[from Mary Powers 9312.10]
Tom (12/9)
Your posts on dynamic systems analysis mavens are killers. My
first reaction was a sick feeling that these people have the DSA
bit in their mouths and there's no possible way to slow them down
as they pursue this latest trendy thing.
But Turvey does leave an opening, in that he does actually in
print state THE intractable problem of psychology - that PCT
solved decades ago: "It is fairly obvious to all of us that any
given coordination will exhibit variations that depend upon the
context...the challenge is to understand the embodiment of this
unvarying task dynamic in variable movement patterns".
An article directed at that single point, called something like
"Is Dynamic Systems Analysis Really Necessary?" might at least
penetrate the skulls of a few people who are only just beginning
to follow the leaders or maybe have some doubts about doing so. I
know there's no percentage in suggesting articles for someone
else to write, but you've got the file full of smoking guns.
Speaking of which, I just got a flyer from MIT/Bradford for a
book by Linda B. Smith and Esther Thelen : A dynamic systems
approach to develoment : applications (1993) and a forthcoming
Dynamic systems approach to the development of cognition and
action. The blurb says: "What do laser lights, crystals, walking,
reaching, and concepts have in common? All are complex dynamic
systems..."
(btw, could you send (lend) me that Current Directions issue with
Nelson's article in it?)
Rick
I got all happy when you mentioned the evolution article and said
you'd give a try at programming it. You give good rant, but you
program better. And in the long (long, long, long) run it's much
more constructive.
Osmo
We can't give you 5 Science and Nature level publications on PCT
because editors and reviewers don't like, understand, or agree
with it. Various people on the net have files full of weird
rejections.
You said >There is still the reference, set or learned by the
system itself or set by an outside observer (like a command given
by a sergeant to a recruit)
This mistake has been around for 50 years. Norbert Wiener said
the same thing. A command from outside is not a reference signal.
It is a perception. IF the recruit has a reference signal for
obeying the sergeant's command, he will do whatever it takes to
obey it, because an unobeyed command is a disturbance - a
discrapancy between a desired state and the perceived state. He
may have the same reference signal as the sergeant, but it is his
own and can't be put in him from the outside. No two soldiers may
have the same reason for obeying - one may want to please the
sergeant because he hopes to be a sergeant himself some day;
another may simply want to avoid punishment for disobedience,
etc. Living control systems differ from artifical ones in that
their reference signals are inborn, or developed, and are not
accessible from outside the system. If you are in management, and
raising three boys, you must know that simply giving orders is
going to lead to all kinds of not-necessarily-desirable results,
depending on the structure of reference signals in the workers or
the boys. They will not obey you unless it satisfies their own
needs - for love, approval, money, promotions, the satisfaction
of achievement, etc., or more coercively, because they are afraid
of what will happen if they don't obey. You can't impose
reference signals - but you can design the environment so that
people can maintain and achieve their reference states by doing
the things you want them to do. You can do this in a way that
enhances them as human beings - helps them fulfill their highest
goals for themselves, or by forcing them to defend their need to
survive (keep your love, hold onto the job), by sacrificing their
integrity in other respects.
There is still purposeful action (behavior) to achieve the
reference, and there is still feedback from the current status of
the desired action.
The feedback is from the current state of the desired perception,
not action. Actions vary in order to produce desired perceived
outcomes.
If the particular actions being used to achieve a desired
perception meet with disapproval, then those actions, as
perceived, can be changed (if the disapproval matters).
Control theory as applied to the life sciences has been around
for half a century, since Cybernetics first came out, and longer,
if one considers the concept of homeostasis. But what is firmly
fixed in the minds of control engineers, biologists, and
psychologists, is the idea that what is being controlled is
output. This is because what matters to an outside observer is
what the plant or the organism is DOING. PCT is concerned with
how that doing gets done, and how it gets done is by controlling
the perception of the actions, not the actions themselves. It is
not output that is controlled, it is input. The input includes
not just what one perceives of one's own actions, but also the
perceived results of those actions, including both the effects on
the physical world and the reactions of other people to what one
is doing.
Mary P.