Re: [Fis] leteral comment - reductionism

Loet stresses that the mathematics of information can be applied to
conceptualize anything. This notion also appears in Pedro's separation
of information as parallel to the scalar levels of reality. Rafael
reminds us of "being inside", with many interesting references. Pedro
makes a beautiful example of energetic value vs psychological value.

I think these issues pinpoint the relation between physics and
information. The "systemic" perspective tried to propose metaphors of
reality: where is the information in this reality? But there is a
parallel world of information, and it can be studied for its own sake.

I will now try to apply some Aristotelian analogies to information,
attempting to define it. The central element is "model", and there are
four causes to it. My "model" roughly corresponds to the notion of the
Sign in semiotics. I think this representation may help illuminate some
of the meta-level disagreements in the recent discussion, and perhaps
reconcile the informational stream and the physicalist stream of
thought. These two streams merely focus at two different aspects of the
same thing: they don't exclude one another.

1. The constructs the thoughts are made of (the materiality of language,
the building blocks of information)

Pedro wrote:
> So, discussing in depth the series of
> bio-economic metaphors raised by Aleks would be quite interesting
> (impossible here, but just as a detail, the ATP is presented as
> 'currency' in most textbooks: wrong, at is just pure 'fuel'; the only
> cellular currency may be established around "second messengers" cAMP,
> Ca++, cGMP, etc.).

In Mongolia the "currency" used to be chunks of dry blood: you could eat
it in cases of emergency - so fuel can be currency. In the very old days
in Europe the currency were chunks of bronze that you could convert into
swords and those you could wield to obtain food. The currency a flower
pays the bee is glucose. If you trace back the origins of currency, you
arrive at fuel. If you trace back the origins of fuel, you arrive at
energy. Pedro is right in saying that the thread is overstretched, but
it's still a thread. The beginnings and the endings are lost in a blur
of vagueness in becoming and ending. When is it currency and no longer
fuel? When does a chicken stop being an egg? When is the egg no longer a
part of the chicken?

"Becoming" and "ending" are a reflection of the "path" metaphor that
Lakoff and Nunez wrote about (as Terry reminded us last year). We wield
these metaphors to comprehend the world. But the actuality of the world
is not made of beginnings and endings. Our *understanding* of the
reality of the world is made of beginnings and endings. Beginnings and
endings are the crayons we use to draw our mental maps.

I think "information" in the context of FIS is the discussion of
metaphors the knowledge is built from, the structure of knowledge.
Shannon information is one of those metaphors. Probability is yet
another metaphor. Both of them rest on the kind of foundations of
ennumeration that Karl often mentions. Igor and Bob speak of network and
autocatalysis metaphors. Stan upholds the hierarchical metaphors.

What metaphors are there? Can we build a dictionary of them? Can we
build a grammar of them?

==

2. The shape of the information world (the formality to information, the
constraints placed upon the information)

Pedro wrote:
> Then, a pressing issue to consider in this tentative revision of social
> ascendancy may be that social information is very largely "decoupled"
> from entropy growth (as biological information itself).

Our minds and computers are all connected. Thoughts jump around, build
up, collapse and flow. Symbols are made, forgotten, mapped and matched.
Sometimes the realities thus constructed are well-connected to the
unknown and often unknowable actualities below. A mental image of a
castle perched on the cliff is matched by an actual pile of rock. A
mental image of a castle on a cloud in the sky is not matched by any data.

So, our minds construct all sorts of ideas. Some ideas are grounded in
energetic reality down below, and some ideas are pies in the sky. Just
as the wind can twist and heat the air to no avail, the same wind can
blow a house away. Some ideas productively modify humanity, and some
ideas make the humanity turn in aimless circles.

We can thus examine how the parallel world of representations looks like
(formal). We can think about what the representations are made of
(material). We can study our minds that host and nurture the information
(efficient). And we can wonder about the purpose of the information
(final). Do these four causes define information science? Am I myopic in
my thinking that "information science" is the study of models?

No wonder that Pedro and others find systemic views reductionist: they
indeed focus primarily on the efficient cause of information (where is
information, what causes it?). Sometimes we look at the final cause of
information (what is information good for? will it sustain itself?)

So, if there is information in a cell: (formal, structure) What is it
like? What does it mean? (material, elements) What is it made of? What
is the language? (efficient, "how") what is its physical nature? How
does it work? (final, "why") Why is it used? What is it good for?

It's easier for a computer database information:
   * material: bits and bytes
   * formal: relational database
   * final: human users' proximate goals
   * efficient: computation

Or for scientific information:
   * material: algebra, geometry
   * formal: empirical data
   * final: truth
   * efficient: maximization, computation

For business thinking:
   * material: situations, actions, decisions (usually not explicit)
   * formal: market, resources
   * final: profit, values
   * efficient: maximization

For (over)simplistic evolutionary models:
   * material: genes
   * formal: environment
   * final: survival, fitness
   * efficient: selection, mutation

One might wonder whether the more specific notion of a FORMal cause
would better capture the notion of "inFORMation". Formal causes are the
constraints, the boundary conditions, but also the correlations. That,
of course, is a possibility, but wait: if we are describing the real
world, and we employ the formal cause as a descriptive metaphor, the
formal cause is a model. The scientific "data" is not objective, it's
itsef a model (made of numbers or ponts, formed in tables or graphs,
collected through measurements, and chosen to satisfy a scientist's
ambitions).

Of course, there may be better foundational metaphors for all this than
my interpretation of Aristotelian causality. But to avoid too much
confusion, "information=model=representation=signs" might be a good
definition that liberates information from its building blocks,
meanings, purposes and mechanisms. It's about meta-thinking, where a
thought, a model is in the focus. We can apply meta-thinking to the
thought of a thinker. We can apply meta-thinking to the thought of a
purpose. We can apply meta-thinking to the thought of a cause.

3. Aristotelian models vs semiotic signs

I will try to map the four-fold causality to the notion of the sign. A
sign is a trajectory from input (object) through mediation
(representamen, vehicle) into an output (interpretant). The model in
Aristotelian scheme corresponds to the Peircean output. The input is the
formal cause of a model. But I decompose the mediation into generativity
(material), mechanism (efficient) and purpose (formal).

For an internalist the world should be pure firstness. But for a
practical discussion, we use various metaphors to convert it into
secondness and thirdness. In the example of scientific thinking, we
pretend that the "data" is a correct representation of firstness. An
internalist scientist will be aware that this is already a simplification.

I would be grateful for comments, criticism and corrections.

-- 
mag. Aleks Jakulin
http://kt.ijs.si/aleks/
Department of Knowledge Technologies,
Jozef Stefan Institute, Ljubljana, Slovenia.
_______________________________________________
fis mailing list
fis@listas.unizar.es
http://webmail.unizar.es/mailman/listinfo/fis