[From Rick Marken (940525.2330)]
Bill Cunningham (940525.1335) asked --
My typing this constitutes behavior, implying that I'm controlling
a perception. What perception?
I (940525.1515) guessed --
you are trying to control your perception of the relevance of
information theory to control theory. You want to perceive that
relevance as high. Your answer to my comments about controlling
uncertainty should help me determine the accuracy of that hypothesis.
Bill Cunningham (940525.2020) replies:
Your hypothesis of what I was controlling for is false.
OK. I'll take your word for it.
So then I take it that the following will create no disturbance:
Bill Cunningham understands that information theory is irrelevant to
control theory.
Your hypothesis that my answer will help determine the accuracy of that
hypothesis is therefore correct, independent of my comments about
controlling uncertainty.
My hypothesis was that you were controlling for the relevance of IT to
PCT. That hypothesis was based on your discussion of my control of
uncertainty. I tested that hypothesis by saying that uncertainty is
not a perception and, thus, not a controllable variable. If you agreed
with that analysis then it would have been evidence against my hypothesis
because my analysis is a way of saying that one important IT concept is
not relevant to PCT; agreeing with that analysis means it is not a
disturbance; I presume that my analysis would have been a disturbance to
the perception of a relationship between IT and PCT. Disagreeing with my
anaysis would have been a correction to the disturbance -- and a bit of
evidence for me that your perception of the IT - PCT relationship is under
control. Since you neither agreed nor disagreed with my anaysis I really
have no evidence either way about whether you were controlling the perceived
IT-PCT relationship. I could count your non-answer as a non-response to the
disturbance -- supporting the hypothesis that you were not controlling
for the IT-PCT relationship; or it could be seen as an evasive response;
the analysis was a disturbance that was defended by avoidance. I have
no idea how to interpret your non-answer, so I'll just take your word for
it that you were not controlling for the perception of an IT-PCT relationship.
This means that I have to reject my hypothesis about the variable you
were controlling. Notice that this is an important part of the test for
controlled variables; when you hypothesize a controlled variable you
CAN be wrong (as, apparently, I was). That's why we TEST; in order
to see whether our guesses are correct.
So the implication of the delightful G&S quote is, I think, wrong in
this case. When I guessed that you were controlling the relationship
between IT and PCT I was not sure that I was right "beyond a shadow of
a doubt"; I knew that there was the possibility that I was wrong. And,
as it turns out, I was wrong (based on your claim; not on the results
of the test).
Look to Martin Taylor's Layered Protocol theory for answer.
But I thought you already gave the answer (that my hypothesis about
the variable you were controlling was false). And why not just tell me
the answer (to whatever you think is the question) straight out?
I know that you were in a rush when you typed this but it does seem
like a rather paradoxical approach to providing answers, coming from
a person who puts the following good (if somewhat oxymoronical) advice
at the bottom of each post: Eschew Obfuscation.
Best
Rick