--- John Vertical <[log in to unmask]> wrote:

> >
> >Any other criteria are, to me, of negliable significance and can be 
> >ignored.
> Do you mean to exclude ease of learning explicitely or implicitely? 

Ease of learning was left out intentionally. The kazoo is much easier to learn
to master than the violin, but that doesn't imply that the kazoo is a "better"
musical instrument than the violin.

> Because 
> you could take this approach to its extreme and have different symbols for 
> the 5 million most common sentences + individual word diacritics to deal 
> with the rest... Then the writing system will be virtually impossible to 
> learn, but it *would* be ridiculously efficient for a hypothetical fully 
> taught reader. It's of no use if no fully taught readers exist, however.
> The fully taught writer would also probably not be all that efficient;

I doubt that the human mind would be that quick at picking out one glyph from
among 5 million. That's why I put reading efficiency above compactness. The
optimum balance would have to be determined experimentally, but I think it
would be mostly one symbol per word with some symbols for common short phrases
like "FYI", "IMHO", "FWIW", "PDQ", etc.

> John Vertical

--- [log in to unmask] wrote:

> li [Gary Shannon] mi tulis la
> > What is the "best" system of writing? 
<snip>> > 
> > 1. Of first importance (to me) is reading efficiency. A thing 
> > written once can
> > be read again and again, so the ease and quickness with which 
> > it can be read
> > outweighs the ease of writing.
> I would tend to say they should be balanced.  If writing isn't easy,
> it's possible that the writer may rush and perform a sloppy job which
> would render the end product difficult anyway.  

Very true. On the other hand, if the writer expects to understood it is his
responsibility to write legibly. On the other-other hand, maybe everyone who
uses this languages types their notes into their pocket computer and the
printer produces perfectly formed printed copy.
> > 2. It should be relatively compact, without sacrificing 
> > readability. 
> I found this could be done by making the symbols as simple as possible.
> Then they can be make small and be written quickly.

I agree. That's one thing I like about Shavian. What I don't like is that when
I read it I'm always hearing the writer's regional accent, which is annoying.

> > it doesn't make a bit of difference whether those 
> > shapes are constructed
> > from systematic phonetic elements or made up of arbitrary squiggles.
> Yes, but it's much easier to learn a phonetic system.  Learning to read
> and write Hanzi/Kanji takes many years.  Learning to write phonemically
> is something that could be picked up very quickly.

That's true, but perhaps the phonetic system could be the stepping stone to the
process of learning. Something like learning full phonetic spelling first, and
then after that is mastered at age 10, learning the "real" system of writing
which is based on full spelling, but more compact and efficient.

> > We do pretty much the same thing with "kite", "light" and 
> > "height" spelling the
> > long-I sound "i-e", "igh" and "eigh", which does a good job 
> > of creating less
> > ambiguous word shapes.
> The differences in spelling are not there because to distinguish the
> words but are a legacy of a time when these words were pronounced
> differently.  

Yes, of course. However, it is this chance "mutation" which is favored by
natural selection not for the reason it initially appeared, but for the "side
effect" it has on making different word shapes more distinct visually. If we
reformed spelling to make it more systematic and consistent it would have a
negative impact on maximum reading speed by making many words appear more
visually similar, and thus less readily distinguished at a glance.