[From Bill Powers (920624.1200)]
Martin Taylor (920623.2100) --
One problem with the study is that the "adapting" VOT for the >(synthetic)
/r@pi/ was 50 msec, and the natural (after /i/) VOT for the >subjects' /p/
averaged 57 msec, 7 of the 20 subjects being shorter than >50 msec. So
there wasn't much perceptual adaptation to be expected.
This is a coffee ground, not a peanut. If the authors reported that
if the subject had listened to /r@pi/ rather than /i/, the voice
onset time was shortened, particularly for /t/.
then they concealed the truth. This finding goes into the literature as a
conclusion about what "people" do, and others will pick up this "fact" and
use it as if it's true of everyone, and waste the time of Rick Marken,
Bruce, you, and everyone else who feels the urge to explain phenomena. It
isn't a phenomenon. Any explanation that purports to show why this
"perceptual adaptation" occurs in "people" also has to explain why, in 35
percent of the cases, the OPPOSITE "adaptation" was observed.
I don't think that anyone exploring behavior using the HPCT model has any
obligation to explain facts that aren't even facts. I think that from now
on we should qualify all phenomena put up for explanation as to their
status: are they false of any large fraction of the people in the study, or
are they real robust phenomena that can reliably be demonstrated in
essentially every subject?
Thank you for bringing this up. This is going to save me a lot of time.
From now on, the first question I'm going to ask (if I remember to do so --
if not, remind me) is whether the phenomenon to be explained is actually a
phenomenon. If it's not, I'm going to turn my attention to more profitable
questions. From here on down, it's all peanuts.
···
----------------------------------------------------------------------
Martin Taylor (920623.2115) --
As soon as you allow same-level loops, you are into the realm in which
oscillations and possible chaos lurk.
I concur. I concluded some time ago that there really aren't any
"configurations of configurations" and so on. That's just a device of
speech, one of the penalties for having to express parallel processes in
sequential form.
I think we perceive legs, arms, seat, back and chair all in parallel. We
can perceive any of these without the others. The Gestalt psychologists
knew this 60 or 70 years ago. In truly hierarchically-related perceptions,
there's NO WAY to perceive a higher-level thing without presence of the
lower-level components, save in imagination. If there are no sensations,
there are no configurations. This is an absolute requirement in the concept
of hierarchical perception.
----------------------------------------------------------------------
Rick Marken (920623.1830) --
Bruce Nevin asked >Is adaptation a disturbance?
You replied:
I call it a disturbance because if affects the perception of the
controlled variable in such a way that action is required to maintain
that variable at a reference level.
This is an important point. We tend to think of disturbances only as
external forces that create error. But it isn't necessary for them to
create error for the control system to oppose their effects. If you're
holding a weight up at arm's length against gravity, you're controlling a
perception of position. Now if someone else comes along and pushes upward
on the weight, the error that's maintaining the muscle forces will
decrease, and the muscle forces will also necessarily decrease. This
decrease in your upward force opposes the upward force that the other
person is adding. So the disturbance is opposed EVEN THOUGH THE OTHER
PERSON IS HELPING YOU HOLD THE WEIGHT UP.
"Adaptation," which I agree is not a very useful term, occurs when the way
a system acts changes. If you hold a dumbbell up at arm's length every day
for half an hour, your arm muscles will "adapt" by getting stronger -- that
is, more muscle fibers will be added. While you were weak, getting the
muscles that existed to hold the weight up required producing a large
driving signal from the spinal motor neurons. If the same driving signals
were produced after the adaptation, the weight would not only be held up,
it would be tossed backward over your shoulder. To keep the same "holding-
up" occurring as before, your peripheral control systems reduce the output
from the motor neurons; the same output now produces more force, so less
output is needed to produce the force needed to counteract the weight of
the same dumbbell, which is always the same force. Therefore the control
system has decreased its neural output to compensate for the disturbance
caused by the increase in signal-to-force transduction in the muscle.
A lot of what's called "adaptation" is just control. And a lot isn't. So
the word isn't really of much use in HPCT.
Thanks, Rick, for anticipating my complaint about junk data -- I wrote my
comment before getting to yours.
-----------------------------------------------------------------------
Hank Folson (920623) --
1. How will a person (living control system) control when s/he does not
know that s/he is a living control system, and does not know that
everyone else is an independent control system that cannot be >controlled
without force?
Exactly the way he or she does now. People will try to make other people
act to suit the actor's reference levels. The others will resist. The
actors will try harder. The others will resist harder. Before long there
will be blood all over everything.
4. How will a person (living control system) control when s/he knows
that s/he is a living control system, and knows that everyone else is >an
independent control system that cannot be controlled without force >in a
control theory _aware_ world?
The futility of (or the penalty for) trying to control other control
systems of equal capability and intelligence will be obvious, and people
will stop trying to do it, except when they have to (to survive, for
example).
Wouldn't the controlling actions be different?
Not so much the actions as the goals.
1A. (A small but important sub-category) How will a person (living
control system) control when s/he does not know that s/he is a livingcontrol system, and does not know that everyone else is an independent
control system that cannot be controlled without force, but has through
their life experience developed a Systems Concept and controlling
techniques that are compatible with PCT?
All methods of control are compatible with PCT. PCT doesn't teach methods
of control; it explains the methods that already exist, and have always
existed. There is no "technique" of control other than comparing your
perceptions with your reference signals and converting the difference into
action (via lower-order goals), as best you can. Some people have learned
how to fool others into submitting to control, or taking advantage of their
ignorance or beliefs to the same end. Control theory doesn't teach anyone
how to do that. It explains what they're doing. It also explains why trying
to control other people creates all the problems it creates, even when it
seems to succeed.
2. How will a person who has lived most of their life and had most of
their experience in life in a Category 1 world as a Category 1 person,
but then founded/pioneered/studied/was seduced by PCT, attempt to
understand and control in our Category 1 world? Will s/he only use the
new paradigm, or will s/he drift back and forth between paradigms?
We could take a poll. I drift back and forth between paradigms, although as
the years have gone by (more for me than most others), I've spent less time
in the system concepts I was raised with and more in a new one. I think
that when it comes to the crunch, I now opt for the new one almost all of
the time. This will be much easier for people who don't have to overcome
the old ways to appreciate the new.
3. How will a person who has been raised in a Category 1 world, and >whose
Systems Concept is rooted in Category 1, and who may even have a >vested
interest in the world remaining a Category 1 world, control >(still in
determined ignorance of their being a control system) when >faced with the
PCT concept of the world?
Stick around and watch. Most of them will probably grow old and die without
ever changing their minds, while continuing to act as perfectly good
control systems without realizing it. They will exert great efforts to
protect their system concepts against disturbance. They will try to attack
the disturbances at the source, or to prevent them from being effective, or
to bring greater forces to bear and overcome the disturbances.
Would you gain any new insights by using these categories were you >reread
and analyse again what is going on in the recent posts on the >newspaper
organization, leadership, Chimps, the Jury comments?
Probably. But don't be shy: tell us what new insights you think we might
get.
RE my quote: don't overlook the part that says "... if we want to find the
significance of the first new concept of human nature since Descartes." I'm
trying to say what we need to do IF that's the goal. For a lot of people,
it still isn't.
---------------------------------------------------------------------
Bruce Nevin (920623.1234 --
An interesting study on the fish. How would you reframe it in PCT terms?
For starters, I'd recommend considering what "aggression" means as a
control phenomenon, and how the relationship of "domination" might work.
There's an interesting parallel here with some vague ideas I've had about
cell specialization during maturation. All cells have all genes,
apparently. But when a certain number of cells with a particular gene
become active, the result, according to the conventional interpretation, is
that additional cells with the same genes are turned off, so we get livers,
fingernails, gonads, and so on.
But suppose that these genes represent control systems controlling for some
aspect of the common milieu. When there are enough control systems for a
given variable, say an enzyme, the controlling systems will experience zero
error, near enough. This means that all the cells with a reference level
for that variable that's less than the amount being maintained will
experience truly zero, or even negative, error and their outputs will turn
completely off. So there's no need for a third party to actively switch
genes on and off: the controlled variable itself will do that.
Suppose that all the male fish have an intrinsic reference signal that
says, in some equivalent form, "What we need around here is a flashy horny
male to keep this society in line." The actual controlled variables would
be more specific things that have this situation as a consequence. One
would suppose, then, that the male who maintains these things at the
highest reference level would satisfy ALL the males' reference levels for
those things, and their control systems would atrophy through disuse. But
when that obliging male dies, suddenly all the males begin to experience
this intrinsic error, and reorganize until one of them manages to satisfy
all of them again. "Reorganize" may not be the word -- this seems too
regular a relationship for reorganization. But you get the idea.
The guy with the highest reference level, and I suppose the highest loop
gain, ends up doing all the work.
---------------------------------------------------------------------
Best to all,
Bill P.