Ray Brown:
> row and columns, etc.

My opinion: row&col format is *more* 2d, but not fully. I suppose I'd
call it "linear-plus", or "double-linear" or somesuch. 2d to me
implies not being constrained to work in a particular direction -
normally, this is "horizontal" (broken vertically, or vice versa,
depending on your script orientation); row-and-col would add one more
to that, but would still be working within a fixed path. What I see is
something more freeform (not necessarily unregulated-freeform, though
I can see that too) - that is, jutting out in all directions,
progressing as a web.

I don't think any lines are linear, though; that's definitely a far
more "zealot"-y position than I take. I would be very intrigued by
other suggestions, but the simplest and most intuitive way to show
connections (in the atoms-and-relations model), to me, is a line.
Possibly a curvy / pretty line with squiggles or whanot to indicate
different types of connection, but still a line. And I believe I've
already made clear that I do agree that some linearity may be
necessary or unavoidable, e.g. for portraying phonetics, or for giving
a cognitively accessible description of events-over-time (as in a
"normal" story).

As Ray says, though  (and I agree) - I don't think there's anything
*wrong* with doing it in row&col or anything else... it's just not my
conception. Whether others want to stick to *my* conception or not is
up to them; I am not militant about my ideas. ;-)

> I'm trying to talk about what Sai envisaged. Whether that is impossible to
> achieve or not, I do not know. I did express my doubts early on, but Sai's
> enthusiasm has rubbed off on me. Unlike you, I do not have this dogmatic
> certainty about its impossibility (or even about it's possibility). As Sai
> wrote on Thursday, May 5:
> "But we're in the business of *creative* linguistics, are we not? I for
> one am not interested in constraining what language can do merely because
> it hasn't been done before in a natural language."

*grin* I am quite glad that my 'vision' as you call it is catching on.
I've had difficulties getting the concept across, as you may have
noticed. :-P

IMHO, it's a very intriguing concept, and possibly quite powerful, but
poses some difficulty in working out a good (i.e. 2d-native [as it
were]) way to *do* it, given how hard it is to break away from one's
linguistic rut.

> wormholes


> fractal WS

*nod* I think it's very much a related point - and one that would be
easier to do in a NL2dWS, though might require higher tech than
pen-and-paper. Zoomability and hyperlinks and all that. Do you have
ideas for how to do 'fractal'-ness other than simple zoom?

> lines != linearity

Yup. Though I think some would - e.g. linear agglutination (per a
previous suggestion; reminds me rather of Chinese / Japanese
compounding) is probably not the most "2d" way to *do* agglutination,
or compounding, or whateverthehell the equivalent would be. So in that
sense, I feel that it is a linear *component*. One way to phrase my
position would be that I would prefer to minimize the number of linear
components, unless they are demonstrably better than their 2d+
counterparts... and we haven't yet worked out an example of 2d fusion.

Ray Brown again:
> I have just come across some stuff by Roger Penrose which seems apposite to the notion of *non-linear* full-2d writing systems. I thought perhaps that quoting some of it might help in making clear how I understand the sort of thing Sai had in mind when beginning this thread (I am quite sure he will put me right if I have got it wrong   :)

*grin* Indeed I would, but you seem to have pretty well grokked my
intent. I like the quotes too. ;-)

It seems to me that an inevitable question that will (and must) come
up with this is the mapping-out of thought itself. I think it rather
self-evident that we (or, at least, Penrose, you, and I :-P) do not
think in words, "natively". To me it is very clear that verbalization
is a translation process - one I do quite well, mind you, just as I
can pretty well translate between the languages in which I'm fluent -
and, more critically, a process of putting into symbols that which is
not natively so.

"The thought that can be named is not the true thought." ;-) Or,
equivalently, "To know a thing is to forget its name."

That is, what we refer to with language is *NOT* really a description
of the thing we are communicating, whatever it is, save for trivial
examples. It is an *evocation*. In poetry, this is most obvious, but I
would contend that it is so for nearly every other form of symbolic
communication between sentients. We merely evoke that which we
(hopefully) already have in common knowledge. (Yes, this is very much
at odds with some [*cough*] theories of linguistics, particularly
computational ones.)

This, btw, is what koans and suchlike are for - they are rather
explicitly merely intended to point at the elephant. As are pretty
much all meditation & other spiritual practices, but it seems to me
that most people forget (or never realize) this, and come to believe
that the practice itself is the elephant.

So what I would like to do is figure out how exactly we *do* think,
natively, and to make this translation-and-symbolization process
mangle it as little as possible. I do not believe it possible for it
to *not* be mangled save for telepathy (and even then...), but it can
get closer, and be more self-aware as a language about the fact that
it is merely evoking, and focus on doing that better. It is a
necessary process - we do need "handles" (to use a CS analogy) into
our memory, to be able to evoke mindstates / memories / thoughts /
reasoning at will, or to do so in others; my notes-to-self are
generally cryptic for that reason, as they are rarely intended to
provide actual information so much as trigger something - and it is
frustrating to me to have it be such a kludgy one at present.

All of that make any sense? Should we start a spinoff thread about
mysticism? :-P

 - Sai