"ong" in American cot-caught nonmerged speech.

DOLL /Q/ may be due to L coloring (dark L makes the preceding vowel rounded in some accents), affecting words like ''culture, pulse, involve, doll, bull, wolf''

<<DOLL /Q/ may be due to L coloring (dark L makes the preceding vowel rounded in some accents), affecting words like ''culture, pulse, involve, doll, bull, wolf''>>

Yep. It's certainly due to that.

Well there have been a number of different conditions that lead to /Q/ (or /A:/) shifting to /O:/. In the original lot-cloth split found in Britain, it was following fricatives; in the extended lot-cloth split found in the US, it was a following /g/ or /N/; and more recently in American English, it can also be a preceding /w/ or a following dark /l/. The linguistic facts speak for themselves; but trying to determine which of these changes are part of the lot-cloth split and which aren't is really just an arbitrary choice.

But perhaps as a standard, we could say that the lot-cloth split consists of the shifts that occured before the father-bother merger (that is, from /Q/ to /O:/) and that later shifts of merged /A:/ to /O:/ (which I think would include cases like "wash" and "doll") are a separate phenomenon.