[From Rick Marken (961002.1200)]
Bruce Abbott (961001.1425 EST) to Bill Powers:
To suggest, as you do, that scientists of the day denied "the simplest and
most obvious facts of human experience" because they were indulging "in the
common human vice of trying to seem wise by denying common sense" is to
offer the most cynical interpretation possible when a far more reasonable--
not to mention charitable--explanation is available that fits the historical
facts.
And the charitable explanation is:
Prior to control theory, Darwinian evolution and its extension to learning
in the form of reinforcement theory were the only available scientific
explanations for the apparent purpose in behavior; the only alternative was
to ignore the scientific (mechanistic) view altogether and attribute
purposeful behavior to the mystical properties of the soul.
So the charitable explanation is that scientists were denying the most
obvious facts of human experience (its purposefulness) because they could
think of only mystical explanations for those facts. These scientists would
have have had to deny the existence of just about everything when it was
first discovered;- ) And when they did have an "explanation" for those facts
(reinforcement theory) the explanation was itself a denial of those facts.
Your charitable explanation sounds an awful lot like Bill's "cynical"
explanation to me.
I'm sure you've noticed that Bill's statement would be a disturbance to
someone controlling for seeing merit in the conventional approach to
understanding behavior. Just another data point;-)
Bruce Abbott (961002.0950 EST)--
What if the participant is using "correct" and "incorrect" to deduce "tone
present" and "tone absent"? This is quite easy for the participant: if she
says "present" and the experimenter says "correct," then "tone present" is
true, and so on.
Whether the experimenter's response is a disturbance or a controlled variable
depends on it's relationship to the subject's actions. If (as Bill P. noted)
the experimenter has set a reference to say "correct" or "incorrect" based on
the subject's responses, then the experimenter's responses are a potential
controlled variable (they are a controlled variable if the subject actually
wants to control what the experimenter says). If what the experimenter says
does _not_ depend in any way on what the subject says -- it only depends,
say, on what happened on a trial -- then it is a disturbance (independent)
variable. In your example above, it would be a disturbance to the perceived
logical (deductive) relationship between the subject's and the experimenter's
responses.
A controlled variable is a perceptual variable that is at least partly
influenced by the actions of the control system (subject).
A disturbance is an environmental variable that is not influenced at all by
the actions of the subject but which influences that state of a controlled
variable.
Whether the KR is a controlled variable or a disturbance variable in a
particular situation can be determined by tracing out the functional
relationships between variables in the control loop.
Best
Rick