On 12 Dec 2014 23:27, "Alex Burka" <aburka@seas.upenn.edu> wrote:
>
> I just realized you were probably using square brackets for IPA. So if I understand correctly (relying on a Wikipedia chart here) [tuwitsku] would be {tu uitsku}, gliding from the [u] to the [i], and [tuʔitsku] is {tu .itsku} with a pause/glottal stop in between. As I understand it Lojban is of the opinion that [tuitsku] is indistinguishable from [tuwitsku].
>
> However, the debated ?{tuitsku} is pronounced [twitsku].
[w] is merely a necessarily-short subvariety of [u], and Lojban should not be making phonological contrasts based on segment duration, because such contrasts are not phonetically robust. That is, it's a bad idea for [tuitsku] to be ambiguous between /tu witsku/ and /twitsku/, and that ambiguity should be remedied and nullified by forbidding one of them. My choice would be to forbid word-initial /w, y/, but since that's not an option under consideration, it must be /twitsku/ that is declared illicit.
And
--