Yes. Football games are purposefully produced interpersonal
conflicts; diets are purposefully produced intrapersonal
conflicts.
OK, those seem obvious. Any ones you can think of that may not be as
apparent?
What about inner personal conflicts or decisions?. If we think of decision
making as conflict resolution then new ways of looking at "decisions" might
be in order.
People often create a conflict that prevents them from controlling some
variable that they "should" control, or from attending to some other
conflict (internal or external) that they "should" resolve. The
justificatory conflict (a.k.a. excuse) can be either internal or external.
The source of the "should" can likewise be internal as well as external.
A person may create a conflict as a distraction of attention. This often
happens interpersonally, the class clown for example. (Conflict can be
funny.) It can happen internally as well--could be the same class clown. Or
a person may provoke a quarrel, for example, to avoid attending to some
other conflict.
A person may create a conflict as a distraction of attention.
This often happens interpersonally, the class clown for example.
Bruce, are you sure you understand the difference between
a "conflict" and a "disturbance" in PCT?
Here's my understanding. A conflict is when two control systems are both
controlling the same variable with different reference values for that
variable. A disturbance is any influence that causes a controlled variable
to deviate from the reference value at which it is controlled, to the
extent that the deviation is not countered by the control system that is
controlling that variable.
In the case of intrapersonal conflict, signals from two output functions
are combined into a single reference signal. Disturbances are quite easy to
distinguish from intrapersonal conflict. However, in the case of
interpersonal conflict, control actions by each control system are
disturbances to the other's control. The only way to determine whether such
a disturbance is a conflict is by determining whether the disturbance is
intentional or a side effect.
A disturbance that is due to an unintended side effect of control cannot
properly be called intrapersonal conflict. However, people do sometimes
control variables that are disturbances to others while unaware of doing so
(or without letting themselves be aware of doing so). And people
experiencing disturbances that are due to unintended side effects of
another's control do sometimes perceive them as intended effects, and much
more frequently they do expect that the other should be controlling for
non-interference with them. And when one does object to disturbance due to
another, the two frequently find themselves in actual conflict over some
variable which at least one of them had not previously been (consciously)
controlling.
But I'm out of the business of controlling other people's
understanding.
Bruce Gregory (991201.1034 EST)
I take it then we should no longer rely on anything you say?
You should never rely on what _anyone_ says. You should
test things out for yourself. That's why I put all those
demos up at my website.
What role _are_ your posts intended to fill in this case?
The intent of my posts is to teach what I know about human
nature based on my own experience with modeling and testing
PCT. My hope is that these posts will encourage people to test
out these ideas for themselves (rather than just play verbal
tricks with them). My hope is rarely realized but when it is,
it can be spectacular (Richard Kennaway comes to mind, though
the real credit there goes to Bill Powers' Little Man model
and, of course, to Richard himself!).
Best
Rick
···
--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates mailto: rmarken@earthlink.net http://home.earthlink.net/~rmarken
Here's my understanding. A conflict is when two control systems are both
controlling the same variable with different reference values for that
variable. A disturbance is any influence that causes a controlled variable
to deviate from the reference value at which it is controlled, to the
extent that the deviation is not countered by the control system that is
controlling that variable.
That is not my understanding. A disturbance is completely compensated for
when control is successful.
Here's my understanding [of conflict and disturbance]...
Check.
Really?
Bruce Nevin (991130.2306 EST)
Here's my understanding. A conflict is when two control systems are both
controlling the same variable with different reference values for that
variable. A disturbance is any influence that causes a controlled variable
to deviate from the reference value at which it is controlled, to the
extent that the deviation is not countered by the control system that is
controlling that variable.
My understanding of "disturbance" is that it is any influence that *would*
cause a controlled variable to deviate from the reference value at which it
is controlled, if the controller were not present. The Test involves
applying disturbances and seeing whether such a deviation actually happens.
The success or otherwise of the controller in resisting the disturbance
isn't relevant to whether something is a disturbance or not. A disturbance
that the controller perfectly counters and a disturbance that overwhelms
the controller are both disturbances.
-- Richard Kennaway, jrk@sys.uea.ac.uk, http://www.sys.uea.ac.uk/~jrk/
School of Information Systems, Univ. of East Anglia, Norwich, U.K.
Here's my understanding. A conflict is when two control systems are both
controlling the same variable with different reference values for that
variable. A disturbance is any influence that causes a controlled variable
to deviate from the reference value at which it is controlled, to the
extent that the deviation is not countered by the control system that is
controlling that variable.
That is not my understanding. A disturbance is completely compensated for
when control is successful.
Depends on gain.
Gain is also relevant to the determination whether there is conflict or
not. There is the theoretical possibility of two control systems that
control the same variable with different references controlling with
lowered gain such that each has "good enough" control of the compromised
value of the CV.
All of this applies directly to questions about social relationships like
coercion, cooperation, agreements, and so on.
Bruce Nevin
···
At 07:12 AM 12/01/1999 -0500, Bruce Gregory wrote:
At 12:41 PM 12/01/1999 +0000, Richard Kennaway wrote:
Bruce Nevin (991130.2306 EST)
A disturbance is any influence that causes a controlled variable
to deviate from the reference value at which it is controlled, to the
extent that the deviation is not countered by the control system that is
controlling that variable.
My understanding of "disturbance" is that it is any influence that *would*
cause a controlled variable to deviate from the reference value at which it
is controlled, if the controller were not present. The Test involves
applying disturbances and seeing whether such a deviation actually happens.
The success or otherwise of the controller in resisting the disturbance
isn't relevant to whether something is a disturbance or not. A disturbance
that the controller perfectly counters and a disturbance that overwhelms
the controller are both disturbances.
My understanding is that control is never absolutely perfect. If I am
wrong, then yes, my statement above does not include the case of perfect
control. It does include the case of overwhelmed control.
I have just one quibble with the belopw, and that concerns the references to
awareness in the last paragraph. Since awareness is not well integrated
into PCT (or at least into BT) at this time, there is no theoretical
justification for including awareness as a factor. I suppose the easiest
way to accomodate 'not letting themselves be awre" would be to assume some
inputs are in imagination mode. Bruce seems to be hinting that some higher
level control systems might use the imagination connection as a means of
controlling error. I wonder if something like this could be set up in
Rick's spreadsheet model?
···
On Tue, Nov 30, 1999 at 11:06:33PM -0500, Bruce Nevin wrote:
[From Bruce Nevin (991130.2306 EST)]
Rick Marken (991130.1740) --
At 05:38 PM 11/30/1999 -0800, Rick Marken wrote:
>
>
>Bruce Nevin (991130.1852 EST)
>
>> A person may create a conflict as a distraction of attention.
>> This often happens interpersonally, the class clown for example.
>
>Bruce, are you sure you understand the difference between
>a "conflict" and a "disturbance" in PCT?
Here's my understanding. A conflict is when two control systems are both
controlling the same variable with different reference values for that
variable. A disturbance is any influence that causes a controlled variable
to deviate from the reference value at which it is controlled, to the
extent that the deviation is not countered by the control system that is
controlling that variable.
In the case of intrapersonal conflict, signals from two output functions
are combined into a single reference signal. Disturbances are quite easy to
distinguish from intrapersonal conflict. However, in the case of
interpersonal conflict, control actions by each control system are
disturbances to the other's control. The only way to determine whether such
a disturbance is a conflict is by determining whether the disturbance is
intentional or a side effect.
A disturbance that is due to an unintended side effect of control cannot
properly be called intrapersonal conflict. However, people do sometimes
control variables that are disturbances to others while unaware of doing so
(or without letting themselves be aware of doing so). And people
experiencing disturbances that are due to unintended side effects of
another's control do sometimes perceive them as intended effects, and much
more frequently they do expect that the other should be controlling for
non-interference with them. And when one does object to disturbance due to
another, the two frequently find themselves in actual conflict over some
variable which at least one of them had not previously been (consciously)
controlling.
My understanding is that control is never absolutely perfect. If I am
wrong, then yes, my statement above does not include the case of perfect
control. It does include the case of overwhelmed control.
That is, control in the face of disturbances is never absolutely perfect.
If qi=r and d=0 with no variation, then control is perfect with qo=0. That
is not relevant since we are talking about disturbances, but the quibble
probably needs to be covered.
That is not my understanding. A disturbance is completely compensated for
when control is successful.
That is only in the case of a system with a pure integrating output
function. If the integrator is leaky, or if the system has a proportional
output function, the error will be approximately the effect of the error on
the controlled variable without feedback control divided by the
steady-state loop gain of the control system.
Also, you must remember that all control systems have limits of output. If
a disturbance is large enough to drive the output to a limit, any further
increase in the disturbance will show up directly as error.
My understanding of "disturbance" is that it is any influence that *would*
cause a controlled variable to deviate from the reference value at which it
is controlled, if the controller were not present. The Test involves
applying disturbances and seeing whether such a deviation actually happens.
The success or otherwise of the controller in resisting the disturbance
isn't relevant to whether something is a disturbance or not. A disturbance
that the controller perfectly counters and a disturbance that overwhelms
the controller are both disturbances.
The second paragraph is true. I have a problem with the first.
It's too easy to think of control either existing or not existing.
Actually, it exists on a continuous scale from none to perfect. When a
disturbance is applied to the variable being controlled by any real control
system, it has some effect on the controlled variable. It must, if there is
to be any change in the output action. But this effect can range from a
large fraction of the value of the variable to such a small effect that you
would need a microscope to see it.
We should not speak in terms of whether control exists or not. We should
speak in terms of loop gain because loop gain determines the degree of
control. The greater the loop gain, the less effect a given disturbance
will have on the controlled variable. If you want to say there is no
control, just say that the loop gain is zero. That says it all. But if the
loop gain is 2, so the effect of the disturbance is reduced to 1/3 of its
uncontrolled value, would you say there is "no" control? And if the loop
gain is 200, so the effect of the disturbance is reduced to 1/201 of the
uncontrolled effect, would you say control is perfect?
It's best to speak in quantitative rather than categorical terms about
control.
Richard, _you_ knew that.
That is not my understanding. A disturbance is completely compensated for
when control is successful.
Bruce N:
Depends on gain.
Gain is also relevant to the determination whether there is conflict or
not. There is the theoretical possibility of two control systems that
control the same variable with different references controlling with
lowered gain such that each has "good enough" control of the compromised
value of the CV.
I am impressed, Bruce N. I needn't have said a thing.