Matt Pearson wrote:
>> And "bath": "bathe". Though IMHO derived forms are weak reeds as
minimal
>> pairs.>
>"Bath" and "bathe" aren't a minimal pair, though, because of the difference
in
>vowel quality.
Aargh. Me brain hurts. I must start taking that gingko stuff again.
>I don't see why derived forms should be any less reliable for determining a
>phonemic contrast than other pairs. As long as (a) speakers are consistent
in
>which sound they use for the base form and which they use for the derived
>form, and (b) the sound in the derived form occurs elsewhere, and is not
>restricted to a particular morpho-phonological pattern.
(b) is the point I had in mind, and it does happen (at least in classical
phonemics). Save for that handful of _real_ minimal pairs in Engl., it's
almost the case with the /T/: /D/ contrast.