Received: by alpheratz.cpm.aca.mmu.ac.uk id TAA20780 (8.6.9/5.3[ref pg@gmsl.co.uk] for cpm.aca.mmu.ac.uk from fmb-majordomo@mmu.ac.uk); Wed, 1 May 2002 19:36:20 +0100 Date: Wed, 01 May 2002 11:31:30 -0700 From: "Douglas P. Wilson" <dp-wilson@shaw.ca> Subject: Re: future language To: memetics@mmu.ac.uk Cc: Robert Neville <rneville@telus.net> Message-id: <001901c1f13e$6a37e940$856c4518@no.shawcable.net> X-MIMEOLE: Produced By Microsoft MimeOLE V5.00.3018.1300 X-Mailer: Microsoft Outlook Express 5.00.3018.1300 Content-type: text/plain; charset=iso-8859-1 Content-transfer-encoding: 7BIT X-Priority: 3 X-MSMail-priority: Normal References: <20020430224629.38010.qmail@web10101.mail.yahoo.com> Sender: fmb-majordomo@mmu.ac.uk Precedence: bulk Reply-To: memetics@mmu.ac.uk
Trupeljak Ozren <ozren_trupeljak@yahoo.com> wrote:
> Personaly, I don't believe that any language could be clear
> without being rather specific ...
Yes, of course, a language should not -- must not -- force us to be more
specific than we can be or want to be. The ability to make vague, tentative
comments and to hint instead of specify is an essential part of language.
But I don't think a lack of specificity or lack of precision is what causes
a lack of clarity. For me clarity comes from an avoidance of ambiguity, an
avoidance of words which have multiple conflicting meanings.
I try to make a clear distinction between precision and ambiguity. When I
think of precision I think of a gaussian distribution, a single smooth hill
on the plains. An imprecise word covers a wide range of related meanings,
and looks like a low hill, just a small mound, while a very precise word
covers a much narrower range of related meanings, and may be thought of as a
high mountain peak rising sharply from the plains.
An ambiguous word, however, has two or more very different meanings, and
looks like at least a bimodal distribution, perhaps multi-modal, like
several hills widely scattered at various, and possibly large, distances
from one another.
Of course the authors of poetry and literary prose often use ambiguity on
purpose for artistic purposes, but where communication (not art) is wanted
ambiguity is undesirable, and clarity is achieved only when ambiguity is
rare.
If you look at the Schedules for the Dewey Decimal library classification
scheme, a kind of ideal language, you will see a system intended to be
utterly unambiguous, but which supports various levels of precision, which
varies with the number of digits in the code, e.g. 512 is the number for
algebra, an imprecise term, 512.5 is linear, multilinear, or
multidimensional algebra, 512.52 is modules and vector spaces, and 512.523
is just vector spaces. (Actually I think their treatment of mathematics is
awful, but you get the idea).
Trupeljak Ozren also wrote:
> There is a huge number of artificial languages in use today - and one
> property that they all have in common is that they are optimised for
> performance of specific tasks.
I wish that was true. Most special-purpose artificial languages that I
have seen are so poor that it is hard believe any attempt at optimization
was made. What examples do you have in mind?
> Don't you think that there aren't people working on a "perfect"
> computer language, one that would replace all of these other ones?
I sure hope so, though I think Ada serves as a warning against seeking
perfection. But to hell with perfection, I'd settle for a programming
language that was merely good. Python is not bad, but limited, not a
general purpose language -- and everything else just seems poor.
> I invoke Godel's theorem on that and clearly state that
> such a thing is impossible.
You may be right in thinking Godel's theorem applies here, since you say
"all of these other ones" which must therefore include logic programming
languages like Prolog or Mercury as well as Turing machine code. But I'm
not convinced that even logic programming languages need the attributes of
consistency and completeness. Post's and Turing's theorems may be more
relevent, but I'm not convinced that a perfect computer programming language
needs to solve the decision or halting problems, either.
There are mathematical proofs which demonstrate that mathematical proofs can
be automatically converted into computer programs, but I don't think it
always works the other way around. I can still daydream about a perfect
computer language, Godel notwithstanding. A perfect computer language
would make the statement of algorithms and correct formulation of problems
easy, I suppose, but surely perfection in the language would not guarantee
solution of those problems.
> By analogy, your perfect future language is also impossible.
You are making some very large assumptions about what "my future perfect
language", oops, wrong tense, I mean "perfect future language" would be
like. Rather than posting book-length messages and earning the hatred of
all list members, I'll dig up a few URLs that might help.
For the same reason, I won't respond to Trupeljak Ozren's interesting
comments on morphemes in this message -- either I'll post another message on
that single topic, or I'll put my reply up as a web page -- or maybe both.
But here's a hint, for the mathematically inclined -- if a language is a
semigroup, morphemes are generators, and internal (or connotative semantics)
are relations. But even the simplest such structures are absurdly large and
unnecessarily complicated if they use the thousands of 2, 3, and 4 letter
morphemes as generators. Why use more morphemes than our 26 letters or
some slightly larger number of phonemes?
An ideal universal language (as ideal as possible with respect to that
smaller set of generators) would then be simply the free semigroup on those
generators, the semigroup defined by those generators but with no relations
at all. The absence of relations is what makes the universal language
completely empty, lacking memetic content. Any natural language that might
use the same set of grapheme or phoneme generators is a homomorphic image of
the universal language and results from adding relations, and these added
relations are, or reflect, or encode, or result from the memetic content of
the natural language.
> If there is *a language* that is so superiorto all the others in whatever
> it is that languages can be superior in, than there is a rather strong
> probabilty that we would all be speaking it today.
We ARE all speaking it today! All words in these homomorphic images
(natural languages) are words in the free semigroup (or ideal language) as
well, and therefore, ... well, this is already too technical and probably
full of mistakes. I only meant to hint at something to be explained
properly on a web page.
> Without language there is precious little content left. For example, we
> have something that (IMO) corresponds to the role of language, and
> exists in all the cultures across the Earth - music.
I'm glad you mentioned this. My ideal language stuff actually originated
from a study of music. Music and mathematics are both very language-like
things. But except for program music which tries to tell a story by
imitating natural sounds, music has no semantics, or none we know about.
But music has at least got sound, which real linguists insist is the vital
sine qua non of language. Mathematics, on the other hand, though it can be
used to describe sounds, has no sonic component. There's lots of semantics
in math, and syntax too, though mathematicians often try to reduce semantics
to syntax (Carnap's logicism), and vice versa.
Well, anyway, math and music seem strangely complementary, a perception
which led me to wonder about a language which had any and all of music as
its phonological component and all of mathematics as its syntax and
semantics. Does that make any sense to anybody? It's science fiction
stuff, of course, and very hard to think about.
> What you see as "baggage", I see as an essential part
> of what language *is*. It is a feature, not a bug...:)
I regret (somewhat) the use of the perjorative term "baggage", and agree
that the memetic content of natural languages is the good stuff, not
something worthless or mistaken -- it is a feature, not a bug.
Neurophysiologists and psychologists have long looked for the engram of
human memory, the place or method where memory is stored, which seems to be
chemical but also involves the pattern of synaptic connections. Human
memory seems to be holographic, not representational -- although awareness
of memories can be envoked by stimulating specific locations in the brain,
there are apparently no specific locations where specific memories are
stored.
We will never be able to find images from that trip through southern France
by focussing a microscope on any specific neurons, since those scenes are
stored diffusely throughout the cortex in some processed or transformed
pattern that we can know about but not directly access.
Sociologists have long spoken of a different kind of memory, cultural
memory, sometimes called racial memory, and sometime called a zeitgeist, or
spirit of the times, or spirit of a nation. That's not something written
down or stored in libraries or on computers, since it is most certainly
visible in non-literate societies, and it is not something conscious or
visible in any way that I know of. What or where is the engram of cultural
memory? I am quite sure it is in language, it is part of that memetic
content, which we can dimly sense the presence of, but cannot directly
access.
I've been watching the Frontier House series on PBS,
http://www.pbs.org/wnet/frontierhouse/ which concludes tonight. It is
about some people who try to live exactly as frontier homesteaders did in
1883, and should fascinate anyone interested in social or cultural history.
It is very well done, but has a fatal flaw. Those families may live and eat
and dress and work exactly like people did in 1883, but they don't TALK like
anyone in 1883 ever did.
Their conversation is full of modern ideas and even when they don't discuss
these ideas directly, their speech is full of words and phrases which
represent and stimulate them. There is a lot of pop-psych jargon from the
self-help movement, and the children know a lot of short cliche expressions
like "no way" and "don't even think about it" which no child in 1883 would
ever have said. (I wonder if I could get a transcript of some of those
conversations, from which I could easily extract the pithy words and
expressions which represent our modern 21st century attitudes -- that would
be illuminating, I am sure).
These children, more so even than the adults, who do it too, clearly know
how to use these little key phrases to dodge chore assignment, start or
resolve conflicts, and manipulate other people in ways nobody in 1883 could
ever have done.
The people who organized this rather amazing exercise in social history
re-enactment have somehow managed to exclude all the modern technology they
could see, but they have failed to exclude some of that "baggage" which
encodes or suggests or stimulates the use of a lot of modern social
technology.
dpw
http://www.socialtechnology.org
===============================================================
This was distributed via the memetics list associated with the
Journal of Memetics - Evolutionary Models of Information Transmission
For information about the journal and the list (e.g. unsubscribing)
see: http://www.cpm.mmu.ac.uk/jom-emit
This archive was generated by hypermail 2b29 : Wed May 01 2002 - 19:48:44 BST