Re: more English orthography

Tom Wier wrote:

>Marcus Smith wrote:
>
>> I guess it's a matter of preference whether you want to consider schwa a
>> phoneme or not.  I can't think of a way to get any other pronunciation for
it
>> in "comma."
>
>Usually the only way to prove that what appears as schwa on the surface
>is underlyingly some other sound is to show that it changes to something
>else when you add on a morpheme or two.  Cf. telephone ['t_hEl@foun]
>vs. telephony [t_h@'lEf@ni].
Absolutely.  But I can't find a morpheme that sounds reasonable attached to
"comma" so I don't think there is a way to be sure what the underlying form
should be.

Besides, that is not necessarily the best means.  I've done "wug testing" on
this matter.  For those who don't know what wug testing is, it is when you
invent a word, tell a speaker that it is real word that they don't know, then
ask them to pronounce inflected forms.  If there is a schwa in the surface
form
of a made up word, then the speaker cannot know what the "underlying" form is
supposed to be.  So if I give him the adjective [h{l@k] (SAMPA) and ask
them to
say the word with -ity suffixed, he could potentially say [h{lVkIti],
[h{lIkIti], [h{lEkIti], [h{lQkIti], or [h{l{kIti].  In fact, the people I
tested were consistent in responding with [h{lVkIti], far too consistent
for it
to have been an arbitrary decision.  Since they are willing to provide a form
for these when there could not possibly be any stored in their head, then I
think we have to be suspicious of any pronunciation of an uncommon word.

Marcus