minimal pairs (was: PIE e/o Ablaut)

[Robert Whiting]

Date: Fri, 31 Mar 2000 19:17:46 -0800
From: Stanley Friesen <sarima at friesen.net>

At 09:00 PM 3/30/00 +0300, Robert Whiting wrote:

>>Most people would not insist on phonemic status for both [th]
>>and [dh] in English on the basis of this minimal pair (although
>>some would doubtless claim that there has been a phomemic split
>>similar to what occurred with /s/ and /z/).  This is because
>>otherwise the sounds are in complementary distribution, [dh]
>>occuring in voiced environments and in deictic words and
>>pronouns, [th] otherwise.

>Personally, I have trouble with this analysis.  "Deictic words
>and pronouns" is NOT what I would call a phonetic condition, so
>I would rule it out as a possible rule for governing allophones.

Yes, perhaps I should have said "many" people rather than "most"
at the beginning.  The fact that this is a grammatically
conditioned environment is preciesely the fact that leads some to
insist that [th] and [dh] must be phonemically distinct (despite
the fact that the only minimal pair that can be produced looks
more like a historical accident than a true minimal distinction).
Others have tried to explain the differnce as resulting from
stressed and unstressed forms and thus provide a phonetic
environment for the distribution rules.  The question becomes how
much grammatical information do you allow to affect the
phonology.

>One approach one can take in living languages is to check the
>speaker's awareness of the distinction.  Often an untrained
>speaker is unaware of true allophonic distinctions.  For
>instance, the aspiration/non-aspiration of voiceless stops in
>English is not generally even noticed by most speakers.  It
>usually has to be demonstrated to them before they can recognize
>it.

>On the other hand, most English speakers I know of seem to be
>quite aware of the /th/ vs. /dh/ distinction.

I'm sure that most English speakers recognize [th] and [dh] as
different sounds.  The question is do they recognize them as
different phonemes.  If you ask English speakers how the plural
is formed they will say that you add -s to the word.  Linguists
know, however, that what is added in most cases is not -[es] but
[ez].  This does not mean that the speaker is not aware of the
distinction between [s] and [z].  The speaker is describing the
spelling rule, not the pronunciation rule.  Now at one time /z/
was not a phoneme distinct from /s/ in English.  The
pronunciation was predictable from the environment.  The
appearance of the two as morphophonemic variants, however, led to
a phoneme split so that now /z/ is recognized as a phoneme, and
so phonemic status is found only in words of recent origin, most
of which are expressive in nature (sip ~ zip; sap ~ zap; fuss ~
fuzz, etc.).  The phones [th] and [dh] are in much the same
relationship as earlier [s] and [z], occurring in predictable
environments (although not always predictable on a phonetic
basis) or as morphonemic variants.  However, there are no new
coinings where [th] opposes [dh] so one suspects that the
speakers do not consider them separate phonemes (yet).

Bob Whiting
whiting at cc.helsinki.fi