Re: A quick question

In a message dated 05/1/02 4:55:25 PM, faceloran@JUNO.COM writes:

<< Sound changes *only* affect the complete forms of the words. >>

    I think the confusion may be in whether you're talking about a diachronic
or synchronic sound change.  For example, final voiced stops in German and
Russian become devoiced.  This wasn't a diachronic sound change; it happens
every single time a German/Russian pronounces a word, and it changes
depending on what cases are applied.  Example:

Russian: /gorod/ realized as [gorot]
But in the prepositional: /gorod + e/ realized as [gorode].

    It all happens synchronically.  That's why I suggested level ordering and
cyclicity.  That way you would have an [x] in all forms.  Of course, you
could just do that by saying /k/>[x], period.  The implications of saying
that /k/>[x] morpheme finally, by the way, is that level-ordering is
affecting the phonology, which DOES happen, no magic wands, or nothing.  All
you're saying is that first the phonetic rules apply, then affixes are added,
and then, if applicable, the phonetic rules apply again, then more affixes,
etc., until you're done and you have your word.  Here's an example of how
morpheme boundaries, level ordering and cyclicity are important in a real
language:

Sundanese:

    Sundanese has this rule that says a nasal consonant spreads nasality to
all the vowels that follow until it comes to another non-nasal consonant.
So, you have words like:

[Na~u~r], [mi~a~sih], [ma~rios]

    This language also has an infix /-ar-/ added after the first consonant
(sorry I don't have glosses for these words; it's in my folder somewhere).
So, if you were to add to, say, the first word, you'd expect a form much like
this: [Na~laur], right?  (Oh, /r/ and /l/ vary, depending on the following
liquid.)  Indeed.  However, it's wrong.  The real form is [Na~la~u~r].  Why?
It seems to go against the rule I stated above.  Well, it's because of level
ordering:

Underlying: /Naur/ + /-ar-/
Phonology: /Na~u~r/ + /-ar-/
Morphology: /N-ar-a~u~r/
Phonology: /Na~la~u~r/
Surface: [Na~la~u~r]

    If you were to switch that order up, though...

Underlying: /Naur/ + /-ar-/
Morphology: /N-ar-aur/
Phonology: /Na~laur/
Surface: *Na~laur

    That's how supposedly "invisible" morpheme boundaries can affect
phonology.  And this all happens synchronically; it's *not* the product of a
historical sound change.
    But anyway, if you want to make it so that /k/ goes to [x] in all cases,
then none of this is necessary, and the morpheme is simply /x/.

-David

"fawiT, Gug&g, tSagZil-a-Gariz, waj min DidZejsat wazid..."
"Soft, driven, slow and mad, like some new language..."
                    -Jim Morrison