Re: Information & Physics

From: Jerry LR Chandler <[email protected]>
Date: Sun 11 Jan 1998 - 04:03:09 CET

Dear Koichiro:

In order to stimulate this conversation, I will attempt to respond to
your very insightful questions from the perspective of a biochemist. I
probably am not adequately trained to understand the physical point of
view; nevertheless, another perspective may help the conversation get
started.

koichiro matsuno/7129 wrote:
>
> Dear FIS Colleagues:
>
> 1) Information of the Population One
> If a population of many individuals is available, it may
> be possible to assign to each representative inidividual its
> information content. In contrast, if the population has only
> One Individual like the Gaia, how could one assign to it its
> information content? Or, is it a wrong question?

I think this is a question about closure - what is the domain of
discourse?
Information theory starts from the notion of a probability space and a
probability distribution. These terms are purely matheamtical in nature
and no material or dynamic constraints exist on them. Thus most any
proability space will do and any probability distribution will do. One
mathematical constraint on any probablility distribution is that the sum
of the probabilies of possible events (objects?) equal one.
If the probability distribution is composed of a single object, then the
probability is one and no uncertainty exists (mathematically). From the
definition of Shannon information, the information content is zero
(since ln 1 = 0).
This first question introduces the general problem of the necessary
condition for congruency between observations, semantics, and symbols.
Clearly, our planet earth is composed from local dynamic processes which
are information-rich. By an appropriate choice of semantics and
symbols, one can contradict the observations (at a different reference
scale!) :-) :-)
>
> 2) Information and Thermodynamics
> Although this dichotomy has attracted lively attention
> from many physicists for almost a century by now, the issue
> does not seem to be settled yet. Is information an
> unnecessary appendage to thermodynamics or vice versa?

By definition, classical thermodynamics is a science of heat
tansformations. Can anyone provide an example of observations which
estabish a one-to-one correspondence between units of heat and discrete
units of information? I believe that intrinsic to your second question
is the issue of the relationship between continuous and discrete
functions. Does information necessarily have to be discrete as proposed
by Shannon? Human bodies can function in accordance with thermodynamic
principles. Are human judgments composed from only discrete
information? Or are continuous functions necessary for human judgment?
Or are human judgments composed from both discrete and continuous forms
of information?

> 3) Information and Quantum Mechanics
> A notorious problem with quantum mechanics is
> measurement, the latter of which seems an indispensable
> attribute of information. In addition, quantum mechanics
> is quite entangled in the manner of specifying its
> wavefunction nonlocally, not to mention quantum teleportation
> of a recent Innsbruck vintage. How can both measurement in
> QM and quantum entanglement be related to the issue of
> information?

Is this question an artifact of the assumptions of quantum mechanics?
In other words, I could view this as a different form of a "closure"
problem.
Does Quantum Mechanics start from a systematic perspective of
interactions between ALL particles in the system and assert that a class
of differential equations will describe the evolution of the system?
Do these differential equations intrinsically assert closure to any
other influence? In other words, in an analogy with a thermodynamic
sense, is a quantum mechanical system an isolated system? Is this the
source of the measurement problem? Is it possible to make a metaphor
between the measurement problem and "begging the question"?
>
> 4) Information and Conservation Laws
> Physics has founded itself upon various conservation
> laws such as energy conservation. How could information of
> an evolutionary and historical character be compatible
> with conservation laws? Are they an odd couple?

Conservation laws come in many forms and have many uses. But, I am not
aware of any conservation law which specifies all possible aggregations
of atoms into chemical forms or the dynamics of interchanges among such
chemical forms. The chemical composition of the matter within living
systems can follow conservation laws and still have access to many
potential dynamical behaviors and many potential chemical forms.
Biological evolution is associated with freedom to explore that which
is not physically or chemically forbidden. Analogously, human cultural
evolution continuously emerges into new patterns despite the
ever-increasing numbers of do's and don't's (rules, regulations, laws,
etc). So, is it the existence of the rules which forbid some behaviors
that motivates systems to search for new behaviors which do not violate
old rules and still achieve the objectives? :-) :-)

> 5) Information and Physics
> An outspoken critic may say this. Information has its
> own business, and physics has its own, too. There may be
> no need to connect the two if you are busy and cannot spare
> your time for doing something else. A second critic may say
> information is much deeper than physics as swaying John
> Wheeler's banner printing "It From Bits". A third critic may
> jump in and remind us of Carl Friedrich von Weizsacker's motto
> on any production process that is certainly physical, roughly
> saying "Information is that which produces information". Are
> all of them sane?
>

A colleague defined physics to me as the art of finding the simple in
the complex. Is physics an extreme form of pragmatism? Is it most
successful when simple linear representations (and occasionally
quadratic forms) yield useful results? Is information theory useful
when one can approximate natural systems with simple distributions, such
as binary choices?
The sentence

 "Information has its own business, and physics has its own, too."

deserves careful inspection from a philosophical perspective. Are the
terms "information" and "physics" more than semantic sub-classes of our
historical efforts to communicate with one another? Is either term
representable in terms of either observations or symbols in the same
sense that we represent the number two with "2" or the atom of sulfur
with the symbol S? Does either of these terms admit closure? In other
words, can we place meaningful boundaries around these terms such that
we can form definitive distinctions between them?

I have sought to raise more questions in order to promote a lively
discussion. Hopefully, I have stimulated further questions which will
help me understand these questions better.

Sincerely
Jerry

Jerry LR Chandler
Krasnow Institute for Advanced Study
George Mason University
Received on Sun Jan 11 04:02:30 1998

This archive was generated by hypermail 2.1.8 : Mon 07 Mar 2005 - 10:24:45 CET