[From Bill Powers (921008.1800)]
Greg Williams (921008) --
But you admit that the environment can have great effect on what
you consider "surface" phenomena.
Yes. Surface phenomena are those like pushing back against the effect
of a disturbance to prevent a controlled variable from changing, or
varying an action so as to keep altered properties of the environment
from disrupting control. When these things are done in such a way as
to keep all an organism's controlled variables still under control,
there are no important effects on the organism. All the important
effects arise when external influences prevent control from being
successful, so controlled variables are no longer under control.
Disturbances can have pronounced effects on the way a person acts. But
if those actions are successful, and don't prevent other control
actions from being successful, such disturbances are insignificant in
the life of the organism.
It appears that you explicitly agree with me that both radical
organismism and radical environmentalism are wrong, because the (to
you) "significant" variables don't depend on the environment and >the
(to you) "insignificant" variables do depend on the >environment.
This is a misstatement of my position. My position is that external
influences are significant to an organism if they prevent it from
controlling its controlled variables. An external influence that does
not alter the organism's ability to control any of its controlled
variables is insignificant to the organism. It causes no change in
organization and it causes no significant error, hierarchical or
Of course, you seem to believe that organismism in some sense is
vindicated if the (to you) "significant" variables do not depend on
I say that the organism is controlling successfully if its controlled
variables don't depend on external influences. A variable is
controlled if it matches its reference level within reasonable
tolerances (whether that reference level be changing or constant).
What's to "vindicate?" That's just how control systems work, according
to PCT. The significant variables (to me) are the controlled variables
(significant for one reason) or external variables capable of
preventing control from succeeding (significant for quite a different
I still think they do [depend on the environment], in important
ways, but I don't need them to for what I've been calling "co-
Are you saying that controlled variables do depend on the environment
in important ways? What ways, when control is successful? Note that if
a high-level disturbance is countered by a change in a lower-level
reference signal, the lower-level perception still tracks the lower-
level reference signal (and so remains undisturbed), while the higher-
level perception also remains undisturbed (where "undisturbed" means
close enough to the reference signal to satisfy all the organisms's
purposes and needs). When the hierarchy is operating properly, all
perceptions at all levels remain close to their respective reference
signals: they remain under control.
This leads to the need to consider why anyone would or would not
decide to be interested in having explanations of or, more
generally, in understanding and dealing with, how someone's (to
Bill) "insignificant" variables can be influenced by that person's
environment (both living and non-living parts). Maybe Clark McPhail
could help begin to answer this question. Or maybe victims of con
artists could help. Or maybe welfare mothers. Or TV watchers. Or
drug addicts. Or parents. Or children. Or teachers. Or students. Or
counselors. Or prisoners. Or most anyone engaged in and/or studying
Nobody in these categories needs any help in dealing with the world as
long as all perceptions are successfully controlled at their reference
levels, and all critical variables remain near their reference states.
Effects of external events become "significant" only when they
frustrate or disrupt or prevent control.
They may also be significant when control has already failed or has
never been learned, and external events make control possible again or
for the first time. They are never significant when successful control
When you mention these categories, the implication I get is that there
is some problem associated with each category. If the welfare mother
or the TV watcher is controlling all the variables that the person is
organized to control at all the levels that exist in that person, and
there is no critical error, the only problem is in the mind of the
observer who sees something wrong with what the person is controlling
for and would like that person to control for something else. Any
attempt to achieve this wish will, of course, result in conflict.
If there is a problem, it is a control problem. There is inner
conflict, or interpersonal conflict. Some control system is badly
organized and is allowing errors that are large enough to call for
reorganization. Some control skill needed to limit critical error has
not been learned. The control problem could be in the TV watcher or in
the psychologist who thinks that watching TV is bad for the person. Or
both. If neither, there is no significant problem.
I hope that this clears up what "significant (to Bill)" means.
The resulting data might help to convince some PCTers that the
slogan "No one can control you" -- meaning that no one make you
want what you don't want -- merits a "Big deal! I can still have
plenty of problems -- and plenty of benefits -- due in part to
others controlling their OWN perceptions which depend on what I do.
And those problems/benefits aren't insignificant TO ME!"
Problems are problems only if you want to fix them and can't. Benefits
are problems only if you can't get them when you want them. If you
have plenty of problems, they are control problems. They will be
solved by solving the control problems. The mechanisms for arriving at
such solutions are inside each organism, not in the environment
between them. Even the mechanism for understanding, accepting, and
putting into practice a solution communicated by someone else lies
inside the recipient. This is the basis on which I would use control
theory to approach a person with such problems -- not by looking for
ways to change the person's environment (although that is not ruled
out) but by helping to eliminate conflict and enhance the person's
ability to regain control. This would apply to everyone involved in
I would NOT explain to the person that the problems are co-determined
by the person and by the environment, even though that does pretty
much cover the possibilities. Even if the problem is partly determined
by the environment, the person is going to have to get that aspect of
the environment under control, as perceived, of course, in order to do
anything about it.