Re: Universal grammar

LEE Sau Dan wrote:
"Rob" == Rob Freeman <groups@xxxxxxxxxxxxxxxxxxx> writes:

Rob> LEE Sau Dan wrote:

Rob> Incompressible strings can be our mathematical model for
Rob> human language, and as a model they can be a more compact
Rob> representation for information than rules.

>> More compact: yes. More appropriate? Probably not.

Rob> Lee Sau Dan, Why?

More compact representations are usually in forms that are more
difficult to manipulate. They also lack redundancy, which can help
doing error-detection or even error-recovery.

Redundancy? I understand. You think I am suggesting we should make our
representations of language more compact by removing redundancy.

That is not what I mean at all. Quite the opposite. What I am
suggesting is that most (all) of the redundancy is already out of the
system.

I think natural language (as a whole) is already incompressible. That
is what I mean when I say "Incompressible strings can be our
mathematical model for human language." The mistake we have been making
is trying to compress them.

It is the same mistake we have made in maths, trying to reduce
everything to an underlying logic. Abandon that assumption, accept this
resistance to logical description that we see (incompressible) strings
exhibiting, and the problem disappears.

-Rob

.

Relevant Pages

  • Re: Universal grammar
    ... Rob> Incompressible strings can be our mathematical model for ... Rob> human language, and as a model they can be a more compact ...
    (sci.lang)
  • Re: Universal grammar
    ... Rob> I think natural language is already ... Rob> incompressible. ... Rob> "Incompressible strings can be our mathematical model for ...
    (sci.lang)
  • Re: Universal grammar
    ... Rob> LEE Sau Dan wrote: ... Rob> Incompressible strings can be our mathematical model for ... and as a model they can be a more compact ... Rob> Lee Sau Dan, Why? ...
    (sci.lang)