minimal pairs (was: PIE e/o Ablaut)

[Robert Whiting]

On Fri, 21 Apr 2000, Stanley Friesen wrote:

> At 09:38 AM 4/14/00 +0300, Robert Whiting wrote:

>> On Fri, 07 Apr, Ross Clark <r.clark at auckland.ac.nz> wrote:
>
>>> ... pathology, authority, anathema, mathematics, Gothic,
>>> Arthur, etc etc.

>> And I am astonished that anyone would present a list of
>> loanwords, however long, and claim that it has some bearing on
>> native English phonology.

> All of the listed words are old loans, and are fully Anglicized.
> They are no longer perceived as "foreign" by the majority of speakers.
> Thus they are indeed quite relevant to the *current* phonemic status
> of the sounds they include.

As I said, "foreign word" is a perceptual category and will vary from
speaker to speaker.  Indeed, most speakers won't even know what a
loan word is.  So we can write off the fact than intervocalic unvoiced
[th] only appears in foreign words as coincidence.  Can we then write
off the regularity of sound change as coincidence too?  I really do
think that when there is an overwhelmingly regular pattern then it has
to be acknowledged.

But I agree that synchronic grammar has to be independent of historical
developments.  Speakers of a language learn its grammar.  They very
seldom learn the history of its grammar.

>>   Loan words do not necessarily follow
>> the phonological rules of the borrowing language.

> Only before they are nativized.  Once nativized, they become relevant.

Sure, but again, this is a perceptual category.  But most speaker will
realize that native words do not have unvoiced intevocalic [th].

>>   In fact this
>> is usually one of the first clues that a word is a loan when it
>> doesn't obey the phonological rules.  This is how you can tell
>> that 'father' is a native (inherited) word and 'padre' is a loan.

> 'Padre' is still perceived as a Spanish word in English.  Few even know
> that 'authority' is NOT originally a native word.

Sure, by people who know what Spanish (although it might also be from
Italian or Portuguese) and know what a loan word is.  Most in the military
think that it is just what you call the chaplain.  So let's consider
'patriot' instead.  Few even know that 'patriot' is NOT originally a
native word.  It's been in the language longer than 'padre'.  Or let's
consider 'patron' or even 'pattern'.  Perceptions of these as foreign
or native words may vary considerably.  Few may even see any connection
between 'father' and 'padre', 'patriot', 'patron', and 'pattern'.  But
this doesn't change the fact that you can see which word is native and
which are loans.

>>> [dh] occurs in word-final position in breathe, bathe, writhe, etc

>> These are morphophonemic variants.  One method of forming verbs
>> from nouns in English is by voicing a final unvoiced spirant.

> And in all of your examples the two sounds are phonemically distinct.
> That is /s/-/z/, and /f/-/v/ are good phoneme pairs, so this is not an
> argument for denying /dh/ phonemic status.

No they are morphophonemic alternations which are predictable by rule.
If you have 'half' and 'halve' the rule tells you that 'half' is a
substantive and 'halve' is a verb derived from it.  If you have 'fife'
and 'five', this is a phonemic distinction because the alternation of
[f] and [v] tells you nothing about the meanings of the words except
that they are different.

> Also, I am not sure I would allow word derivation processes to use
> NON-phonemic changes.  The very fact that a sound difference can be
> used in word derivation is, to my mind, evidence that the difference
> is in fact phonemic.

Again, this is a difference between synchronic rules and diachronic
(historical) rules (changes).  The synchronic derivational rule is not
based on the historical change, but they both have the same result.
Historically, the spirant in the infinitive was voiced because it was in a
voiced environment when the change that voiced spirants in this
environment took place.  The derivational rule came about because the
spirant was not devoiced when the attrition of the infinitive ending
left it in word final position (in contrast to the final [th] of the
ordinal numbers when it became word final).

>> But before you get too deeply involved in trying to find
>> something, consider this also simple fact:  If it is not possible
>> for English speakers to determine the pronunciation of <th>
>> as [th] or [dh] entirely by rule, how is it possible for the
>> graphemic system to get by with only one grapheme for the two
>> sounds?

> In the same way that Hebrew can get by with a writing system that does
> not represent most vowels, and the same way Mycenian Greek could get
> by with a syllabic writing system that failed to represent the
> pronunciation of the language.

Hebrew gets by without writing most vowels because the vowels are
predictable from the environment.  There's that word again -- predictable.
Mycenaenean Greek got by because the writing system was used for extremely
limited purposes and in administrative texts and inventories they drew
little pictures to tell the reader what the word was so he didn't have
to rely on the lousy phonological fit between the writing system and
the language.

> Answer: a native speaker has the vocabulary *memorized*, so they
> *know* which words are pronounced which way, and read that *into* the
> written word.

No, the answer is that people learn to speak *before* they learn to
read and write (in fact, some never learn to read and write at all,
but this doesn't affect their ability as speakers).  If there is a
good fit between the writing system and the phonology of the langauge
then the reader doesn't have to know anything about the vocabulary.

When I was in Italy many years ago, I encountered someone who promised
me that he would teach me to read, write, and speak Italian in 6 hours
(1 hour a week for 6 weeks).  He was as good as his word.  At the end
of that time, if I saw something written in Italian I could speak it
flawlessly; if I heard something spoken I could write it easily.  The
only problem was that I couldn't understand a word of it.  He taught
me the relationship between the writing system and the sounds of the
language.  Someone (Italian) later handed me an invitation written in
Italian and I read it to them, and then said, "but what does it mean?"
The reply was, "but you read it perfectly, you can't possibly not know
what it means."

The same is not true of spoken and written English.  Finnish children
learn to write from dictation at the age of 7 or 8.  English speaking
children never do.  Spelling in English is more of an art than a
science.  That is why the spelling bee is such a popular entertainment
in English speaking countries.  In Finnish or Italian what would be
the point?  Once you hear a word spoken properly you know how to spell
it.  Once you see a word written you know how to pronounce it.

But the point is that speakers don't have the entire vocabulary
*memorized*.  They couldn't.  How many words are there in an English
dictionary?  When they hear a word that they don't recognize they
have to guess how to spell it.  When they see a written word they
don't recognize they have to guess how to pronounce it based on
certain rules.  If one knows a couple of rules one can always guess
how to pronounce <th> in English.  That's why the distribution of
sounds represented by <th> is predictable and not arbitrary distinctions.

Bob Whiting
whiting at cc.helsinki.fi