Re: Optimum number of symbols

From: "Nik Taylor" <fortytwo@...>
> "Mike S." wrote:
> > You fail to mention that syllabic script characters will need
> > be more complex as well.  On average, you are probably making
> > close to the same number of strokes per syllable.
>
> In my case, there are fewer strokes on average, as many characters are
> single-stroke, especially in certain variants.  Out of 81 basic
> characters, every single one may be written with a single stroke, tho
> some of them have 2 stroke variants that are more common, 9 to be exact,
> thus 72 out of 81 characters are *always* written with a single stroke.
> All of the diacritics are one-stroke, too, so, for example, klaaf, which
> must be written with 8 strokes in romanization, can be written with just
> 3 strokes in the native script.
Sounds efficient.

> > Phonemic scripts are also known for the feature of being written
cursively.
>
> A few alphabetic scripts are written cursively, but many are not.  And
> syllabic scripts can be, too.  For what it's worth, even in the
> alphabets I've made, I've never used cursives.
>
> > If syllables are more intuitive to learn and use, then there should
> > be no temptation to encode any phoneme-level information into
> > a syllabic script.  Explicitly encoding phoneme-level information
> > into a syllabic script is analogous to encoding phonetic data such
> > as +/- voice, +/- velar into a phonemic symbols.
>
> Your point?  Such things DO exist in alphabet scripts, such as long
> vowel markers and diacritics, not to mention digraph conventions, like
> the use of -h to make certain sounds in many roman-using language, like
> English _ch_ or _th_.  I see my coda diacritics as being equivalent to
> things like the tilde in Portuguese or Spanish, or the acutes, graves,
> and circumflexes of Latin, or the raised dot of traditional Irish
> spelling/ -h digraphs of modern Irish spellings.
You seem to have a good point here; some syllabic systems require
phonemic patches just as some alphabets have phonetic patches.
However, looking at the western languages, I have to say that
patches in the form of diagraphs or diacritics are more a testament
to the astonishingly conservative nature of the Roman alphabet
than to the any need of the alphabetic system to rely on phonetics.

Actually, most of the time, these patches are either unphonetic
or unproductive as phonetic markers.  I'm not sure what to call
the "h" in English diagraphs, but a phonetic marker, it is
certainly not.  Likewise, the Spanish tilde only marks one grapheme
--arguably phonetic but hardly productive.  The German umlaut
*seems* like a phonetic marker--until we learn that a"u
is pronounced [OI].  What's that all about?

In short, *very* few modern alphabetic have implemented any
phonetic markers productively, and *no* alphabet relies heavily
on phonetic markings in a systematic way.  Nor could one even
if it were desired.  Phonetic systems are not suitable for written
mass communications; phonetics are always far too complex.
On the other hand, it's trivially easy to design a script that
systematically encodes the phonemic contents of a syllable
and call it a syllabic system; my point is that to the extent
it exploits phonemic markers to its advantage, it's really
(by definition even!) a phonemic system in disguise, and serves
as evidence that a *phonemic* system simply makes more sense.

> Besides, there are too many syllables in Uatakassi for a pure syllabry
> to work.
Nonsense.  Would you need more characters than Chinese?
You can create a symbol for every syllable in your language
if you wanted to.

> > If understanding phonemic distinctions are needed, or
> > at least helpful, in learning or using a syllabic script, then I
> > can't see any reason not to use a phonemic script in the first place
>
> First off, you can start off by learning only the basic characters.  To
> teach a child to read, you must first get the child to be able to
> understand how to connect abstractions like /k/ and /a/ to make a
> pronounceable syllable /ka/.  You do not need to do this with a
> syllabry, you already have /ka/ right there.  After they get the basic
> concept of reading, then you can go into the abstract diacritics.  For
> that matter, I'm not sure if you'd even need to teach the diacritics as
> characters.  Does anyone know if Japanese children learn the double-dots
> as an abstract diacritic, or do they simply learn _ka_ and _ga_ as if
> they were separate characters, the way English-speaking children aren't
> actually taught the relationship between s-sh and t-th?
You can learn the sounds of combinations of letters such as "ka"
and "th" just as easily as as you can learn those of syllabics.
But I'll tell you what you can't do.  Upon seeing a new syllabic,
you can *not* guess what it sounds like.  However, in a alphabetic
system, upon seeing a new word, you have a fighting chance of
getting the sound on the basis of the letters you know.  The
analytical nature of the alphabet, which you fault for being
overly abstract, is the alphabet's greatest feature, because
once that's mastered, you have a chance to read any word.

> > That's about it.  With a little ingenuity you probably
> > *can* create a *true* (as opposed to pseudo-phonemic)
> > syllabic script that is *more* efficient than a phonemic
> > one, if you choose to go to the trouble.
>
> But, it would be unnaturalistic and, IMO, boring.  My modified syllabry
> is descended from a true syllabry, the non-syllabic aspects are due to
> sound changes.  I don't know of any natural script that is based on a
> principle of more common sounds using simpler characters.
I wasn't trying to generate a naturalistic script; I was trying
to generate an efficient, practical one that could compete
successfully against a phonemic system.

And actually, it's overly inefficient orthographies that are
unnaturalistic.  People in all ages and places naturally want
to save time, and orthographic symbols naturally assume simpler
forms over the years.  My procedure simply rationalizes and speeds
up the process.  Plus, you can still have creative fun making up
all the less common characters in your system.  Not boring at all!

>
> And I would NOT describe my system as "pseudo-phonemic", it's basically
> a syllabry that has additional complications to permit the larger number
> of syllables in the Classic form as opposed to the Common Kassi for
> which it was originally designed.
>
> > But as the rule rather than as the exception, I stand
> > by my position that phonemic scripts have the edge.
>
> I wasn't saying that syllabries were necessarily better than alphabets,
> only that there were advantages to syllabries.  There are advantages to
> alphabets, too, I will admit.  Which one is better depends on the
> language.  One that has a limited number of syllables would, I believe,
> be better served by a syllabry, while one, like English, that has a
> large number of syllables would be better served by an alphabet or
> perhaps a mixed system.  I think a system that had single characters for
> common clusters would be advantageous.  Perhaps you could write /strajk/
> with, just 3 characters, say, str-aj-k.  The details of such a mixed
> system would, of course, have to be worked out by analyzing the
> phonetics of the language, to come up with a compromise between
> conciseness of writing and fewness of symbols, which would, in large
> part, be a matter of personal choice.
Both systems exist in the real world and they both obviously work.
There are advantages to both and in some cases, a syllabic system
might be the way to go.  However, I would think if you had more
than a few hundred syllables to encode, it would make sense however
to start regularing the characters by reference to their phonemic
components.  That's really nothing more than smooshing together
alphabetic symbols, regardless of what you call it.

Regards

Replies