Well, there is a first time for everything. Thanks Mike Tintner. This is a
useful explanation.
Cheers,
~PM
From: tint...@blueyonder.co.uk
To: a...@listbox.com
Subject: Re: [agi] Semiosis
Date: Sun, 7 Apr 2013 13:12:22 +0100
PM:
What
if signs = concepts. Does that change anything?
This
misses the whole point of semiosis/semiotics (although to be fair many
semioticians don’t get it, as wiki indicates). You, like linguistically
oriented
semioticians, want to pin reference down.
The
reality is that there are a vast and ever proliferating array of sign systems –
and EVERY SINGLE ONE has a unique function in portraying any object.
Take
cats.
There
are words that label cats, cartoons of them, photos of them, sculptures of
them,
paintings of them, diagrams of them, graphics of them, x-rays of them, maps of
them, sound recordings of them, music of them, movies of them, mathematical
descriptions of them, geometrical portraryals of them, logical references to
them...etc.
And
EVERY SINGLE ONE portrays some different aspect/dimension of cats. Fact ... or
certainly truth.
So
your concept of cat or any other object has to be a v. fluid affair, that can
link and metamorphose into any and all forms of signs. This is v. hard not just
for AI-ers but anyone in our present culture to grasp, because we are still not
yet philosophically a true multimedia (vs unimedia, logocentric) culture that
appreciates the value of all sign systems.
[METAMORPHOSIS
is a key dimension of all concepts - all concepts have to be capable of
endless metamorphosis - because the objects they refer to are capable of
endless metamorphosis].
Figures
are the natural form for any concept because they can be fleshed in to form
almost any other kind of sign, and because they represent the framework of
objects. And it looks exceedingly like that is what the brain uses, because
that
is how we visibly use figures/graphics as actual, external signs. Note how the
Cartesian grid is the basis for a vast array of signs – and how the atlas of
maps forms a similar basis for many other signs.
Figures
are also the natural form for concepts because they link straight into body
movement – provide a structure for body movement. No other sign system can
easily do that. And again, you can look at our externalised use of sign
systems, and see that is exactly how we use figures – for directing movement
- e.g. the figures of an Ikea manual, or outlines of a choreographic
routine, or battle plan, or route-map.
The
link between representation of objects and movements and enaction upon objects
and movements is of massive and central importance. Language doesn’t connect to
movement. Logic doesn’t connect to movement. Maths doesn’t connect to movement.
But figures/graphics do.
To
sum up so far: we need ALL our sign systems to represent objects, not just
one or two or three. Not just the 3 r’s for example. And a brain needs its
natural equivalents of all these. We need our own mental movies, sound
recordings, still photos, rough outlines/shadows,
words/names/symbols....
But
graphics/figures provide a natural central framework for linking all these
different sign systems – and for linking representations of the world to
actions
upon the world, for linking outlines of objects to guidelines for actions upon
them.
Thirdly,
one last fundamental dimension of concepts to realise is their “holy
ghostliness”. The literate mind wants to pin concepts down, as PM quite
naturally does – to be definitive.
But
concepts don’t have and don’t need any definitive form. It doesn’t matter what
concept/figures of a cat your brain starts with. Whether its first figures are
say a cat sitting or moving, or curled up. Since every concept is “by
definition” metamorphic, it is fundamental to concepts that they can and almost
certainly will evolve - in line with your knowledge and experience of the
real world objects. Concepts of objects can start anywhere and evolve
anywhere. Words may appear to have a definitive dictionary meaning
(although in truth there is nothing definitive about dictionary definitions,
but
concepts dpn’t.
P.S.
I doubt that there is a single multimedia AGI-er – a single person who doesn’t
think that an AGI will be able to be human-level inteliigent with just a
handful
of sign systems, esp. language, logic, maths and maybe some movie capability –
whereas human-level intelligence actually involves a capacity for ALL sign
systems.
AGI | Archives
| Modify
Your Subscription
-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424
Modify Your Subscription:
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657
Powered by Listbox: http://www.listbox.com
