--- In qalam@yahoogroups.com, "Peter T. Daniels" <grammatim@...>
wrote:

> suzmccarth wrote:
> >
> > --- In qalam@yahoogroups.com, "Peter T. Daniels" <grammatim@...>
> > wrote:
> > >
> > > That depends, of course, on what the meaning of "diacritic" is.
> >
> > That is a very good question. Here is the Unicode definition. "A
> > mark applied or attached to a symbol to create a new symbol that
> > represents a modified or new value." I sometimes think that the
> > term diacritic is used interchangeably with 'combining

character'.

> > Both a pulli and the dependent vowels are called combining
> > characters, but are they both considered diacritics?
>
> What a surprise that they came up with something vague and useless.
>
> I'd say diacritics do NOT create a new symbol -- that's the point
> (Turkish dotted and dotless i don't involve a diacritic) -- but

modify a

> symbol to give it a modified reading.

Well if Turkish dotted and dotted i don't differ by a diacritic, nor
within Turkish do the pairs <o> and <ö> and <u> and <ü>.

In German, they are modifications created by a live grammatical
process. In French (and Greek) they are modifications like an
Arabic hamza. As such differences use identical and identified
('unified') notation within the varieties of the Latin alphabet, the
definiiton has to cover such vagueness.

Does it matter whether <G> is <C> with a diacritic? Devanagari has
a diacritic (by a graphical criterion), 'nukta', used to create new
consonants. It corresponds well with most of the new characters in
the long-established Tai alphabets (Thai, Lao, Lanna), where this
unacknowledged diacritic manifests itself in up to two ways:

1) Stretching the top right 'tail' of the consonant;
2) Denting an upper horizontal portion (Thai, Lao) or adding a large
hook to the top left of the consonant (Lanna).

(Viet Tai also seems to have a standard way of making a new
consonant.)

Does it matter whether this modification in the Thai alphabets is
seen as a form of nukta or not. Synchronically, I think not.

Where the use of the term diacritic does matter is, of course, in
the input method. What determines whether a diacritic is composed
or precomposed is the number of combinations that arise and how
stably it is connected with the morpheme. Again, it is a matter of
the effort involved. Replacing a precomposed e acute by an e grave
is not a problem. Replacing a composed e acute by a composed e
grave is irritating. In manuscript one would just change the accent.

As the number of keys on a keyboard is limited, there are two
possibilities for dependent vowels - independent characters or
diacritics. As there is a limit to how many vowels etc. a character
can correctly bear, a desire to preclude invalid combinations forces
them into a dependent status for editing.

Richard.