[NGL] -ig suffix question
Jack Durst
Date: Mon, 11 Oct 1999 01:59:16 -0230
From: Stephen DeGrace <c72...@morgan.ucs.mun.ca>
Subject: Re: [NGL] -ig suffix question
From: Stephen DeGrace <c72...@morgan.ucs.mun.ca>
On Thu, 7 Oct 1999 14:49:55 -0700 (PDT), Gerald Koenig
<j...@netcom.com> wrote:
[...]
>Any marriage:
>
>nex v marry
>nex-xi n-->
>nexi n matrimony
>nexi-i adj-->
>nexisi matrimonial
>nexisi-ig adv -->
>nexisisig adv matrimonially
>CVCVCVCVC
>
>
>I'll propose this if it's wanted.
>
>Jerry
Hmm. The benefit to this idea is that it would allow unambiguous
parsing of constructions which are now ambiguous (in writing and in
speech of any dialect which doesn't make a point of pronouncing
unwritten long vowels) by the reduction rules. The disadvantage is
that it's going to make some words significantly longer in our
language which already suffers from a tendency towards longer words
simply by its phonology. It could be that context can distinguish most
ambiguities currently present and that this additional "precision",
for want of a better word, is not really required. Maybe we could just
try it for a while and see? How does that come across?
Stephen
--------------------------- ONElist Sponsor ----------------------------
Get EXPERT CONTENT at ONElist!
Join PROS&PUNDITS. For details go to:
<a href=" http://clickme.onelist.com/ad/prospun1 ">Click Here</a>
------------------------------------------------------------------------
Jack Durst
make parsing easier, but at a significant cost in terms of word length.
Going from 3 letters and 1 sylable for -cig to 7 letters and 3
sylables for -cisisig (or similar) would cause a serious increase in
adverage word length with *no* increase in parsability (given that it's
unambiguously an abstract adverb anyway) in many cases and only a slight
increase in the already ambiguous cases.
I'd like to counter-propose a different possibility based on something
that's already well accepted in the language: Stephen's contraction
marker. That would require no increase in word length and would still
disambiguate where nessicary. -cig becomes -cīg, no increase in letters
or sylables in the 8-bit version, only one letter increased in the 7-bit
version, and it still disambiguates in almost all cases.
In the abstract, this could even be applied to a consonant reduction; just
place the ^ on the vouel preceeding the consonant, or place it directly
before the consonant itself as a seperate character.
I'd like to make it optional, as, in a great many cases, disambiguation is
unnessicary, but I'm flexible on that count.
On another note, It is specifically *not* a stated goal of NGL to be
machine parseable, the language clearly envisions itself being used by
humans, and so a level of ambiguity that the adverage native speaker could
easily resolve from context is perfectly permissable. On that count, it
is, strictly speaking, unnessicary to disambiguate in most cases.
However, I have no opposition to doing it as it could well prove better in
terms of reading spead and ease of learning the language.
The old timers will recall that I was opposed to the double letter rule
when originally proposed for that very reason.
Sincerely,
Jack Durst
Sp...@sierra.net
[this posting written in Net English]