Criticizing Linguistics/Shared Cognitions (2)

[Rob Freeman]

Anyone interested in this perspective of "language as computation",
and which is thus beyond abstraction into classes, which I assume
Steve Long is hinting at here, should take a look at these recent
threads on the Corpora list:

Is a complete grammar possible (beyond the corpus itself)?

ad-hoc generalization and meaning

corpus syntax (and how we can use it to code meaning)

http://www.uib.no/mailman/public/corpora/2007-September/thread.html

And more recently the discussion here:

http://groups.google.com/group/grammatical-incompleteness

-Rob Freeman

On 10/5/07, Salinas17 at aol.com <Salinas17 at aol.com> wrote:
> In a message dated 10/3/07 12:42:03 PM, amnfn at well.com writes:
> << By the same token, if by language, we mean the coding of information using
> a limited number of recurring subunits that can be recombined to form an
> unlimited number of messages with an indeterminately large degree of complexity,
> then clearly there can be language without cognition. >>
>
> And that might be the problem with that definition of language.
>
> You speak of "messages"  -- in the ordinary sense of that word -- messages
> from whom to whom? Can a message be from no one to nobody?  Is this message part
> important to your definition of language?
>
> And why should the "recurring subunits" be limited?
>
> Why can't every bit of information be assigned its own non-recurring
> "subunit" -- on into infinity?  So we have a language where every thing, every action
> and relationship or process is its own "subunit," its own code and its own
> syntactical slot.  Wouldn't this kind of language be more efficient than juggling
> "recurring subunits"?
>
> And this coding business -- isn't the process --> information, encoding,
> transmission, de-coding, back to information?  So where is the de-coding in this
> definition of language?
>
> And why does the information need coding in the first place?  Why take the
> step of putting a stand-in -- a symbol -- in place of the raw data?  Can't we
> make an "unlimited number of messages" right out of the raw data without
> unnecessary coding?
>
> Why is the language you are describing carrying so much baggage?  Could it be
> that its constrained by a function?  Could that function be communication?
>
> <<Computer code and DNA code are two examples.>>
>
> Actually we have the whole universe and everything in it.  Coded information
> right down to our tiny recurring subunit atoms, hadrons, quarks, spin, gravity
> and dark matter, etc.  In fact, in this view, language is just a form of
> computation - just a spec in the algorithmic theory of everything, not the other
> way around.
>
> (See Steven Weinberg's review of Stephen Wolfram, "Is the Universe a Computer?
> http://www.nybooks.com/articles/15762
> Also Jürgen Schmidhuber's web site on the Zuse Hypothesis: The Universe as
> Computer -- http://www.idsia.ch/~juergen/digitalphysics.html
> Also Seth Lloyd, Programming the Universe: A Quantum Computer Scientist Takes
> on the Cosmos (2006))
>
> <<Even writing left behind by people long dead is evidence of this principle,
> at least in one direction. Every time we read a message without meeting the
> person who wrote it, we process language that we acquired from an inanimate
> object.>>
>
> Interesting idea.
>
> Some say that you can process language acquired from an inanimate object even
> if it doesn't have any writing on it.  "Signatures of all things I am here to
> read," wrote James Joyce, courtesy of Jakob Boehme.
>
> On the other hand, I can be in the physical presence of at least half the
> people on this planet and not understand a word they are saying, because I don't
> understand the languages they are speaking -- even with my handy book of
> Universal Grammar tucked under my arm.
>
> Another interesting inanimate object is the computer I'm sitting in front of
>  It is the culmination of thousands of thousands of separate "cognitions" on
> the part of thousands and thousands of individuals who go back in time to
> Euclid and before, the people who gradually worked out over a long, long time the
> materials, the numbers, the concepts, the electronics, the processes, the
> software, the distribution and even the marketing.  I am, as an individual,
> incapable of building this machine from scratch, and so are my individual
> "cognitions."
>
> The machine is the culmination of thousands and thousands of "cognitions" on
> the part of thousands and thousands of individuals over along, long period of
> time.  What linked them all together communally is, I believe, language.
>
> Regards,
> Steve Long