Error during command authentication.

Error - unable to initiate communication with LISTSERV (errno=111). The server is probably not started. LISTSERV 16.5 - CONLANG Archives

Print

Print


I am still thinking about this, and trying to read what I can about
related linguistics on Google Scholar & Google Books. So far this has
mostly been interesting stuff arguing that there can be no such thing
as a written language. :)

Anyway I wanted to mention an example of what I would want this theory
to be and why the theory would involve more careful investigation and
work than it would nice insights. () below means something like a
sidenote in a really rough draft which will have to be taken out
before it can be made final; {} means a more thoroughly irrelevant
comment.

Many scripts have a letter which is a zigzag or curvy line, like the
English S, because this letter is easy to think of, write, and
recognize. (I do not actually know how many or what percentage or
anything.) However, the letter is unstable because it can be
conceptualized multiple ways. An added flourish can easily be seen as
an extra curve {I have been playing with using up- or down- flourishes
as units of meaning added onto an alphabet, and saw that the top end
of S could either curl down into a spiral or up into the start of
another curve}. Since the letter itself is only curves this can go so
far as to make it reverse. Therefore this simple letter can quite
easily develop a fairly complex history or spread of allographs. (A
good example of this is the German esszet, which for a time was
ambiguous between being a digraph of 'ss' and of 'sz' partially due to
allographs of z having an extra flourish making the difference
negligible. However, that was also possibly influenced by the
phonology: <sz> = [sts] was easy to reduce to [s].)

Another example of this can be seen in the letters <M> and <N>, which
over time led to the letters <m> and <n>. Though <N> is up-down-up,
<n> can often be observed in handwriting as down-up-down, and if it
were not for the pressure of differentiating from <u> as well as the
influence of the typewritten form it might well become fully flipped.

:) OK, done pretending to write a paper. The fact is I know nothing
about "curvy" letters outside Europe, and need to read more. And
besides, English sure has a lot of them under at least loose
definitions, showing that these rules I seek to uncover sure don't do
much constraining of the possibilities. Still, I hope there are good
evolutionary tendencies in alphabets; I want to know a lot more about
what makes a 'realistic' constructed script, and if there are no
universals regarding scripts then anything goes!

Let me know if this is too off-topic for the list.

On Tue, Sep 8, 2009 at 4:41 PM, Daniel Demski <[log in to unmask]> wrote:
> Is there such a thing as ... forgive this horrible name ... an 'IPA
> for written language'? Naturally that's absurd because you can just
> write a written symbol to represent it, but what I mean is: that the
> IPA chart lists the common variations produceable by the human mouth
> and distinguishable to the human ear used in human speech, which
> encompass most of the sounds used across the world. It seems to me
> there are analogously variations in symbols which are produceable by
> pigment on a surface and distinguishable to human eyes which encompass
> at least a fair fraction of the world's writing systems.
>
> Naturally it seems the space of possible written symbols is more
> complex, but that doesn't keep similar letters from popping up in
> diverse world writing systems.
>
> Some of the variations which can be either meaningful or not are,
> which direction a letter is facing, how tall it is, whether a
> particular shape is closed ( <c> versus <o> versus some peoples' <u>,
> some peoples' <u> versus <a>, perhaps <F> versus <P> ... am I right in
> recalling <>=written?), boldness of a line, 'crossed-ness', etc.
>
> However, such a system might need to admit that letters are
> fundamentally made up of 'strokes', which obviously would make things
> pretty complicated.
>
> Ideally such a system would allow us to take an arbitrary lexeme and
> describe at least its fundamental features without needing to draw
> it-- we could describe the <m> as essentially double-peaked, with an
> upper left corner initiation, that is, starting point for the pen; and
> there should be some good way of saying the peaks are connected such
> that three vertical bars are prominent to the letter. <M> on the other
> hand does not have the initiation or the connection, just two pointed
> peaks and two prominent verticals.
>
> (I am often surprised that googling exact phrases even as short as "we
> may describe this letter" (trying also all the can/may, the/this, and
> other variations) doesn't get me any exact matches. Nobody has said
> "we may describe the letter m as two peaks"?!!)
>
> What initially got me thinking about this was that English fonts vary
> things which don't change the letters' meanings in order to get
> variety and beauty. But these things could be meaningful in other
> languages, and there are things like rotating or reflecting letters,
> or adding loops, which can make more problematic fonts.
>