On 10/23/2001 08:57:53 PM Kenneth Whistler wrote:
>> >If you examine the UCAS, you'll see that the orientation of the
>> >base consonant symbols (rotationally), and the placement of the
>> >dots (or dashes), correlate with the vowellings of the syllables.
>>
>> Which it would seem means that this has properties of an abugida, I
would
>> have thought.
>
>An abugida is basically a *consonant* writing system.
>
>UCAS has a set of forms for open, vocalic syllables, like
>the Hiragana a i u e o, as well as a complete set of CV
>syllables (or actually several sets, since the entire system
>has mutated as it has spread from language to language).
The presence of forms for open vocalic syllables is such a big difference
from Ethiopics glottal fidel series?
Hiragana is quite difference since there's no regular difference between
the different vowel forms for a given consonant.
>Conceptually the UCAS system is an enumeration of syllables.
No more or less than Ethiopic and Hiragana.
>What makes it featural is that you can find systematic correlations
>in how the vowel ranks are indicated.
Which is also basically true of Ethiopic, and which is the basis given for
the definition of abugida by Daniels and for considering Ethiopic to be a
prototypical abugida.
>Also, I don't know of any true consonant writing system that
>indicates vowels by rotations and inversions of the consonants.
I don't think I know of *any* other writing system that does exactly this.
The closest thing I can think of is Pollard, which (if I recall) uses
relative position of the vowel part to indicate tone.
>> and for this reason I had heard Ethiopic described as an abugida. So
what,
>> then, is the difference between a featural syllabary and an abugida?
>
>See above.
>
>If the basic enumeration is a list of consonants, you are dealing
>with an abugida.
>
>If the basic enumeration is a list of syllables, you are dealing
>with a syllabary
I can go along with the principle. I just can't see a good reason to
suggest that UCAS basically enumerates syllables while Ethiopic basically
enumerates consonants.
> -- and you then determine whether it is featural
>or not by examining the graphemes for phonological correlations.
But abugidas also have phonological correlations for the regular
variations -- and regular variations are part of the definition given by
Daniels.
>> >and the placement of the
>> >top "butterflies" corresponds with manner or other distinctions among
>> >the consonant ranks.
>>
>> Writing systems for Mayan languages often add an apostrophe after a
>> consonant to indicate glottalised forms; writing systems of many
languages
>> use a tilde over vowels to indicate nasalistion; many W. African
languages
>> use grave and acute to indicate tone levels; etc. Does that make all of
>> these writing systems featural?
>
>The alphabets themselves are generally not, since they are just
>extensions of Latin, where the letter shapes are arbitrarily
>related to sounds.
>
>However, if you do a graphological analysis of the orthographies
>for these kinds of systems, and treat the various digraphs as
>units of the overall system, then I'd say you can make
>featural attributions for the grapheme formation. Thus, if I
>have a system like: p t k ph th kh t' t' k' and so on, the
>digraphs or use of modifier letters are clearly featural.
Which leads exactly to the point I was trying to make: this is stretching
the definition of "featural" to a point at which it applies to *a lot* of
writing systems that would certainly normally have been considered to be
of a different type. Latin writing systems are alphabetic, but you've
found a way to call them featural; Hiragana is a prototypical syllabary,
but the existence of the voicing mark makes it featural; Arabic is usually
considered an abjad, but shaddah makes it featural; Ethiopic is usually
considered the prototypical abugida, but you've shown it to be featural;
some would consider Thai an alphasyllabary, but tone marks makes it
featural. (If you view Indic scripts in the same way that Daniels does,
you consider them to be abugidas, and that means you view them in terms of
a regular graphical variation that corresponds to a vocalic variation, and
so you'd consider any of them to be featural.)
If "featural" is taken to the extreme of meaning that there is a regular
graphic variation hat corresponds to a regular phonological variation,
then any writing system could potentially be argued to be featural. Is
that what "featural" should really be taken to mean?
If so, then "featural" is not a type within a classification that
partitions scripts between abjads, alphabets, etc. That doesn't
necessarily bother me since in any discussion of that partition I've ever
seen the "featural" class consisted of exactly one script, Hangul, and a
class of one in this case has always suggested to me a breakdown in the
system of classification rather than a well-motivated basis for
classification. But the definition you've given here for "featural" is
also somewhat different from that which Daniels used in application to
Hangul: if I recall, it was the metaphoric similarity between the graphic
shape and the shape of articulators / point of articulation that led him
to call Hangul a featural system. That is a rather narrower definition
that what you are using.
- Peter
---------------------------------------------------------------------------
Peter Constable
Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485
E-mail: <peter_constable@sil.org>
This archive was generated by hypermail 2.1.2 : Wed Oct 24 2001 - 05:42:39 EDT