Peter,

[you wrote]
I'm not questioning the view that different hangul consonant jamo correspond to points of
articulation -- the same would be true, for that matter, of the Latin letters "p", "t", "k"; that's
not particularly surprising. But these structural units (both the jamo and Latin letters) are
representing phonological objects comparable to phonemes.

It is not the case that a single jamo identifies just point of particulation (e.g. velar) and must
be combined with a different jamo that identifies manner of articulation (e.g. oral stop vs. nasal
stop), as well as other jamo for other similar things. (Imagine a Spanish writing system in which
the phoneme /k/, which is a voiceless velar oral stop, was represented by a sequence of characters,
"^o|" where "^" represents velar tongue position, "o" represents oral stop and "|" represents a lack
of voicing.) But that's what it would have to be like to say that the writing system had structural
units that corresponded to phonological *features*.
[mariano]
I like your way of thinking and have somewhat long comments. I understand -somewhat reinterpreting-
that you mean that an alphabet made of symbolic traits for each phonemic distinctive trait (as far
as the concept of *phonemic distinctive trait* is valid) would be a complex system and quite
difficult to learnt, and probably not economic for effort (but let's say, by the way, that the
distinctive phonemic traits are less in number than the number of phonems, so that with, may be,
just 9 symbols we would have enough for many more phonems). Le me call this kind of alphabet an
*phonemic distinctive traits writing system* because it is not and alphabet and makes explicit
expression of each phonemic trait for each phonem. Never has been made such a kind of *phonemic
distinctive traits writing system* in wich each phonem would be represented by a cluster combining
some of a minimal small set of symbols. I do not think that it is an imposible task, for example, it
might be implement by means of elements like meaningfull main strokes and smaller secondary and
terciary strokes -or more as needed- that would be combined at least in part by overlaping- but I do
not feel confident with respect that the distinctive phonemic traits that have been described in
phonetics are the relevant ones.

I think that the *han-gul*, and may be also the Bell's *visible speech* are systems that use an
abstract iconicity for articulatory traits. That is, both make use of just some of the phonemic
articulatory traits from the phonems they represent but not all of the articulatory traits of those
phonems -because that I use the term "abstraction". So, that, if I have understood you, these
systems are not a *phonemic distinctive traits writing system* but systems that make abstraction of
those articulatory traits of the phonems by, in the best case, choosing the prototipical ones to
represent each whole phonem.

I think that one of the important questions to be very aware is that both, *han-gul* and *visible
speech* are consciously made systems by a human researcher and that they take into account not the
acoustic traits of phonems, but their articulatory traits. I think it is important, because what
really likely does matter in cognizing speech sounds as phonems are not the articulatory traits, but
mainly the accoustic traits and probably secondarily some of the articulatory traits.
Some time ago I was trying to make an alphabet (for Spanish) by watching the phonems spectrograms,
in the spectrograms what you see are the acoustic traits (for example the frequency formants for
vowels, the vertical bars for interrupted phonems or the irregularity of the frecuencies for
fricative phonems) an I realized what I think is that the latin alphabet already has somewhat a kind
of abstract iconicity for the acoustic traits of the phonems in quite a number of its letters, as
well as that there are a few letters that accumulate many irregularities and a lack of such abstract
iconicity. So I mean that there are some evidences for saying that the latin alphabet in its letters
and with respect their more widespreed pronuntiations has happened by a process anologous to
*evolution* a change towards an abstract iconicity for the acoustic phonemic traits, but still
bearing some arbitrariness.

For example if:
A. <t>, <d>, <p>, <q>, <k>, <b> are understood as interrupted and having a vertical stroke as an
abstraccion of it.
B. <s>, <f>, <j>, <z>, <y>, <g> are understood as "fricative" (irregular frecuencies trait) and
having a wave like or zig-zag stroke as an abstraction of it.
C. <m>, <n> are understood as voiced nasal and an horizontal line stroke as abstraction of it.
D. <l>, <r> are understood as liquid and a loop big or small stroke as abstraction of it.
E. <i>, <e>, <a>, <o>, <u> are understood in that order as gradually lower frequency voices and a
circular stroke that (like musical notes) is drawn in the highest position for <i> and the lowest
position for <u> (here might be done a reasoning about how that circular stroke would deform
according with the position, the space constrains for the letter character and the constrains for
visualization that does not allow unmarked too small figures)

A background idea is that the brain do like patterns and works by a law of the least effort, better
than with aleatory marks;
a thing that psicologist and neurologist may study by making the apropiate experiments.

I have recieved about these ideas in other forums no response, some agreement, mild criticism and
also rejection.
What do you think?

Yours cordially,
mariano