Hi all,

On Mon, 10 Oct 2005, Jim Henry wrote:
> On 10/10/05, Yahya Abdal-Aziz <yahya@...> wrote:
> > > 3) So let's use the final vowel as a lexical class marker.
> > Here I have a problem. Does this perhaps imply that the
> > number of lexical classes must equal the number of vowels
> > in the phoneme inventory? If so, are we sure that gives us
> > enough lexical classes to construct almost any interesting
> > utterance?
>
> If you look back in the group archives, you'll see
> we talked about a possible scheme with four semantic
> categories (substance, action/process, relationship,
> and quality) which correspond to four "parts of
> speech", noun, verb, preposition, and adjective/adverb.
> But I think Larry finally went with a simpler scheme
> that folds relationships into the other categories.
>
> Jim Henry

Thanks for this, Jim.

On joining the list, I scanned the messages page and
discovered that the list has been running since March.
Rather than read the entire archive, I chose to start at
the beginning of September, hoping this way to avoid
discussing any dead issues. Of course, in doing so I
accepted the risk that I would reopen old issues in
ignorance of the previous discussion and conclusions.

From reading the (latest) Konya description, I expected
it to have exactly four semantic categories with
corresponding lexical classes. Jim's message suggests
that Konya now has only three. Could it be that we will
discover a need to make the number five, say, or seven?
Unless we can definitively fix the number of categories
AND match it to the number of available vowels, it
seems likely to me that we may either have a category
we can't express, or a potential waste of a significant
fraction of the possible shorter morphemes.

For your consideration: Perhaps this one-to-one
mapping of vowels and semantic classes is not quite
flexible enough?

Regards,
Yahya

--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.11.13/126 - Release Date: 9/10/05