[From Chris Cherpas (951114.0939 PT)]
[re: > Rick Marken (951110.2000)]
[re: >> Chris Cherpas (951110.1551 PT)]
RM:
The
terms "cause" and "control" are apparently used interchangably by behaviorists;
but not by control theorists. "Cause" describes a particular relationship
between variables; x causes y if variations in x lead to concommitant
variations in y, all other things being equal. "Control" describes a different
relationship between variables; y, is controlled if it is protected from the
known effects of x, by systematic compensatory variations in z.
cc:
Thanks, Rick, this moves me toward much more clarity.
cc:
Does one behavior EVER follow another behavior in time?
RM:
Sure. But remember that, in PCT, behavior is controlled perception -- a
component of a continuously (and invisibly) operating control loop. A
sequence of behaviors (in PCT) is a sequence of different settings of the
reference specification for the level of a controlled variable. The sequence
of keypresses I make to produce this post is a sequence of perceptions; it is
the result of sequential changes in my setting of the specifications for the
perception of a key being pressed.
cc:
OK. The biggest problem I've had has been understanding the different
conditions under which the two verbal communities -- behaviorist and
control theorist -- invoke certain terms, namely, control and behavior.
cc:
In all sincerity, I think I may be able to understand you but what happens
to time?
RM:
The control loop operates continuously over time; there are continuous
temporal variations in output that protect the controlled perception from
continuous temporal variations in disturbance. Time is always there,
don't worry.
cc:
OK. By the way, I'm moving along in B:CP, and it's truly a pleasure.
Best regards,
cc