Re: About "features"

At 13:29 -0600 2001-11-13, Peter_Constable@... wrote:

>On 11/13/2001 10:00:31 AM Michael Everson wrote:
>
>>>Why? What is it that makes us willing to use "featural" to describe
>>>a script? Obviously it has some limits; what are they?
>>
>>Shapes. A featural syllabary uses its glyphs to help you remember and
>>recognize the sounds. "All these ones that look alike have a similar
>>sound". A non-featural syllabary is a collection of shapes that have
>>no systematic relation to the sounds they represent.
>
>It sounds, then, like every abugida would by definition (assuming your
>definition) be featural.

Certainly not. Why would you think that? It's clear, Peter. Lots of
scripts have glyphs, and they point to phonemes and such. When there
is a graphic system to the glyphs where glyphs have relationships
between themselves which are analogous to the physical properties of
related phonemes, then the system is featural.

Abugidas do not share this property a priori. Abugidas have inherent
vowels, that's all.

> >Look at the script. See if it has any systematic relation between
>>shapes and sounds, or if they're just conventional, unrelated to one
>>another.
>
>That's a little more explicit. I think it should be fleshed out more,
>though.

But it is simple.

> >>The shape P can also easily be morphed
> >>into the shape D, and there is some commonality in terms of articulation
> >>(they're both stops) in the sounds.
> >
> >Stopness seems pretty weak.
>
>That's exactly one of the points of the example: the definition you gave
>didn't give any idea of what the limits are. The second point was a
>corrolary: you could take it and identify a lot of more or less ad hoc
>pairs and claim that they fit the description you gave.

The limits are whatever the inventor chose to consider significant
units of speech and whatever graphs he chose to represent those in a
patterned and isomorphic way.

>Virtually every occurrence of "o" in English relates to a certain limited
>ranges of sounds. Similarly every occurrence of "i" or "w" or "t" relates
>to certain limited ranges of sounds. There are, therefore, systematic
>relationships between glyph shapes and sounds, and thus English
>orthography fits your definition. If the conclusion isn't what we want,
>then I'd suggest that the definition isn't adequate.

No, Peter! Use your imagination! Visualize, or go look in a book of
scripts. Think of situations where glyphs with similar graphic
features to other glyphs are used to represent sounds which also have
similar phonetic relationships to other sounds.

Your comment above about Latin does not in any respect go any where
near this very simple idea.

If you don't understand this please go and read Appendix E of the
Lord of the Rings. Tengwar is a perfect example of a featural script.
--
Michael Everson *** Everson Typography *** http://www.evertype.com
15 Port Chaeimhghein Íochtarach; Baile Átha Cliath 2; Éire/Ireland
Telephone +353 86 807 9169 *** Fax +353 1 478 2597 (by arrangement)