Richard Wordingham wrote:
> --- In, "Peter T. Daniels" <grammatim@...> wrote:
> > Richard Wordingham wrote:
> > >
> > > --- In, "Peter T. Daniels" <grammatim@...>
> > > wrote:
> > > > Richard Wordingham wrote:
> > > > > Perhaps the critical thing about the most abugidas is that the
> > > > > vowel marks' positions are usually scattered about the
> > > > > consonant. How does the set of positions affect the
> > > > > cognitive processes?
> > > > No, the thing about all abugidas is the inherent vowel.
> > > > The positioning of the marks is a historical accident.
> > > I think the educationalists will tell us that the placement is a
> > > very significant accident. Would Tamil be any less of a
> > > neosyllabary if pulli indicated /a/ and the base form the syllable
> > > final consonant?
> > "Neosyllabary" means nothing to me, and you won't provide a definition
> > for the way you're using it, so I cannot answer your question.
> I did, but you rejected it because I'm not persuaded that a real
> neosyllabary in that sense actually exists. To me a neosyllabary is a
> descendant of a phonemic writing system, but a descendant in which the
> originally phonemic nature is so obscured that the system is best
> worked with as a syllabary. As there are degrees of obscurity, it is
> legitimate to talk of a system being more or less of a neosyllabary,
> just as in chemistry the difference between covalent and ionic bonding
> is a continuum.

So you say that the thing does not exist, but you continue to use the
name as a label for things. How is that supposed to be interpreted?

> > No language has VC syllables to the exclusion of CV syllables, so your
> > hypothetical seems counterfactual.
> Which is an indication that you misunderstood it! I will therefore
> rephrase the question.
> Let us ignore marginal exceptions which have no bearing on the basic

Which exceptions are we ignoring? Hard cases make bad law, but marginal
data are often the most revealing data in linguistics (and probably
elsewhere, too).

> system. Native Tamil words have a (C)V(C) syllable structure. The
> final C is written as base consonant plus pulli; the CV part is
> written as base consonant alone, or base consonant plus vowel symbol.
> Some of these combinations have sufficiently obscure forms that Tamil
> is considered to be partly a neosyllabary as I understand the term.

That doesn't help ...

> Now suppose, as actually happened in one form of Old Tamil Brahmi,
> that CV was always written as base consonant plus vowel symbol, and
> the the final C were written as base consonant without additional
> mark. For the sake of definiteness, let us further assume that the
> vowel form for /a/ were identical to the actual pulli. (That is not
> so very far from an actual Thai/Khmer symbol for short /a/ in closed
> syllables, and does not change the percentage of vowels that are
> indicated after the consonant.)

Thai is not Khmer, and Thai/Khmer is not Tamil, but let that pass ...

> Should this hypothetical but plausible modification of Tamil have more
> or less or fewer obscure combinations than actual Tamil?

I have no idea what "obscure combinations" may be, but it sounds like
you just described an alphabet. Like Lao. (I wish Bright had made Diller
say that Lao is an alphabet -- I finally found the statement "Lao marks
all vowels overtly" (WWS 463).

> > > You can argue that Cree is an abugida because
> > > the West(?) Cree finals are reduced forms of the a-forms, and
> >
> > The way it notates final consonants has no bearing on its typology.
> On the contrary, it shows that the a-form is the default form. The
> other CV forms are derived by rotation or reflection, and the virama
> equivalent is shrinkage. West Cree thus has an inherent vowel - /a/ -
> and is thus an abugida!

No, it's a "sophisticated grammatogeny," so it has no place in the
Peter T. Daniels grammatim@...