Re: Optimum number of symbols

At 3:55 PM -0400 5/23/02, John Cowan wrote:
>David G. Durand scripsit:
>
>>  In Semitic languages (like Egyptian, Arabic, Hebrew, etc.) the vowel
>>  inventory is not especially small, most have 5-8 vowel systems if I
>>  recall. However the grammatical function of vowels is relatively
>>  light, and they are generally predictable with little to no ambiguity
>>  in context.
>
>However, the Arabic script is used for a great many non-Semitic languages,
>often with great pain.  In Mongolian (which is also an abjad, and is
>derived from Arabic or something like it by several steps), the same
>glyph shapes can represent final a, e, or n.  There is furthermore
>no orthographic difference in any position between o and u, o" and u",
>t and d, k and g, or y and j.  Thus urtu 'long' and ordu 'palace'
>are written identically.
Of course! The application of Arabic script conventions to many
languages mets cultural or agendas at the cost of representing those
languages accurately. I take this as more grist for Christophe's
mill, which says that extreme caution should be exercised in
evaluating writing systems cross-linguistically, and
cross-culturally. I think that such evaluations are possible in
theory, but only under agreed criteria (which is unlikely) and
experimental evidence (which is sparse, and some of which will never
be available).

Mike's alphabet fandom reflects certain realities, but that depends
on a lot of unanalyzed factors.

>  > Another factoid for the mix: subtitles for movies in Chinese and
>>  Japanese are much more complete than those in alphabetic writing
>>  systems, because reading speed for logographic scripts is higher in
>>  terms of words-per-time-unit.
>
>And since reading dominates either learning or writing, in terms of the
>amount it is done, this is arguably the correct engineering tradeoff.
Right. If you're optimizing only for time spent reading/writing. If
the ability to make usable mechanical typewriters or typesetting
machines is important, then the tradeoff is different. If you have
computers with keyboards, it's different again. If you have graphics
pads and intelligent parsers, the balance might shift again. And
that's only one set of technology issues.

>  > The inarguable difficulty of learning
>>  such a system does have some payoff. Further, even more than in the
>>  case of English, the Chinese writing system unifies a set of _very_
>>  divergent dialects, that would are mutually unintelligible at the
>>  phonemic level.
>
>Well, not really.  If you mean dialects strictly, viz. dialects of
>Mandarin, then they are no more divergent than those of French or German.
>If you mean other Sinitic languages (Cantonese, Shanghainese, Hakka,
>Min, etc.)
>then speakers of those languages *can* "read out" hanzi texts in their own
>languages, but the result is frequently ungrammatical: to learn to
>write Chinese,
>they must learn the written conventions of Mandarin, not merely the
>symbol-sound
>correspondences.  The same is equally true of Chinese people who speak Tai
>languages.
I was referring to the larger group of languages. The claim is
frequently made that the script enables communication across these
groups. I can't comment as to how divergent the grammars really are,
and how hard that learning process is. For that matter, people tend
trained in a written or academic tradition tend to overlook the work
that was actually involved in that transition, or not to understand
how much harder that can be for people coming from different
circumstances.

   -- David

Reply