Re: 5 phonemes, was: Another phonological extreme

On Thu, 20 Jul 2000 15:18:57 -0700, Danny Wier <dawier@...> wrote:

>--- Vasiliy Chernov <bc_@...> wrote:
>
>> Yes, I know I’m abusing a certain understanding of phonology...
>
>I don't think you're abusing phonology!  In fact, you're explaining
>quite well the relationship of phonemes to phones.  And the effect of
>history on the move from phonemics to phonetics.
Personally, when dealing with natlangs, I prefer to think of phonology
as of something much closer to the surface level. I think the underlying
level in its relation to the surface is better termed 'morphonology'.

But, OTOH, I remember that when I first heard _try_ pronounced with
the American accent, I thought it was /tSwaj/ and wondered what it
could mean... :) close enough to what I'm doing with Ija and Miituu...

>The "deep" phonemes should be written, I think, with a phonetic script
>other than Latin or Arabic.
Not available in ASCII, I'm afraid :(

>Here a syllabry would be more practical,
>since it has a very small phonology.  Also, syllables are going to
>become somewhat alphabetic because of how your "deep phonemes" relate
>to "surface phones".
>
>Plus with a syllabry you only have to learn six characters; in an
>alphabetic script only one less!
You forget the V, CVC and VC types, also possible in Miituu on the
'deep' level. The total number of 'deep' syllables will be 36, I think.
(Ija had 15). Still practical for creating syllable-letters, though.

>Are your "true geminates" grammatical in origin (as in certain Arabic
>verb classes)?
Partly. I still don't know much about the grammar. Only that Miituu
(and Ija) will have a lot of infixes, in order that their weird
morphonology could show all of its beauties :)

>That would double your consonant phonemes, since
>gemination is not simply a VCCV group with identical C's.
Yes, I'm afraid this interpretation is possible, too :(

What is worse, I didn't really need all those closed syllables. Just
wanted to get some diphthongs, etc. I think I'll design a syster
dialect without such complications. Already know its name: Mwakupu.


>With /r=/ being a vowel, you have a vowel to consonant shift, kinda
>like Proto-Indo-European.
Or Spanish, or Portuguese, or Arabic (I mean the w/u, j/i thing).

>Is there a high-falootin' technical name for
>such?  (I can only think of "reduction", PIE having /er/ ~ /or/ ~
>/r=/.)
I don't know...

<snip details about Tech>

>30 consonants and 3 vowels?!  And Tech is said to have hundreds of
>consonants and dozens of vowels?  Where are the retroflexes?  The
>voiceless pharyngeal/epiglottal ejective affricate?  The velar nasal?
>The voiced implosive stops?  The voiceless and implosive nasals?  The
>fricatives?  All the sibilants?  The long, nasal and pharyngealized
>vowels?
>
>A Techian would just say that those extra phonemes are "contextual
>variants".
Yes, I like this type of things!

Did you check the distribution rules for the 'contextual variants'?

>
>But maybe they're right.  Most likely, Proto-Tech (remember, it's still
>an isolate unless you belive in Nostratic) probably had only the
>aforementioned phonemes, so these would be what Vasily called "deep".
>I probably left out a few I should've listed.  The legion "surface"
>phonemes came later as the language evolved (and revolved)...
>
>As for the "surface" phonemes, I won't get into that.  I posted the
>full phonology, and boy was it a scary sight...
>
>DaW.
>
Basilius