Rachel Klippenstein wrote: > Do voiced velars tend to cause an i-type off-glide on > a preceding voewl? I think I may have one in leg - it > doesn't sound like quite the same vowel as in ten pr > bed or peck, and more definitely, I have an i-like > off-glide after & (=ae digraph) before [g] and [N], so > I have a different vowel in bag and bang (something > like [&I] or [&i] than in back or bad (just plain > [&]). (My brother even tends to say these with a > vowel something like [eI], so bag is for him [beIg] > Yes. You're likely to get a transitional vowel sound any time the tongue has to move from one extreme position to another, as from low front [&] to bunched-up-in-back [g]. Also, the fact that Engl.stressed vowels are lengthened before voiced final consonants abets this, and in a sense the glide-vowel is substituting for part of the length-- so you don't hear much of a glide before a final k. Your particular examples suggest there may be a little Southern-US influence; is that possible? The exact quality of the glide probably depends on the individual; personally, I have more of a schwa-glide in the case of [...&g]-- but like you, an i-glide in [...&N]. Here the front & causes the N to move forward to a position very similar to [i] or [I] (contrast "bong"-- no glide, because the tongue only moves slightly up from low back [O] to [N] and the N is a true velar, much further back than in "bang").