Lars Marius Garshol wrote:

>
> * i18n
> |
> | I wonder if there is any sort of correlation among the historical
> | use of certain categories of writing systems with certain categories
> | of languages.
>
> * Peter T. Daniels
> |
> | Yes. There is no such correlation.
>
> It seems to me that in most cases when a script begins to be used to
> write some language the users of that language either adopt or adapt
> an existing script and that they usually do not look around for a
> suitable candidate, but pick the one most obvious to them. It also
> seems that very rarely when adapting an existing script do they change
> it to such an extent that the new script belongs to a different
> category.
>
> So I guess that the reason for this lack of correlation might be that
> very rarely is a script extensively tailored for the language it is
> used to write, so that the influence of the language on its script is
> (generally) quite weak.
>
> In fact, in my scripts "database"[1] I can find 12 cases of a script
> changing type when being adapted to a new language, out 114 total
> cases, as follows:
>
> PREDECESSOR SUCCESSOR PRED. TYPE SUCC. TYPE
> ========================================================================
> Classical syriac Modern syriac Abjad Alphabet

Not really; the pointings merely changed from optional to obligatory.

> Linear A Cypriote syllabary Logosyllabary Syllabary

Do you mean Linear B? or what? Why do you say Cypriote "succeeded" LinB?

> Arabic script Thaana Abjad Alphabet

Hardly an adaptation; a reuse of symbols, not unlike (if it's what
happened) the reuse of some hieroglyphs for consonant signs on the road
to the Semitic abjad.

> Mongolian clear script Buryat Abjad Alphabet

To what extent is Mongolian not already alphabetic?

> Pahlavi Avestan Abjad Alphabet

No -- Avestan didn't develop out of Pahlavi; some Av. letters come from
Phl., some come from other sources. Skjaervo sent me his suggestions
after I discovered Hofmann's(?) (in EncIran s.v. Avestan) suggestions;
they should have been in the WWS article.

> Aramaic script Kharoshthi Abjad Abugida
> Sabean/Minean script Amharic script Abjad Abugida

These are the parade examples -- but why don't you include Brahmi?

> Orkhon Hungarian runes Abjad Alphabet

??

> Chinese script Man'yoogana Logosyllabary Syllabary

Why not the kana generally?

> Lanna script Tai Lue script Abugida Alphabet
> Phoenician script Greek Abjad Alphabet
> Proto-Elamite Old Elamite Logosyllabary Syllabary

OEl is in cuneiform syllabary; since we can't read "PEl" we have no idea
what language it represents, and there's no reason to suppose that
Sumerian cuneiform, adapted for Elamite, developed from PEl!

(I think you've slipped from examples of script adaptation to examples
of [supposed] earlier and later stages of a language that are written
with different scripts.)

> The clearest example of a language -> script influence I see in this
> list is a tendency for non-semitic languages to abandon the abjad in
> favour of script types that denote vowels. (Though this could hardly
> be called a scientific conclusion.)

Except, of course, that Iranian didn't.

> [1] It's actually a topic map.
>

You'll be wanting to look at my article in *Studies in the Linguistic
Sciences* 30/1 (2000), the King Sejong anniverary symposium at the
University of Illinois at Urbana-Champaign; it's also summed up in the
Blackwell *Handbook of Linguistics* chapter.
--
Peter T. Daniels grammatim@...