From Tom Bourbon [940421.0822]
This will probably end up trapped in the local queue, or the dumper, but
here goes.
[from Wayne Hershberger]
Bill Powers (940420)
I'm not sure what you mean by "a control system which
transmits information." What does a control system transmit
information from, and to?
From environment (i.e., disturbance) to organism/mechanism (i.e.,
output). The more closely the output mirrors the disturbance the
more fully the organism/mechanism is informed about the state of
the environment.
The control system transmits this "information?" The system is "informed"
if its output mirrors a disturbance? For certain? Or is this merely
another case of "metaphor creep?"
I wish we could get past the endless conflation of "information, the
descriptive statistic after Shannon" and "information, knowing about and
being informed." The two have become hoplessly intertwined once again, in
this exchange. In which sense(s) do you mean "information," Wayne?
The fact that this information is to be found in
the system' output rather than its input is neither here nor
there. As you say,
It's not necessary to trace the flow or know the state of
information inside the system between the disturbance and
the output variable.
Nonetheless, couldn't one quantify the amount of information
transmitted in this way (forget the ideal system; I'm not sure
what I was driving at, anyway)?
After some additional time for reflection, did you remember what you were
driving at? 
As a reader, I wasn't sure what you were trying to say.
I am not suggesting for a second that Rick is wrong when he argues
that the information in question is not to be found in the
controlled input.
Why not? Why do you think his argument is correct, or do you just want to
avoid luring him out of his lair?
For a few weeks now, I have been frustrated by the local network problems
that kept me from throwing a little high octane gasoline on the inferno
raging over "informed perception." Bill had posted some data showing a
small correlation between the handle movements (the integral or the
derivative) and various data and derivatives that pertain to the
disturbance. Bill, Rick and I saw the low correlation and the associated
high coefficient of alienation (coefficient of uselessness -- the
correlation "accounted for" only about 2% of the variance in the
relationship) as evidence that the person's handle movements were independent
of the derivative of the disturbance; in contrast, Martin invoked the
technically correct idea that any non-zero correlation is a non-zero
correlation -- he saw evidence in favor of dependence -- he liked the 2%
solution.
I have data from analyses of multiple runs on a pursuit tracking task. I
believe thoes data can add a little fuel to this all-consuming fire. I will
post them when I have some degree of confidence that my message will leave
near-office orbit and make its way into csg cyberspace.
Later,
Tom
As a reader, I wasn't sure what you were trying to say.