> Dan Clemmensen wrote:
> > We are now confronting the near-term possibility
> >of remodelling our brains, via uploading or computer augmentation.
> >Will we need new linguistic constructs to use the new hardware
> >effectively? can we get some idea of what we need before the
> >hardware is available. This is roughly analogous to designing
> >a new higher-level computer language in conjunction with a
> >new processor.
Language in general does not efficiently encode emotional states or
spatial/temporal relations. Just for fun, ask someone to describe a
spiral staircase. It seems impossible to do this in casual speech without
making a spiraling motion with the hands.
> What about expressing multivariable dynamically-changing probabilities?
>
> What about ways of expressing the passage of time for one of you that's
> still and another of you that's moving at .99c, when you're "talking to
> yourself"?
>
> Personally, I've always been at a loss for ways to adequately express
> joy and wonder, and I could hope that a new language would help with
> that, too....
Language, as great as it is, is not a perfect communication technology.
We think lots of things we can't express in normal language; that's why we
also use mathematics, music, visual arts, body language...
A superior form of language, one which does all that modern human language
does as well as efficiently describes spatial/temporal relationships,
emotional states, etc., would necessitate an evolutionary change. A
language which does not conform to Universal Grammar and the proposed,
concomitant Universal Semantics would _not_ be comprehensible to us right
now.
Consider apes who have been taught some sign language. They can use
words, which is no mean achievment, but they cannot use syntax; it is a
system which they are cognitively unable to process. Consequently while
a signing ape can understand a biologist signing "eat banana?", they are
incapable of understanding the difference between "You eat the banana" and
"The banana eats you" which is entirely a syntactic distinction. This is
demonstrated by signing apes' strings of words, along the lines of "me eat
eat eat me eat me me eat".
The stages of child language acquisition give us an idea of the previous
evolutionary stages of human language. First, single words: "cat" "doggy"
"milk" (actually this comes after non-linguistic babbling and then
babbling using the phonemes of the child's linguistic environment,
including hand gestures for babies around signing parents, but anyway.)
Then come two word utterences, at which point the child already
demonstrates a basic sensitivity to syntax. Then comes the "telegraphic"
stage: "Kitty come table?" "We go home." "Look pretty bird!" Both these
stages show a prefered word order (used something like 80-90% of the time,
even in languages with flexible word order like Russian). The vocabulary
grows fairly steadily through these stages, slowly enough to keep a fairly
accurate count of the words the child uses. THEN comes the explosion:
the child starts using full sentences and demonstrates a vocabulary too
large to be counted easily.
At each of these stages, children seem to be able to understand one degree
of complexity greater than they can produce. This gives me the idea that
the next stage of linguistic evolution is something adult humans may be
able to understand but not employ themselves.
Hey, come to think of it, this is already the case. I can understand
Shakespeare, but I seriously doubt I could produce language of the same
level. Great writers encode thoughts more efficiently than can non-great
writers.