Re: Dublex/Katanda hybrid

From: "And Rosta" <a-rosta@...>

This post contains a few last remarks to And's critique as well
as a few phonological revisions I have made to the model language.

> Mike S:
> > [snip]
>
> The scheme is ingenious and simple, and I can see how it could
> be argued to be an improvement on Katanda.
>
> But as a practical scheme it seems both wasteful and unnecessarily
> verbose. Wasteful, because some meanings whose frequency and
> basicness warrant their being assigned to a primitive (e.g. say),
> do not require hundreds of different derived roots for different
> sorts of saying (unlike bird, tree, etc.). unnecessarily verbose,
> because other -- perhaps less elegant -- schemes could be more
> concise.
>
> --And.
It is puzzling that you cite the root "say" as an example of
a wasteful primitive.  Not only does Morneau use /da/="say, tell"
as the basis of deriving potentially dozens of speech act verbs,
e.g., "explain", "tell the truth about", "complain", "praise",
etc., but he also feels compelled to designate a separate morph
/ta/ to serve as the corresponding root-starter, e.g. /taji/=
"talk-bird (parrot)", many others.

Nevertheless, I am sure you are correct in pointing out that
there will be many primitives that will be somehow necessary
yet relatively unproductive in their capacity either as root-headers
or as semiroots.  I stated earlier, and I wish to emphasize now,
that for all its charm this system will likely be rendered
mediocre or disappointing without an extensive amount of study
and planning prior to primitive selection.

Nor am I convinced that the result of even a careful, planned
attempt would be more concise--or more anything else--than any
other scheme, and I certainly make no claim to that effect.
Such could only be determined after much further development.

Regards,

---   Mike

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Phonological Revisions.

The two affricative phonemes /c/ and /j/ are hereby scrapped.
The reason for this is that there are too many consonant
clusters that, unbuffered, would be confused for them, e.g.
t + c, t + h, c + h. (formerly h=[S]).  As a result, there
has been reshuffling; henceforth: c = [S], j = /Z/, h=/G/.

The two graphemes freed in the reshuffling, y and w, are
hereby reassigned as front round vowels like French "u"
and "eu" respectively.  If these were the only changes,
our sets of primitives and of semiroots would be:

Permutations of (Semiroots ::= C V) = 18 x 7 = 126.
Permutations of (Primitives ::= C V X) = 18 x 7 x 26 = 3276.

Increasing the number of available semiroots from 100
to 126 will definitely pay dividends over the difficulty
that the front round vowels will present some people.
Selecting semiroots properly is the central design concern
of this system; 26 extra semiroots will make that task
a little easier.

There are two last revisions.  I am going to hazard that
we will not require as many as 3276 primitives.  In order
to ease pronunciation, let's eliminate all morph-final
voiced stops and fricatives.  Unvoiced stops and fricatives
shall be fortis initially and lenis finally;  in front
of voiced stops and fricatives they shall also be voiced.
(Alternately, you could regard final stops and fricatives
as voiced with unvoiced allophones.  It makes no difference;
I prefer to spell them with the unvoiced graphemes.)

Lastly, gemination shall be allowed.  The result of all
these changes is that the "native" accent will now never
require a buffer schwa in any location.  The buffer schwa
does however remain an option for those who opt to use it.

The final tally for primitives:

Permutations of (Primitives ::= C V X) = 18 x 7 x 19 = 2394.

I will try to put together a revised description incorporating
all changes tomorrow.

---   Mike