[Martin Taylor 931129 12:30]
The following is a lightly edited version of a private posting to Bill
Powers, which he asked me to post to CSG-L. The editing is to remove
references of a private nature.
Martin
···
==================================
Bill,
I can tell you a little about where I was once coming from in respect
of PCT and dynamics.
Firstly, we have to recognize that our other discussion (the "reality" of
CEVs defined by perceptual signals) applies. A student of dynamics
has a set of perceptions of the world that work together very nicely.
The system of dynamics is quite coherent, and perceptions using its
relationships as functions are nicely controllable. You disturb it
a little, and the disturbance is resisted. What is necessary is to add
new functions, not to disturb the old, but to allow them to continue
to operate.
Specifically, intentionality in control is always WITHIN an organism.
Seen from outside, as part of the larger universe, there is none. There
are energy flows and shifts of entropy, and the like (all perceptions
in people who have appropriate functions with which to perceive them).
It doesn't matter WHAT the organisms do to stabilize their own local
environment, because that doesn't affect the overall governing factors
of energy conservation, self-organization in non-equilibrium flows, and
so forth. It is only when you shift viewpoint, and start to look (empathise)
from within the organism that you see control and intention. That is to
say, you start by asserting that it DOES matter what the organism does
to stabilize its local environment.
Think of the whole issue of life as a long-term (>4 x 10^9 year) dynamic
problem. All you can see from this viewpoint is one (and only one) self-
organized structure, which alters very slowly, just as do the individual
vortices in a vortex train. There isn't any intentionality here. Each
vortex can be studied individually, but is "really" only an aspect of the
whole vortex train. Likewise, from this point of view, an individual
organism is only an expression of a long dynamic evolution.
Within any high energy flow (far from equilibrium) system, there are
feedback loops, meaning that events that occur in one place have effects
that can later be felt at the same place. If any of these feedback loops
tend to be negative, that "place" is more stable than are other places
in the flow. The flow can be (and often is) described in terms of attractor
basins separated by repellor surfaces. This nomenclature obscures the
processes in favour of a description of the results of the processes.
Negative feedback produces attractors, positive feedback produces
repellors, and lack of feedback (very rare) produces metastable regions.
Nothing in this suggests purpose.
Purpose arises only when the negative feedback has a changeable reference,
and not just from the simple fact of there being an attractor (there being a
negative feedback process). But if you just look at the attractors
and their basins, you can't tell whether there is a changeable reference.
You have to look at what is sometimes called a "superdynamic," which is
a picture of how the dynamic--the arrangement of attractors and repellors--
changes shape as a function of the value of other parameters. Even then
you only have a picture from outside, and you don't see the processes
that generate the picture.
Intention can't be seen from the dynamic picture of the total system,
because it isn't there. As a whole, the universe is purposeless. The
purpose exists only within specialized small pieces. We call them ECSs,
and we use them as a shorthand to describe a process. The purpose of the
ECS is delivered to it from outside, in the form of the reference signal.
As seen by the dynamicist observer, this reference signal --eventually--
is part of the energy flow of the universe, and represents no purpose.
"Eventually" there may mean 4 billion years ago, since the "purpose" of
gene propagation must go back about that far.
Evolutionary "purpose" sounds like teleology, and is resisted for that
reason. There is no reference level for what the gene should be in the long
term, and hence no "purpose" in the way it changes over evolutionary time.
It can change in multitudinous ways, and those that survive will be those
later seen. All the same, there is a purpose and that purpose is stability,
a.k.a. the approach to a dynamic attractor. Genes that collaborate to
work through structures that control their perceptions are more stable
than genes that don't. So the "purposes" of organisms can be seen, from
the dynamical viewpoint, as being not purposes at all.
I think that the key is always to ask "from what viewpoint are you looking."
If you are looking from within an organisms, the purpose is there. It is
strong. It is to propagate genes. The organism exists for this end alone.
All else is to support it, and as one looks "up" from within the organism,
one sees why one tenses a muscle (a purpose generated from the need to
move the arm), and why one gets a job (a purpose generated from the need
to make money) and so forth. From that view, purpose is everything.
Perception is the way purpose is achieved, but without purpose, perception
would be pointless. From the outside, though, there is still no purpose.
I have a strong feeling that [the dynamicists] are
looking for a particular KIND of explanation. Are we seeing here
a modern version of the rejection of purposive behavior that has
been going on for this whole century, and a good part of the past
one?
I think it may not be so much a search for a kind of explanation, but
a search for a target of explanation. Undoubtedly different kinds of
explanation satisfy different people. You are well aware that what
satisfies you is different from what satisfies me. You like to see
a model work, whereas I see that as showing only that particular
circumstances can be shown to allow the model to work. I like general
principles that apply under circumstances generically like the particular
instance. You find that unsatisfying because critical assumptions or
analytic methods may be hidden or wrong. There isn't a right or wrong
about it. What would be wrong would be to assert that any kind of
explanation is, in itself, wrong. The explanation may be wrong, and
the kind may not fit your own perceptual function set very well, but if
it fits the dynamicist's and it works, then there is no ground for criticism.
They seem to be trying to pre-empt the term "intention"
and give it some other meaning that makes it more consistent with
their beliefs about causation. And of course, in the process,
abandoning its original denotation.
We all use words in ways that suit our personal ways of perceiving the
world. When we develop theories, we coerce the words into straight-jackets.
Other people coerce them into different forms, and we lose the ability
to communicate. Too bad, but not easily prevented. Your "intention"
is quite specifically constrained by the hierarchic structure of PCT.
It sounds as if it has much the same effect as the loose, everyday,
meaning of the term, but when one comes to use it precisely, those
connotations have to be dropped. You couldn't really use "intention"
in the same way if we had only a very complex single-level set of control
systems. But the general public wouldn't have to change.
Martin