Allegory, Recognizing control

[From Rick Marken (960409.2215)]

Bruce Gregory (960409.1145 EDT) --

Re: Statistics -- An allegory

I guess I don't get the point of the allegory. Aren't both social
scientists missing Phil Runkel's main point: that the study of group
characteristics tells you nothing about the nature of any individual
in the group?


The rubber band demo just lets us show all the components of controlling
very clearly -- components that are not always visible in everyday

Stefan Balke (9604009.1730 CET) --

I imagine as continuously as possible a figure ranged from zero to a few
thousands representing the amount of error (stress) I perceive in the moment
right helps me to make visible at least to me the amount of the
actual error... if there are circumstances like time-pressure, screaming
children, unpleasurable people standing in front of me in the queue, I
perceive figures up to 500

This is an attempt to make visible the error, the goal, the perceived
variable, the action which I choose to counteract the disturbance and the
changes in the amount of error. So I perfectly see, whether the action


Just try it and tell me whether it makes sense to you and whether it works!

I think this makes great sense because error is a sure fire sign of control --
or, at least, attempted control. But I think that, once error is noticed --
regardless of its magnitude -- the important (and more difficult) thing to
try to notice is the reason for the error: what perception you are trying
to control and what its reference state is. For example, the "circumstances"
that you mention as causes for a perception of large error are really just
states of perceptual variables; the error results from the fact that these
variables are not in their reference states. "Time-pressure", for example,
probably refers to a perception of the relationship between the current
time and the time at which some event starts; the perception of this
temporal relationship creates error if you have a reference for a
particular value of this relationship -- viz., that the current time be
well before the time the event starts-- so that you don't perceive this
relationship in the state "late" or even "too early". "Screaming
children" is just a perception of the loudness level produced by the
kids; "screaming" creates an error _only_ if you have a reference for
loudness that is more like "quietly cooing".

When error is large it is probably very difficult to concentrate on the
fact that you are controlling; in that case you don't really want to
spend time identifying the perceptual variable that you are trying to
control and the reference, in yourself, for the state of that variable.
You just want to make things better, fast -- get to the appointment on
time or get the kids to shut up. When there is a lot of error it can
sure seem like the perception itself is causing the error.

Similarly, when there is no error at all it is difficult to even notice
that you are controlling; you get to the appointment on time, the kids
smile happily when you put them to bed and you don't have to wait in
the queue. When there is virtually no error while you control (as is
usually the case) everything you want just happens, effortlessly.

I think the best time to notice control is when there is some intermediate
level of error; when things are neither effortless nor painful; when the
level of error on the "Balke scale":wink: is about 50. Then you can take a
moment to try to notice the controlled perceptual _variable_, your own
reference for that variable, what you are doing to try to bring that
variable to the reference state and (possibly) what it is about the outside
world that "pushes" that variable from the state you prefer (what might be
the disturbing variables).