Re: Knowings, Information, and Matter


Dear Aaron,                              December 30, 1997

You quote Stan (Klein):  

>> On the issue of whether the fundamental 'stuff' of our world 
>> is information
>> I think this is just a semantics issue of what one calls it. If one wants
>> to call the underlying stuff information how can one object,

[Aaron (Sloman)]
>I guess it's a little more than semantics if the forms of the equations
>or laws that actually work, have something suggestive of items of
>information or events in which information is acquired, transformed,
>transmitted, stored, etc. I take it that's Henry's claim.

[HPS]
Yes. I first want to distinguish the "stuff" out of which the universe
is made from "classically conceived matter": the latter is characterized
by local-deterministic laws governing local parts that are completely
characterized by a location and a few numbers assigned to that
location. The world as conceive in QT is very different due to the very
different character of the dynamical laws, which involve sudden jumps
in distant places due to events happening here. This non-matter-like
behavior is what forced the inventors of QT to formulate QT in terms
of "our knowledge": our knowledge (or information) about whether or not
the quantum of energy is likely to be found in the faraway region is
suddenly changed when we find it here. But in orthodox thinking one does
not grant that the "stuff" was all here before the measurement was made.

The words do play the important role of characterizing the way the stuff
is believed to behave.

>
>What's more likely is that NONE of our existing familiar concepts
>("information", "knowing", etc.) is really adequate to express the
>semantics of the theory, so we have to grope around for a while until a
>new set of concepts emerges. (Compare the emergence of the distinction
>between inertial mass and weight as gravitational force, or the
>distinctions between kinetic energy, momentum and velocity.)
>
>I take it Henry is trying to accelerate the groping, though some bits of
>his paper leave me very puzzled, for instance I find this VERY hard to
>understand:
>
>   Von Neumann said it all when he showed, back in the thirties, that
>   one could push the boundary between the world described by the
>   quantum formalism and the world described in terms our classical
>   concepts all the way to the boundary between brain and mind without
>   disrupting the predictions of quantum theory, and noted that there
>   is no other natural place to put the boundary, without disrupting the
>   integrity of the theory. In fact, it is, in principle, {\it only} if
>   one pushes the boundary all way to the brain-mind interface that one
>   obtains, strictly, the prediction of orthodox quantum theory:
>
>I can't unpack the metaphor of "boundary" and "pushing' etc. But I'll
>try re-reading.

The orthodox Copenhagen interpretation says that the theory allows
one to compute probabilities associated with the appearance of classically
described outcomes of experiments: e.g.," the pointer on the device lies 
between the numbers 7 and 8 on the dial." or " the light numbered 5 lighted
up". Generally one uses also a classical model of the device to figure
out what distribution (wave function of the quantum object) is being 
detected when the device acts in one of the allowed classically describable
ways.

von Neumann considered a situation in which there was a sequence
of measuring devices, each set up so as to detect the response of the 
preceding one. Then he showed that one could take any one of the detectors
to be the one that defined the boundary between the part of nature that was
described by the classical description and the part that was described by a
quantum wave function. All possible choices gave FAPP (for all practical
purposes) the same predictions for the probabilities of the various 
possible observable outcomes. He imagined pushing this boundary between
the classically described and quantum mechanically described worlds right
into the brain: first the transition element would be the cells in the
retina, then some higher-level system of neurons until finally the entire
brain would be described by a wave function, and the experience would be 
the experience associated with some classically describable aspect of
the brain dynamics. As at every other stage [of this sequence of 
alternative possible (but FAPP empirically indistinguishable) ways of 
making the separation between the classically and quantum mechanically 
described parts of nature] this classically described aspect singles out 
as actual only one branch of the many-branched wave function of the 
quantum mechanically described part of nature. Thus there is a jump or 
mismatch between the QM and CM described parts. But the von Neumann 
analysis shows that FAPP there is no way to detect where along the chain 
the jump occurs from many branches [representing alternative possible 
classically described worlds] to the one classical world that appears 
in our consciousness. 

von N. seems to suggest that the only natural placement of the
"Heisenberger Schnit" is between mind and matter.
Wigner is explicit, and attributes the idea to von N. 

My approach is to accept that human thoughts are coonnected in this way 
to a quantum collapse in human brains, but to extend mind beyond human 
consciousness, so as at least to allow animals to have rudimentary 
experiences. 

Thus experiences of some generalized kind, associated with collapse events 
associated with physical systems, will be a general feature of nature.
The experiential aspect of the event will be dynamically important, because
only certain sufficiently coherent experiences are allowed. This provides
the needed criterion for when an event can occur. (There seems to be no
natural criterion within the purely "physical" description supplied by QT.)

Within this scheme there would be some "coherency condition" on thoughts,
at least in principle. If a continuum of thoughts---and hence corresponding
collapses--- are allowed, then one has trouble formulating the probability
calculus associated with QT. The normal rule is based upon the notion of
setting up a set of  alternative possibilities characterized by
distinguishable classically described experiences. 


[Stan, continued]
>> as long
>> as that information doesn't need a conscious being to observe it.

[Aaron, continues]
>Information needs only something which does one or more of:

>	receive, transmit, store, transform, combine, compare,
>	detect, use (e.g. in selecting an action) etc.
>
>(Note that these are not concepts of classical physics: in classical
>physics they require higher order mechanisms, e.g. switches,
>controllers, cameras, homeostatic devices, recorders, computers, etc.)

>Rudimentary forms of those processes don't require anything like human
>consciousness. A thermostat suffices. Or a bacterium?

There is the interesting question of whether "information"
is defined by a structure itself, or only insofar as there is another
system that can "interpret" it by reacting to it in a characteristic 
way. 

I think of information as more than a mere structure itself:
an interpreting system is needed. Thus a nature made of information
would be a more tightly bound system than one made of classically
described matter.  The latter kind of world although dynamically
complete is only very loosely bound: each tiny part is logically
and ontologically self sufficient. 

I think the main message of quantum
theory is that Nature is not as loosely connected as that: the whole
is more like a web of components defined by their connections to
the whole. This seems to make it more like a structure of information
than an aggregation of classically conceived tiny material elements
behaving in accordance with local-deterministic laws. The problem is
to fill in the details of this more complex structure in a way that 
can be tested.

Henry