[From Bill Powers (960130.1400 MST)]
Martin Taylor (960130 13:15) --
Why change the perfectly good word "control" just because some
people use the same word for something different?
Because I can't get any of the people who are participating in important
discussions on csgnet to use "control" in the same way I use it. People
are still claiming that the environment controls behavior, when in fact
if behavior changes, the environment will NOT change as needed to
restore the behavior to a particular state. I have pointed that out over
and over, but to no avail. Saying that the environment influences or
affects or even determines behavior is still considered the same as
saying that the environment controls behavior. This makes rational
discussions almost impossible.
By changing my word to retrofaction, I can avoid these problems. Let
people use "control" in any shifty sloppy way they want to. But if they
want to say something about retrofaction they are going to have to use
my definition and be precise about it, because I made up the term and I
will recognize no other definition as valid. If someone says that
disturbances retrofact behavior, I will be able to say they are simply
wrong. If they want to say that schedules of reinforcement retrofact
behavior, I will tell them they are flatly mistaken. Deprivation does
not retrofact behavior; reinforcements do not retrofact behavior; memory
associations do not retrofact behavior; nothing in the environment but
another retrofactive system can retrofact behavior. If you want to say
that A retrofacts B, you are doing to have to prove that for every
disturbance acting on B, A will produce a change in action on B that
counteracts the effects of the disturbance (as well as meeting the other
criteria that used to be called the Test for the Controlled Variable and
are now known as the Test for the Retrofacted Variable).
Don't blame me for this ugly new word, folks. You drove me to it as my
only means of counteracting an error that has been driving me nuts for
several years.
We've been through all this so often that I don't see how anyone
can fail to make the distinctions you now go to such pains to
restate
I've given up trying to understand it. I'm changing the rules.
But again to be serious, your description of retrofaction of today
does not distinguish it from ball-in-a-bowl equilibrium, so far as
I can see.
OK., if the ball is perturbed from the bottom of the bowl by a steady
10-gram tangential disturbance, what will the bowl or gravity do to
counteract this disturbance and return the ball to the bottom of the
bowl, or even _partway_ to the bottom? And how do the bowl or gravity
sense the position of the ball?
···
-----------------------------------------------------------------------
Bruce Abbott (960130.1355 EST) --
Oh, here's a little news item. Apparently this guy lost
retrofaction of his car last night and ended up, unharmed, in a
ditch. He was quoted as saying that he didn't know what happened;
he had tried and tried to get the car to retro but it just wouldn't
and just veered off the road, out of retrofaction! (:->
It looks as though the proportionality factor in the environmental
feedback function became so small that retrofaction ceased to be
possible with the means available.
"Retro" is "in" this year, so this is probably as good a time as
any to introduce the new term.
Oh, you think the word's too trendy? All right, I'm willing to change
the term to "spontefaction" if you like. Just so you understand that the
term means that if A spontefacts B, then for every disturbance tending
to alter B, A will change its effect on B to cancel the effect of the
disturbance. Plus the other provisos in the former Test for the
Controlled Variable.
----------------------------------------------------------------------
I don't think you guys are taking me seriously yet.
----------------------------------------------------------------------
Best to all,
Bill P.