Replies to Juan, and John & Norbert

Juan:

In responding to your request for feedback on your article, I would like
to stress that our viewpoints are very similar. In general, I am very
enthusiastic about your paper. If I stress some (relatively) minor
differences between us, it is only because that is more interesting than
discussing points of agreement!

I am in basic agreement with you when you write that "in a natural
system we cannot have 'information alone,' detached from any interaction
process past, present, or future: information is always there for a
purpose . . .structural order alone does not represent
information---information appears only when structural order leads to
specific change elsewhere . . ."

However, I would like to make the two following observations:

(1) You seem to be saying that all pragmatic information use is
communication. Here I disagree. I think we have to allow for a single
organism interacting with the inorganic part of its environment as being
an instance of genuine information use, as well. In my work, this is the
interaction I have focused on. It seems to me that "communication"
between two organisms is derivative of this more basic kind of
interaction. (There is the further problem of whether every interaction
between two organisms should count as "communication"---even a
predator-prey interaction---or whether "communication" should be
restricted to cases where one organism is hoping to be "understood" by
the other. In the predator-prey case, obviously information is
"transmitted" in some sense[unless deception is perfect], but not in the
same sense, I think, as in case of communication aimed specifically at
adjusting behavior.)

(2) Our pan-informationist colleagues are not going to easily agree with
us that "in a natural system we cannot have 'information alone'" unless
we can produce an analysis showing how their version of information as a
fundamental constituent of the universe on an ontological par with
matter and energy can be reinterpreted in the light of the
bio-interactionist model.

Maybe we need to lay it out like this. There are two kinds of cases to
consider:

A - some inorganic structure

B, C - organisms

So, the question of the locus of information must be posed for each of
the following four different cases:

monadic:

(A) inorganic, intrinsic

(B) organic, intrinsic

dyadic:

(B-A) organic-inorganic interaction

(B-C) organic-organic interaction

In regard to my first remark above, I simply want to emphasize that the
dyadic relation (B-A) represents pragmatic information use just as much
as (B-C) does. Indeed, I feel it is the more basic kind of interaction.
I am not sure whether you will agree with me about this point, or not.

In regard to my second remark above, most here at FIS want to say that
(A) embodies intrinsic information, while you and I (and Pedro and
Jerry, I think) want to deny it. How can we clarify the point at issue
between the pan-informationists and the bio-informationists?

Well, we all agree that (B) embodies information. What I think is going
on is that there is an unspoken belief on the part of many that at
bottom there can be no difference (or no difference that really matters)
between (A) and (B). That is, an organism is viewed by most scientists
as having no irreducible properties, as being nothing more than the sum
of its parts. To deny this is viewed as "vitalism," which is beyond the
pale. Now, I think that this reflects a mistaken view of
ontology---basically, a bill of goods that has been sold to the rest of
us by the high-energy physicists. The condensed-matter physicists mostly
do not agree. That is one problem with pan-informationism. But there is
another factor contributing to the confusion that we bio-informationists
must also address.

There is nothing wrong with speaking about the "information content" of
inorganic systems, so long as we remember that in doing so, we speak as
human beings looking out on the world. As such, we are instantiating
the relation (B-A). It is when we abstract away from the relation (B-A)
and claim that information exists already in the monadic relation (A)
that we err. I have tried to highlight this problem by introducing the
following terminology:

Call the kind of use of information instantiated in the dyadic relation
(B-A) "autotelic" information. It is information ABOUT the world FOR an
organism. But now, if we want to focus on (A) itself, what can we say?
Well, insofar as it is a part of the dyadic relation, I guess we must
allow that there is "information" in A, so long as we always remember
that it is information ABOUT A and not FOR A. I have coined the term
"allotelic" to describe information in this special sense---information
about one system for the purposes of a second system.

Now, with these definitions in hand, and looking back at the monadic
relations, I will say that there is no information at all in (A) before
the origin of life, and after the origin of life, there is only
allotelic information. Information in the intrinsic or autotelic sense
is restricted to (B), and simply did not exist before organisms came
into existence. I'm thinking that this distinction may possibly help to
bridge the gap between the pan-informationists and the
bio-informationists, although there still remains the question of the
utility of speaking of allotelic information abstracted away from the
dyadic relation (B-A). There may be perfectly good reasons why we will
want to speak about allotelic information, but we must tread very
carefully here. If we are not careful, we may end up building the human
observer into the very foundation of existence, thus landing in the
cul-de-sac of idealism. It is the dyadic relation (B-A) (understood as a
special kind of dynamical interaction) that constitutes the ontological
ground of information in the universe, in my judgment, and we forget
this fact at our intellectual peril.

John and Norbert:

Certainly, it is no accident that the Shannon equation takes the same
form as the Boltzmann equation. After all, one might say that what
Shannon was doing was analyzing the statistical mechanics of
communication channels. But that does not mean that information and
thermodynamic entropy are the same thing. After all, it is no accident
that the equations describing electromagnetic and mechanical waves have
the same form, either. They are both WAVES.

But just as light and water differ in many other respects, so too do
information and thermodynamic entropy. The main way they differ is that
information is intrinsically MEANINGFUL for some cognitive agent. To use
the word "information" in a way that neglects its meaning is to misuse
words, in my view, and to sow confusion.

Meaning is the heart of the matter. Or, more precisely, the VALUE that
structures have for cognitive agents in light of their goals is the
essence of information. So, value is the heart of the matter.
Naturalizing value---explaining how purpose and meaning can come into
existence out of a physical universe that is just minimizing
energy---that is the crux of our problem. And conflating information
and thermodynamic entropy merely sweeps this crucial problem under the
rug. It does nothing to help solve it.

James
Received on Mon Jun 3 14:46:51 2002