Re: Conscious Layer Speculations

Jerry--

Thanks very much for you comments. Judging by the responses, I managed to
take a very simple idea and make it needlessly difficult. While I had a
pleasant meal with William in Vienna, during which we discussed this
matter, I cannot presume to speak for him. But I can try to clarify my own
prior statement.

I hope the list will forgive me for the poor etiquette of the copious
quoting...

>First, Prof. Dockens describes:
>
>> The interfaces between psychology, anthropology, endocrinology, and
>> pharmacology are crucial to a behavioral science approach to the problem of
>> consiousness. Here, consciousness manifests itself from three perspectives:
>>
>> 1) Inside, looking out,
>> 2) Outside, looking in and
>> 3) Inside, looking in.
>>
>Prof. Dockens perspective is close to my own views, but, not restricted
>to the nature of consciousness, but to causality in living systems.
>Thus, one can classify causality as:
>: bottom-up
>: top-down
>: inside-outward
>: outside-inward.
>
>Such a classification allows for much more than layering; it allows for
>a hierarchical logic of organization which includes scalar
>relationships. I think the bottom-up term is necessary to account for
>energy flows and for biochemical processes. Such biochemical processes
>must underlie both pharmacolgy and endocrinology which are included in
>the "problems of consciousness."

We've gotten off on the wrong foot. The key concept from "behavioral
science" that matters is autonomous organizing by some means. Causal
information (information that makes things happen) is-I think we all
agree-central to that. The metaphor of consciousness doesn't translate
directly because people (and groups of people) are conscious. Cells are
not, at least not in the same way. When formally dealt with, such
differences are normally considered in terms of abstraction. This provides
a reasonable way for talking about the nature of the information and its
communication at the level of atoms...

>
>Ted Goranson wrote:
>> >You'll recall that we supposed that there are two kinds of information,
>> >Info (1) which is the ordinary kind, but which I wish in the present
>> >context to limit to kinds of abstractions and messages that support (the
>> >mechanics of) a science and Info(2) the impetus to organize. I submit that
>> >they are different and the challenge is to have them merge, which is more
>> >likely possible in the social layer.
>
>The separation of info(1) and info(2) into these two descriptions is a
>bit uncomfortable to a chemist. Chemists believe that a fundamental
>nature of atoms is that they attract to one-another in very specific
>ways (manners, accomodations, associations, co-ordinations, ...).
>Molecules are organizations of atoms. Biomacromolecules are
>organizations of specific biomolecules.
>Within chemical thermodynamics, the "impetus to organize" is described
>in terms of energy relationships; these phenomena underly the need for
>nutritional sustainance of living systems. Of course, more than
>"imputus" is needed to form new biological organizaions; the specific
>structures must be obtained from the ecoment (surroundings) or from
>storage sources and reformed under the actions of the existing
>organization. Thus, it would be most interesting to learn how the
>notions of "info(1)" and "info(2)" can be made inclusive with the
>perspectives of chemistry and with the notions of causality as expressed
>above.
>
>Ted Goranson further writes:
>> >3. The laws which codify these new abstraction spaces (and ideally the
>> >mechanism which defines them) will be the most basic discoverable laws.
>> >This is to say that the mapping between info(1) and (2) will be not in
>> >terms of information per se, but the definition of the abstraction spaces
>> >which define information. Since Category Theory is the mathematics of
>> >abstraction, it may provide some useful formal techniques.
>> >
>
>I agree with the premise that abstraction spaces are necessary to
>represent complex systems. A basic question is: Are two abstraction
>spaces sufficient?
>Can we consider the term "abstraction spaces" to mean simply "info(1)
>and info(2)" or analogously, microscopically and macroscopically (as in
>Boltzmann)?
>Rosen argues, from a categorical perspective, that Newtonianism is
>insufficient.
>(Boltzmann's equation is closely related to a Newtionian perspective.)
>>From my perspective, I have argued that category theory can be used to
>construct a hierarchy of "abstraction spaces," but that these "degrees
>of organization" (in my terminology) are linked in a manner different
>from (info(1) and info(2). Rather, it is argued that degrees of
>organization are composed as a sequence of metalanguages within the
>"imputus to organize" generated by chemical thermodynamics. Thus, the
>number of "degrees of organization" is open-ended; the logic of the
>grammers must be defined in terms of observations of natural systems.
>It is asserted that within a cell (or any living organism), the
>"imputus" to organize can be classified into the set of directional
>causes given above; proof of this assertion will not be easy.
>Obviously, genetics play a role in the "inside-outward" causal
>processes.
>
>I am reluctant to comment on further aspects of Ted's post because the
>meanings of the wider assertions are unclear to me. Perhaps Ted would
>find the time to ground his assertions within an explicit context space?

..So I proposed two ideas concerning abstraction.

The first is quite simple actually. The way that humans reason about the
universe, say our understanding of how atoms organize into molecules,
involves a set of abstractions that are restricted by our mind's hardware
as well as the conceptual tools we have devised. I call that an abstraction
space, and it turns out to be one we know it rather well, because it is the
home for our own logic, science and language, as well as the simulcrums
we've built in silicon.

But atoms, if they can be metaphorically said to reason, reason quite
differently than humans. With license, one can say it will be similarly
restricted to their hardware and the communicative means that has somehow
evolved at the atomic level. This is stretching the metaphor, but usefully
so. Whatever information is exchanged and which results from molecular
aggregation is native to the space of atom-thought (whatever that is). Now,
what attracts me about the FIS community is the belief that we can get a
handle on this world through considerations of information. My statement
was that the abstractions that constitute the information that involves
atoms and their societies is different than those which humans use. These
differences in nformation spaces will be understood as differences in
abstractions, and certain information mechanics will therefore differ. The
key in my mind is how these differ.

The simple observation was that when people (abstract and) reason, they do
so about the real world. And when they create a science of molecular
binding, they do so in their (human) abstract space. We can say that this
is information (the human kind) about information (the intrinsic atomic
kind). And that we may be able to better understand the abstractions of the
information impetus for atomic aggregation by understanding that difference
and working backwards from our own human-colored abstraction space.

The second observation is more speculative and dates (in my own mind) only
from the Vienna meeting. There would be not two information spaces, but
many. Information related to molecular binding will be inherently different
than that involved in cellular aggregation. I suggest that each will have
its own abstractions; its own domain of thought if you will. I proposed
that we can do even better than looking at two abstract spaces: human and
atomic-instead we can look at a sequence of layers with some primitive
information (read "abstract") space at say quarks, then a molecular level,
then cellular, then animal, then societal-or whatever layering you wish.
The additional leverage this gives us is a set of boundaries between
abstract spaces. Information of some kind is passed among these spaces, say
from Conrad' neurons to his animal's mind. And the consideration of these
boundaries will help understand the nature of abstractions that define the
kind of information that drives the universe.

Category theory is the basis we would use to reason well about such
abstraction spaces, and I have some developing ideas about this. My purpose
is backward from yours, Jerry. It turns out that we do not well understand
the abstractions of societal dynamics, so we have the so-called "soft
sciences" which are forced to use imperfect Bayesian abstractions,
numerical probabilities, the least powerful of abstract entities. So, if we
can understand these abstractions (of say molecular binding) and project
them to the top layer, we might have more robust tools to reason about
collaboration among people as well as atoms.

We can take a shot at what types of abstraction atoms (or better for me,
elementary particles) exist in. Would you like to try that?

The conversation about information and natural language discouraged me.
Natural language only illuminates the human abstraction space. The question
is: what is the "natural language" of atoms. Matsuno's contributions, in
this forum and others, has intrigued me. He focuses on the notion of tense,
and I believe this to be most cogent. It approaches the abstraction problem
through the causal component of language which is intuitively leverageable.

Best, Ted
_____________
Ted Goranson
Sirius-Beta, Virginia Beach USA
757/426-6704, fax 757/721-0781
Received on Thu Jan 07 10:28:00 1999