[Martin Taylor 951124 11:00]
Bill Powers (951123.1830 MST)
The problem, I think, is that you have in mind a conclusion, and you're
trying to set up assumptions that will lead there.
Why don't you just come right out and say what you're trying to prove,
and pick an example from real life to illustrate it?
I'm not trying to prove anything at all. I'm trying to illustrate the
connections among several different threads of discussion. I'm picking
illustrations that show the extreme conditions of certain problems, so
as to ease them into a continuity. I don't remember who it was that
introduced "error-free learning" into the discussion, but that's what
I'm trying to achieve, by making the different conditions as different
as possible, to reduce the likelihood that extraneous paths will be
followed.
The jacket-flipper whose ritual cannot be varied is one extreme of a
particular problem of loops completed through side-effects, Jack the cat
who will escape by side effects almost no matter what ritual he chooses
is the other extreme. How different will be their two possibilities for
turning the important side-effect into a perception that they will
deliberately control?
What happens with most rituals, even if performed with
the utmost exactness, is that they fail most of the time, but people
continue to have faith in them because they think they have no other way
of getting what they want.
You are right that most rituals don't achieve the desired results because
of anything in the ritual itself (though some do, and I did give a real life
example that you now ask for--willow-bark tea). However, I disagree that
people have faith in them because they have no other way to get what they
want. If that were the case, one would expect the rituals to be destroyed
by reorganization. We need some new postulate to explain why that doesn't
happen (a question to which we may return some day--I think it has to do
with language and the maintenance of cultural artifacts by the mechanism
I proposed in Durango 93).
In many rituals the desired result sometimes comes about for reasons
unrelated to the person's actions (and therefore the ritual is frozen in
the reorganization process as if the ritual truly worked). When rituals
don't work, having been incorporated into a person's organization, there
are often other means, which would allow the ritual not to be disorganized.
And in humans, at least, imagination plays a role, which we can verbalize
in two phrases: "God willing" and "God helps those who help themselves."
If God is not currently willing, the ritual must be performed with more
emphasis, in the same way as a conflict often results in an increase of
the outputs of the conflicting control systems, before reorganization
eliminates the conflict.
When a side effect completes the action side of a high-level control loop,
it seems to me that there are two possible end points of the reorganization
process: (1) ritualizing, and (2) discovery (by which I mean discovering
a perception directly affected by the side-effect action, so that controlling
the new perception becomes a reorganized part of the high-level control
loop). When random changes in the action sequence (before it is ritualized)
are quite likely to maintain the useful side-effect, then discovery seems
to be a likely outcome, whereas (as in the jacket-flipper situation) when
random changes almost certainly eliminate the useful side-effect, the
result seems likely to be ritual.
But this isn't what I am "trying to prove." It is one aspect of tying
together several different threads by the use of intersecting examples.
Real life is often more complicated than the examples, and complicated
examples lead in more directions than is fruitful for particular
discussions, even when the connection between the simple and the
complicated situations is quite smooth and direct.
Jack the cat scientist is like an S-R psycholgist, but the S-R psychologist
is a much more complex case. Jack's situation is understandable, and
at least in principle wholly predictable from PCT. But if Jack had a
whole slew of interacting sticks, buttons, pulleys and levers that he
could use in a ritual that worked 30% of the time, his behaviour would
probably be both less predictable and more ritualized. When Jack leaves
his ritual, his success rate drops, so why leave the ritual? (or to put it
in more mechanistic terms, behaviour that differs from the ritual is
more likely to be reorganized away than the sometimes successful ritual
behaviour).
It takes Jill the kitten to find the piece of connecting rod that moves the
latch, and Jill doesn't need Jack's ritual if she is to escape. But Jack
will probably keep telling Jill that she has to follow the ritual or she'll
never learn anything.
Perhaps the problem you're dealing with is imaginary.
Of course it is. That's what science is about, isn't it? How did Jill find
the effective connection to the latch without imagining the possibility?
How did you develop PCT other than by imagining the problem?
The relation between religion and science is not an imaginary problem. It
is coming to the fore in the politics of many nations, not least the USA.
I think PCT has something to say about where this unnecessary conflict
comes from, in a quite direct way. But that's not "what I am trying to
prove" because I am trying, not to "prove" anything, but to illustrate
and explore a great deal, of which this issue falls out as a by-product.
"Explore" is the key word. I got into this thread about Thorndike's cat
because nobody had mentioned the stereotyped nature of the cat's actions,
and because I had a vague puzzlement in my mind. As we, together, explore
the different changes in the situation, the issues and linkages have become
much clearer. The exploration seems to me to be well worthwhile continuing,
and not only in the direction emphasized in this message. I hope that
you think so, too, and that others will start to contribute to this thread.
Martin