IPA and something about phonetics before we talk about phonemes, and before anything much about dialectal variation and sociolinguistics. We’ve had quite a bit of confusion. I have fairly standard Australian English vowels (although maybe my back vowels are a bit more rounded than standard, and my diphthongs are a little different). But I have a fairly clear length distinction along with the tense/lax distinction (so my /i/ is clearly much longer than my /ɪ/, for example, and I would transcribe it [iː]). We are using Finegan’s textbook (Language: its structure and use), which presents a variety of I guess you’d call it Standard Average American. At least half of my students are from Texas, though, and have a rather different vowel system. They regularly merge [ɑ] and [ɔ], for example, and there’s quite a bit of monophthongisation of /aɪ/.
This is, of course, an excellent illustration of why IPA is useful, and why orthography doesn’t let you capture differences in pronunciation, but it still adds a lot to the steepness of the learning curve for students who are still pretty new to the idea about thinking about language in this way. When I was an undergrad we just learnt standard Australian English and were graded wrong if, for example, we wrote a w instead of a velarised l for words like “pool”. But I don’t want to teach phonemic transcription when we’re supposed to be doing phonetic transcription! One solution is to do what some of my colleagues do, and do a unit on sociolinguistics and variation first, before the more technical parts of the class, but I don’t want to do that. I like being about to talk about variation and sociolinguistics after they know something about IPA and phonemes and morphology and syntax, because I think it makes it easier to see why linguists talk about socially stigmatised speech varieties as not being “bad speech” or “lazy ways of talking”.