[Martin Taylor 2010.09.01.10.43]
(edited 2010.09.06.15.06)
[From Rick Marken (2010.08.31.0830)]
Martin Taylor
(2010.08.30.23.05)–
However, that's not the
point, is it, as I gather from your smiley. The point is
that the need for PCT can be derived from the basic laws of
physics with no need for any model fitting.
The point (of the smiley face) was that I disagree with that
emphatically.
I sketched that
derivation in my Editorial for the PCT Special Issue of
IJHCS.
Could you give a quick (1 _short_ paragraph) recap of that?
I’d be especially interested in knowing whether your
derivation shows that S-R theory cannot be derived from the
basic laws of physics and is, thus, not needed.
Since your reasoning for "emphatically" disagreeing with the
possibility of deriving the need for PCT from basic physical laws is
simply that you don’t like the idea, I guess I don’t have to alter
the argument.
Here's a (not very) short paragraph, which I follow with a point by
point description. Neither is completely rigorous, but the paragraph
version is more casual, the point-by-point version more technical.
The point-by-point version may be more intelligible after you have
read the short paragraph. Or maybe the reverse. Both should be
treated as abstracts rather than as complete arguments.
···
If the structure of an entity is to be self-sustaining (i.e. it is a
live organism) it must have a way of repairing damage to its
essential structure. In other words, it must have some way of
detecting and acting to correct deviations of its critical
(“intrinsic” in PCT) variables from their optimum states. The act of
correction implies the export of entropy to the environment which
requires an energy flow through the organism, which, in turn, means
the organism must have a means to select a low-entropy source of
energy and to output high-entropy energy. The organism must have a
way of detecting and opposing environmental influences that might
damage the structure of intrinsic variables, which implies
low-energy detection systems and high-energy action systems that act
on the environment (the perceptual control part of PCT). Intrinsic
variables may change in the absence of direct environmental
interaction – the entropy of an isolated system tends to increase
– so the same kind of perceptual control system that mitigates the
influence of the environment on the intrinsic variable structure can
be used to act on the environment to influence the intrinsic
variables toward their optimum values (food foraging activity is a
prime example, in that effective control of the perceptions involved
in food gathering results in the import of low-entropy energy that
is a necessary component of any restorative process). Reorganization
over evolutionary time or within an individual lifetime makes it
more likely that the environmental perceptions being controlled will
serve to maintain the intrinsic variables near their optimum values
than that they will not.
Point-by-point version:
Preamble
1. Any structured system is in a non-equilibrium state. If left to
itself or if in contact with an external environment the structure
will increase its entropy (decay toward equilibrium). The structure
of a living system must be maintained.
2. The maintenance of structure despite ongoing interactions with
the environment is a characteristic of life, so I will call the
structured entity an organism,
3. To maintain structure implies that there exists some internal
record of what the structure should be, and a means of correcting
the difference between what is and what should be. In PCT language,
this is the control of intrinsic variables.
Control of intrinsic variables
4a. One of the requirements for maintaining structure is the
acquisition from the environment of low-entropy energy (e.g. food)
and the disposal of high-entropy energy to the environment. Failing
this, the organism’s structure of intrinsic variables will decay.
4b. To acquire low-entropy energy requires either that the organism
be located so that it can obtain the energy passively and dispose of
high-entropy energy passively (like a plant), or that it act to get
the food and to dispose of the waste (like an animal). If passive,
control of the intrinsic variables could be entirely internal to the
organism (it would be perceptual control, but it would not involve
perception of or action upon the external environment). If active,
the organism must detect some property of the environment in order
to bring it to a state that facilitates the import of low-entropy
energy.
4c. To detect a property of the environment is, in PCT language, to
generate a perception. To bring that perception to a particular
state is to control the perception by acting on the environment.
Control of perceptions of the environment
5a. The intrinsic variable structure gains entropy from its
interactions with the environment. Any mechanism that reduces total
energy input from random interactions will ease the requirements for
restorative maintenance action.
5b. Two ways to reduce the effect of environmental interaction are
passive shielding and active countering. Active countering demands
that something about the environment be detected (in PCT-language,
that a perception be created) so that action to counter the
influence and maintain the appropriate state of the perception can
be performed (in other words, so that the perception can be
controlled).
Biased selection of environmental perceptions to be controlled
6a. There are too many possible perceptions for them all to be
controlled simultaneously. Thermodynamically, the ones that matter
are those whose control serves to maintain the intrinsic variable
structure. The expenditure of energy on others serves to increase
the entropy of the organism, which implies the need for further
entropy export by increasing the intake of low-entropy energy and
the output of high-entropy energy (in everyday language, inefficient
work requires more food and gets you hot).
6b. Therefore, some mechanism must exist so that perceptions for
which the control actions tend to enhance the maintenance of
intrinsic variables are more likely to be controlled than
perceptions for which the control actions are irrelevant to
maintenance of intrinsic variables or tend to damage the intrinsic
variable structure. (In HPCT, this mechanism is “e-coli
reorganization”, but whether e-coli reorganization is necessary or
not, some such mechanism must exist).
Summary: Thermodynamic reasons require that there be perceptual
control of intrinsic variables, control of perceptions of
environmental states, and a way to make it more likely that
controlled perceptions of the environment enhance the maintenance of
the intrinsic variable structure than that perceptions are
controlled that are neutral with respect to intrinsic variable
maintenance or for which the control actions influence the intrinsic
variable structure detrimentally. Succinctly, what is required is
PCT with reorganization based on the reduction of error in the set
of intrinsic variables. Nothing in the analysis distinguishes
whether the required reorganization happens over evolutionary time
or within an individual.
-----------
You ask whether the "derivation shows that S-R theory cannot be
derived from the basic laws of physics and is, thus, not needed."
“Thus” is inappropriately used. “Not needed” doesn’t follow from
“cannot be derived from”. Lots of things are true but can’t be
derived from the laws of physics. For example, why is Glendale
situated where it is, and why does it have that bilingually
redundant name? I guess you might respond that Glendale is not
needed, and I couldn’t argue with you, since I know nothing about
Glendale except its name. However, I would not conclude that
Glendale’s being not needed could be derived from the inability of
the laws of physics to predict its location and its strange name.
I don't think there is anything in the derivation that suggests the
impossibility of situations in which a stimulus automatically evokes
a response and there is no corresponding feedback pathway. I think
that the derivation suggests (but does not prove) that if such
situations exist, they must be rare, and they must exist alongside
perceptual control structures that maintain the intrinsic variables
near their optimum values. Possibly the same kind of argument might
be extended to suggest that any such built-in S-R linkage would more
probably be detrimental than helpful if the environmental conditions
changed, in which case S-R linkages would tend to be eliminated over
evolutionary time (and perhaps within the lifetime of an individual,
though that would be harder to argue). Anyway, it’s not something I
thought about until you asked.
Martin