On Wednesday, November 14, 2001, at 10:15 AM, william bright wrote:

> i'm surprised how often the present correspondence refers to tolkien
> scripts and other scripts invented by philologers (which tolkien was) or
> linguists or hobbyists, for fun or for experimental purposes rather than
> for practical use. surely there is very little limit to the typological
> characteristics of scripts invented by imaginative people. but what is
> interesting *to me* at least is: what characteristics of scripts WORK for
> communication in human societies?
>

I've spent the last couple of years actually using both Deseret and
Shavian and it's been an interesting experience. Both are seriously
flawed IMHO as reading scripts, but for different reasons.

In the case of Deseret, there are no descenders or ascenders. Except when
a word is capitalized, each word is rectangular, which makes it hard to
distinguish words without parsing them. Moreover, there are *faux amis*
among the letter shapes, such as a "C"-like letter which is used for the M
sound.

Strangely, Shavian was worse. It has ascenders and descenders galore,
true, but on the whole I tended to find the letter shapes, particularly
for the vowels, so similar that it was hard to learn them. In fact, I
tended to read it almost as if it were an abjad, just getting the
consonants and then adding the vowels that best made sense in the context.

Of course, in both cases I'm a Latin reader coming at it as an adult. The
kilometerage of a child learning the scripts from scratch could well vary.

==========
John H. Jenkins
jenkins@...
jenkins@...
http://homepage.mac.com/jenkins/