[Martin Taylor 2014.10.28.14.50]
[From Dag Forssell (2010.10.28.0950 PDT)]
Bill Powers did exactly that when I drove him back to the
airport in 1991
(See Ten Minutes at my site).
He asked me: "Would you like to see a really tight control
system?" – “Yes”.
So as I drove at 65 MPH feet away from the concrete divider on
the
freeway, Bill pulled on my arm. Needless to say I resisted.
Runkel used this as an example in People as Living Things
(without
attribution).
To me, Bill's pulling was a disturbance, plain and simple.
Position in lane is a controlled variable at one level, position
of
steering wheel a controlled variable at another level.
And shoulder and elbow angles, and muscle tensions in
the related muscles, are a raft of controlled variables at yet
another pair of levels.
If Bill had pulled on Dag's steering wheel, the disturbance would
have been to the perception of wheel angle. Since he pulled on the
arm, it would have been on the joint angles. The effect of changing
the values of joint angles would, of course, have been to change the
steering wheel angle, and thereby to change the perceived position
in lane. Dag controlled the position in lane by controlling the
lower-level variables. Had Bill disturbed the steering wheel
directly, the effect would have been to also disturb Dag’s joint
angles, which would have required changes in the reference values
supplied to the muscle tension control units. So it’s not easy to
disentangle the effects of Bill pulling on Dag’s arm and his pulling
on the steering wheel, since the feedback paths in the real world
are quite tightly coupled.
Either way, the disturbance, like any disturbance, results in less
precise conformation of the perception to the reference than would
be the case in the absence of disturbance, but it does not influence
the ability to control, which is the ability to keep the perception
near its reference value despite the disturbance.
Since control is never absolutely perfect, the effects of low-level
disturbances do propagate up the hierarchy and must be compensated
at each level, but because control is better than no control, the
effects are diminished as one goes up through the levels. So my
answer to Fred’s question: " Is that a disturbance to
the
car-in-the-lane controlled variable or is it a disturbance to some
variable related to my behavior – perhaps “controlling for
unimpeded
ability to steer”?" is that it is a disturbance to Dag’s joint
angle perceptions, and thus a (lesser) disturbance to Dag’s
steering wheel angle perception, and thus to Dag’s “car-in-lane”
perception – but not a big one. It affects the precision of Dag’s
car-in-lane control, but not his abiity to steer (unless Bill
pulls very hard).
Martin
···
Best, Dag
At 09:09 AM 10/28/2014, you wrote:
[From
Fred Nickols
(2010.10.28.1205 EDT)]
I have a question of clarification about PCT as applied to
human
performance.
Disturbances are those other actors and factors that also
affect the
controlled variable. In human performance technology, these
would
fall under “Conditions.” But, in human performance
technology,
“Conditions” also includes factors that affect the behavior of
the
performer and that can impede, hinder or even help the person
to
perform/behave as necessary.
Let’s use the driving the car example. My aim is to stay in my
lane. A side wind might be a disturbance. But what if you, a
passenger, tug on my arm? You are interfering with my ability
to
steer and thus maintain lane position. Is that a disturbance
to the
car-in-the-lane controlled variable or is it a disturbance to
some
variable related to my behavior – perhaps “controlling for
unimpeded
ability to steer”?
I’m a little puzzled. Any help greatly appreciated.
Regards,
Fred Nickols, CPT
Managing Partner
**[ Distance
Consulting LLC](http://www.nickols.us/)
The Knowledge
Workers’ Tool
Room
*“Be sure you measure what you want.”
“Be sure you want what you measure.”
***