Kirk wrote:Travis B. wrote:Kirk wrote:Yes, your data are certaintly interesting. As you said, I can't find rules that work as consistently as they do for my speech. I was somewhat surprised you pronounce "golf/gulf" the same. You don't do the same for "doll/dull" right? Or do you?
I myself pronounce
doll as [dɔːɫ] and
dull as [dʌːɫ]. The main thing here is that the change of historical [ʌ] to [ɔ] can only occur if there is another consonant (and I would suspect an obstruent at that) after /ɫ/, but even that is not consistent, as shown by my aforementioned examples. One note though is that there is
consistently [ɔɫ] for what was most likely historically [ɑɫ] in my dialect. Furthermore, there is an interesting alternation for what was most likely historically [ɒɫ] where such has been shifted in a syllable structure-dependent fashion to either [ɔɫ] or [ɑɫ] depending on whether such is split by a syllable boundary, which has been morphologically frozen in different related wordforms based on the same root, such as
doll and
dolly, which for me is [ˈdɑːɫi].
Yeah, I remember talking with you about that phenomenon in your dialect before. Really interesting.
Yeah, it's quite interesting to me as well. The first thing is that it has to be a quite old set of shifts, by NAE standards, as the upper bound on it, time-wise, is the point where the
father-
bother merger occurred in the dialects from which the English here today originated. My guess is that it took the form of three different shifts, two happening at around the same period of time, and one happening later. These would be [ɑɫ] -> [ɔɫ], [ɒɫ] -> [ɔɫ], and [ɒ] -> [ɑ], with the order of the first two being unclear, but definitely occurring before the father-bother merger.
Edit: One note I forgot to add is that such sound changes other than the father-bother merger likely originated in the present-day UK, since there are many dialects there which have similar vowels in the same positions as me here, even though they don't necessarily use the same vowels; for example, many English English dialects have [ɒɫ] where I have [ɔɫ] corresponding to likely historical [ɑɫ], indicating that it likely was a related shift, which most likely occurred in the same general time frame, which just happened to have a slightly different result.