[From Bill Powers (940904.0840 MDT)]
Rick Marken (940903.2300) --
Just by way of amplying your discussion of means and ends:
It's true that in order to control a variable, the means used to control
it must be free to change, because the variable in general is subject to
disturbances. So no one value of output, whether it be a physical effort
or a choice of principles, can possibly keep the controlled variable (as
seen from the next level up) in a specific state.
We have to remember that the controlled variable is at a _higher_ level
than the action that controls it. This fact gets overlooked because we
all tend to speak of controlled variables as if they were in the
external environment, so they seem to be of a lower level than the
mental actions which control them (just because of the way we draw
diagrams, with the external controlled variable at the bottom). The
output of the control system, up here in the diagram, controls the
controlled variable, down there in the diagram.
But that is confounding the external and internal points of view. From
the internal point of view, the controlled variable doesn't even exist
until the signals from the environment have been perceived, re-
perceived, and re-re-perceived to create a perception of the state of
the controlled variable for comparison with a reference signal entering
the control system from above. The output that affects the controlled
variable leaves the control system headed downward, and its effects
eventually exit the organism, pass through the environment, re-enter the
organism, and pass upward until they finally show up as a change in the
controlled _perception_.
So now we can see why there is a confusion between ends and means -- why
people can say that the ends don't justify the means. The means, the
outputs, affect the environment in more than one way. As a result, the
effects re-enter the organism and alter more than one kind of
perception. At the same level where the output originated, they can
affect the inputs of many other control systems, causing errors in other
systems at the same or higher levels.
When people say that the ends don't justify the means, what they really
should be saying is that achieving one end doesn't justify frustrating
the achievement of other ends. The wisdom in the basic admonition is
contained in the fact that we have to use actions to achieve ends, and
that actions can cause problems relative to other ends if we don't
choose them wisely. You don't swat a fly with your best vase.
···
---------------------------------------------------------------------
Bruce Buchanan (940803.2005) --
How about prefacing your posts with
[From Bruce Buchanan (yymmdd.hhmm)]
?
This would mean that I wouldn't have to scan down to the end of the post
to see who sent it. I know it's unusual to put the signature before the
salutation, but it's convenient for the reader.
The data format suggested allows subtracting one time-stamp from another
to determine immediately which is earlier, very handy for sorting posts
by time of reception. I don't know a single person who has ever taken
advantage of this fact, but it's the principle, don't you see?
-----------------------------
It's really wonderful to have a person with your experience on our
little net. You bring a whiff of sanity to all discussions. It's nice to
know that there are reliable sources of health-care data. The only
problem for most of us is that we have to (or choose to) rely on others
to go through it and pick out relevant summary information, and there's
the rub. Give the same sources to Democrats and Republicans, and you
will get two different views of how well the health care system works.
Also, lots of people simply don't believe anything published by a
government (despite what you say about how foolish it would be for
governments to fudge the data). No matter what the data say, the matter
still comes down to who you trust: the people who reported the facts,
the people who collected them for publication, the people who decided
which parts to publish, the people who reviewed the data base to see
what it said. You can't get away from the people who are always involved
with producing and interpreting data. The mere fact that you were a
Ministry official will lead some people to pounce on that: AHA!
--------------------------------
A point of vocabulary. A controllING variable is a variable that
controls something else. As variables generally aren't control systems,
what this usually means is a variable that DETERMINES some other
variable, as through a physical connection. A controllED variable, on
the other hand, is a variable that is controlled by some control system;
its state is determined by something other than itself. When we speak of
a controllED perception, we mean a perception that is caused by the
actions of a system to come to a preselected state, and that is
maintained in that state by the actions. In a control system, the only
controllING variable is the reference signal; it controls in the sense
that it specifies the value to which the control system will bring the
perceptual signal. It doesn't actually control in the technical sense,
but only in the colloqial sense of having a determining influence (in
this case because of the way the control system reacts to it).
-----------------------------------------------------------------------
Best to all,
Bill P.