Re: Optimum number of symbols

Thomas R. Wier <trwier@...> wrote:
>
>Quoting "Mike S." <mcslason@...>:
>
>> On Tue, 21 May 2002 21:08:46 -0500, Thomas R. Wier <trwier@...>
>> wrote:
>> >Quoting "Mike S." <mcslason@...>:
>> >
>> >> > [snip]
>> >> > >structures, I think the simplicity and efficiency of the phonemic
>> >> > >system easily trumps all contenders.
>> >> >
>> >> > English???
>> >>
>> >> I believe you are knocking down a strawman here.  No one would
>> >> suggest that English orthography is the ideal model of a phonemic
>> >> system.  It might be questioned whether it is a phonemic system
>> >> at all.  I would argue that it a combination of phonemic and
>> >> morphemic approaches.  [...]  Whatever the benefits or
>> >> shortcomings of this combinational approach, they have little
>> >> or no bearing as criticisms of a true phonemic system.
>
>[...]
>
>> Interestingly (assuming my German doesn't fail me the way it
>> failed me the other day), the words you picked are *almost* an
>> example of both cases at once:  <Rad> and <Rat> sound alike, but
>> are spelt differently; <das Rad> und <die Raeder> contain stems
>> that are spelt *almost* alike, but are pronounced differently.
>
>Actually, I was referring to different forms:  <Rad> and <Rat>
>are homophonous in the singular, and yet are phonologically
>distinct in the plural:  <Räder> /RE:d@/ and <Räter> /RE:t@/.
>This does bear on the criticism of a true phonemic system, since
>a true phonemic system will fail to capture phonological
>neutralizations like that in the German data I presented.
The basic feeling of a phonemic purist is that a true phonemic
system should not fail to make a distinction anywhere the spoken
language makes one, but that there's no valid reason to insist that
a writing system make distinctions that speech does not, and
that deliberately introducing irregularities into a writing
system typically does more harm than good.  From this perspective,
it would simply be argued this is not a valid criticism of the
phonemic system.

That being said, let me state that despite possible appearances
I am not a purist.


>That is, /Rat/ is really two distinct words: /Rat/-1 and /Rat/-2,
(Just out of curiosity, is this the conventional analysis of <Rad>?
Having studying a little German on my own, my way of memorizing
<Rad> is to think of the root morpheme as being /rad/.  Then
I remember as a general rule that any root-final voiced stop
becomes unvoiced.  Thus, /rad/ + (null inflection) -> /rat/.
I have to say that my method strikes me as a little easier than
trying to remember /rat/-1 (or was it 2?) + er -> /rE:d@/.
In other words, it seems that knowing the general unvoicing
rule beats having to know hundreds of irregular voicing rules.
Perhaps I am missing something though.)


>which German's *morphophonological* spelling system captures
>because it *ignores* the phonological homophony in order to
>satisfy another goal, semantic difference.
Yes, but my understanding of German has this rule as so regular
as to look like a phonemic rule in disguise.  My gut feeling, the
way I *want* to look at it, is that word-finally /t/ and /d/
somehow share an allophone, [t].  If this could be accepted,
then no violation of the phonemic system occurs in the first place.
However I don't accept this; the phonemic system is violated.

Now, if there were a couple hundred German words floating about
that ended in a <d> that was pronounced [d], causing confusion,
I might argue that the rule was not very beneficial, and I bet
a lot of Germans would agree with me.  But there aren't, and I don't.

So I agree with you.  It is far more beneficial here to go with
the sheer practicality of accommodating morphemics than to insist
on pure phonemics.


>In other words, purely phonemic scripts, which would insist on
>one written form for the two meanings of /Rat/, may in fact be
>less useful to the speakers involved.
Quite true.  Likely every script would be made more useful
with an occasional tweak here and there.  I would hope as regular
as the German unvoicing rule.


> And afterall, writing's
>_raison d'etre_ is to serve speakers interests, in whatever
>way seems to them most useful. You cannot tell objectively
>(and I think this was one of Nik's points) what the speakers
>will feel is the most salient interest.
I agree up to a point.


Regards