I would like to thank Terry for his detailed analysis of my comments on his
work. I should repeat that I consider his theory as a necessary part of any
emerging theory of information and going beyond Shannon. I also commend him for
indicating where it is 'incomplete' (sic), subject to differences of opinion as
to what may be relevant from other approaches which have not been explicitly
discussed in his paper.
One interesting place to start might be the following statement by Terry: Only
the linkage between them (JEB: the molecular phenomena of the model) that
constitutes autogenesis lacks a known empirical exemplar. It is an empirical
question whether this can occur, and what conditions and types of molecules
this would require. I see no physico-chemical reason to doubt this
According to my view of real molecules as instantiating both actual and
potential properties, the linkage between them does also. If this picture is
correct, we have a correct way of looking at the phenomena themselves. We can
then accept the value of the model, which does not violate the principle but
ignores it, but not forget this additional principle when returning to reality.
My view is, admittedly, dependent on acceptance of the reality of quantum
entities and their most complex (non-Boolean) properties as the foundation of
the dualisms at higher levels of reality. However, I believe I am not alone
here. I therefore look forward to further discussions of Terry's approach to
information in which the additional physics and its dynamic logic might be
explicitly taken into account.
Many thanks again,
De : dea...@berkeley.edu
Date : 31/01/2015 - 00:10 (PST)
À : joe.bren...@bluewin.ch
Cc : email@example.com
Objet : Re: [Fis] Fwd: Re: Concluding the Lecture?
Indeed there is much more to discuss than I could include in this
already too long discussion paper. The related absence issues are of
course critical to my thinking. I value your continued feedback on
these issues as well.
I think you do a quite adequate job of restating the autogenesis
hypothesis in your first paragraph. I also agree with your comment
about the model of autogenesis being incomplete because it does not
specify the necessary stereochemical properties of the interacting
molecules, or for that matter the energy flux that is required to
drive reciprocal catalysis, the shapes and charges of molecules that
tend to self assemble into containers (like viral capsids), the
rate-coupling required for reciprocal catalysis and self-assembly to
be reciprocally supportive, and the entropy production of the whole
process, etc., etc. Yes, much simulation and lab work lies ahead.
I actually don't see a problem there, however, nor do I think this
results in circularity. Nothing at the molecular level smuggles in
properties that define information in the model. All that matters for
my purpose is that I am not postulating any unrealistic atomic and
When Ludwig Boltzmann used an idealized thought experiment for
formulate his atomistic account of the 2nd law of thermodynamics with
particles that didn't even interact, it was sufficient to model the
general logic of entropy increase. No real atoms, no real physics,
just the logic of time and random change in position. The model
captured what was minimally necessary and no more. Yes, Gibbs and
others fine-tuned the account, adding the role of free-energy and many
dimensions of interactions, but Boltzmann's thought experiment laid
the foundation. So I don't consider the abstraction involved in the
autogenesis model to be an intrinsic fatal flaw. The question is
whether or not it is too simple, or whether it violates some basic
physico-chemical principles. I can't see how you can doubt that it is
a realistic model, since both component processes are well-studied
molecular phenomena with innumerable exemplars available. Only the
linkage between them that constitutes autogenesis lacks a known
empirical exemplar. It is an empirical question whether this can
occur, and what conditions and types of molecules this would require.
I see no physico-chemical reason to doubt this possibility.
Your question about qualitative signification and my concept of
work saving seemed to lead inexplicably into a comment about human
and social history. Lost me there. But you also seemed to suggest
that the autogenic model provided no fixed ground for making a
qualitative assessment (significance). I believe that it does.
In the autogenic model this depends on there being a fixed amount of
chemical work required to reconstitute an autogenic complex from a
specific state of disaggregation. This differential can be assigned a
finite repeatable value (again not specifying specific molecules).
This functionally defined threshold provides the reference value that
I argue is required for assessing the significance of information