Data and meaning

Dear Colleagues,

Jerry Chandler wrote (Sept 28th), answering John Holgate (sept 20th):

>As John notes, the separation of meaning is not restricted to grammar
>and Shannon information theory. Most philosophers of mathematics
>claim that no meaning is attached to mathematical rules or
>structures. By way of contrast, the language of chemistry and
>chemical grammars are meaningful - a matter of common experience.
>If one assumes that formally, "information" has no meaning, then one
>must turn elsewhere to identify the source of meaning and
>communication.
>As a logical parallel, I would point out that the nature of entropy
>(as a physical phenomenon) depends of the co-existence of temperature
>and energy.
>In bare bone terms:
>you get nothing from nothing;
>you get something from something.

The two points brought up by John and Jerry underline aspects of
our discussion that could deserve further developments:

1) Separation of meaning from information:
As already written in a previous post, I feel that a meaning
cannot exist without an information carrying it.
In other words, a meaning is an information, a meaningful information.
(Information being the component of a signal, a variation of energy.
Ex: sound, light, chemical element, protein,...).
In our brains, mental states are related to variation of
neurotransmitter patterns. The meanings in our brains are based on
information status in neurons (understanding that a lot is to be done in
this field, specially regarding the nature of consciousness).
But this does not mean that all information is meaningful. Many
information are meaningless, but most information can participate to the
creation of meaningful information Ex: Noise from thunderstorm is not
a meaningful information per se. But this noise in going to participate to
the creation of a meaning in our brain. For instance, the created meaning
will be "rain is comming". And this meaning will generate an action
"look for shelter".
So there is an important difference between a meaningful information
generated by a system, and an incident information that will participate
to the meaning generation. The incident information can be meaningless.

2) Origin of meaning:
If we agree that a meaning is a meaningful information, we have to look
for the generation of this information.
As some already know, I feel there is a way to modelize a
"Meaning Generator System" (MGS) as being a system submitted to a
constraint that has to be satisfied.
If the MGS receives an information related to the constraint, it will
generate a meaningful information that will participate to the triggering
of an action needed for the satisfaction of the constraint.
For us, as MGS hearing the noise of thunderstorm, the constraint we
have to satisfy is "avoid getting wet in the rain". The generated meaning
with corresponding action will be "rain is coming, lets look for shelter".
Such examples of meaning generation correspond to simple cases where
the MGS can be cleary defined. I feel that access to more complex cases
is possible via the analysis of corresponding MGSs with identification of
existing constraints.

Also, I feel that these subjects could be part of Luciano Floridi's PI.

Regards
Christophe Menant (crmenant@free.fr)
Received on Sun Sep 29 14:59:54 2002