R A Brown, On 18/11/2011 08:56: > On 17/11/2011 20:52, BPJ wrote: > > So why can't /paa/ be [pa.a], [paha] or [pa?a] ad lib? > > No reason at all, that I am aware of. > > The problem is that as things stand at the moment, it seems to me > that neither ['pi.a] and ['pija] nor ['pu.a] and ['puwa] will > realistically be distinguished. The only way of doing that is to > insist that /a/ _must_ be preceded by [h] or [?] or some other sound. > If it must be so proceeded then that sound is surely no longer > 'ad_lib' but assumes phonemic status. To me it (i.e. [?a, ha]) would look like an obligatory allophone of /a/. --And.