I had intended to send this to the entire AUXLANG list.  When
it didn't show up, I just realized that it had gone to the original
poster.  I am so accustomed to assuming that a reply-to will go to
the entire list that I didn't even notice the address.

---------- Forwarded message ----------
From: Paul O. BARTLETT <[log in to unmask]>
To: Simple Language <[log in to unmask]>
Date: Thu, 21 Oct 2004 15:51:11 -0400 (EDT)
Subject: Re: A piori vs. a posteriori auxlangs

On Thu, 21 Oct 2004, Simple Language wrote (excerpted):

> Large number of root words makes it easier to define compound
> words well. On the other hand, the root words should
> not be longer than two letters to keep the compound words
> short.

    At the expense of potential confusion.

>        Ygyde solves this problem by giving special meaning
> to pairs of root words. [..]
>
> cake = owafyby = "noun happy foam food"

    Why not some kinds of whipped pudding?  My point is that many of
the compounds seem arbitrary and therefore not self-explaining.  And
if they are arbitrary and not self-explaining, why go to the trouble
of pretending that the language is so clear as to its vocabulary?

> Three letter long compound words are not good because they
> produce consonant clusters or vowel clusters.

    Recently Koboseya, the "Kobold" language was announced at
http://koboxora.free.fr/index.htm (all the explanations are in French
at the moment).  A few words have an initial vowel, but otherwise all
words are CV syllables of varying numbers.  The vocabulary is based on
871 (I think it is) roots, with a number of lexical affixes.  I haven't
looked at it in depth, but it seems to me to be more attractive than
Ygyde and less confusing.  I am not necessarily advocating Koboseya,
only that an easily pronounceable language can apparently be built up
of a manageable number of roots and affixes that is not so restricted
as Ygyde's.

> It seems that you dislike a priori languages. So far you have not said
> anything positive about them.

    As applied to constructed languages, "a priori" actually has two
slightly different uses.  The first use is a vocabulary is built on a
classification scheme of some kind.  Notable examples are Real
Character, Ro, and Oz.  I myself honestly do not think that they are
workable, for several reasons, and so do not give a whole lot of
attention to them except to the inflated claims of their creators.  The
second type of a priori language is one merely in which the vocabulary
items are created without reference to exiting natural languages, but
not according to a classificatory scheme.  Suma comes to mind.

>                               I dislike natural languages and euroclones,
> but I do not worship Ygyde.

    That is your choice, of course.

>                             I like to talk about compound languages because
> their potential has not been fully explored yet.

    I think that they have been tested and found wanting.

> > The *existing* words are too close
> >together in sound to enable reliable distinction in a noisy environment.
> >Adding more phonemes might actually make the discrimination better.
>
> I dislike the letter 'r' and the consonant clusters. Chinese speakers claim
> that tones (rising, falling, modulated) are easy to distinguish. Perhaps
> tones or accent can make Ygyde's words more distinct...

    If you intend your language as an international auxiliary language
 de novo usable by adult learners without an a posteriori vocabulary,
then you may have a case for not having *both* of /r/ and /l/, as the
distinction between them is difficult for some people.  Similarly with
tones and many consonant clusters.  Certain phonemes (such as /T/ and
/D/) are actually quite rare and may be avoided.  However, that does
not mean that I think that the overall design of Ygyde approaches
optimality (whatever that might mean).

>                                                         Ygyde is not
> protected by copyright, so anyone can make his own Ygyde-like language.

    Does that mean that I can design my own language and call it Ygyde?

--
Paul Bartlett
PGP key info in message headers