[From Bruce Abbott (951020.1245 EST)]
Bill Powers (951020.0545) --
One question that seems to be overlooked is why chicks do NOT start
following large moving objects as soon as they can see and walk.
In asking this question and offering the speculative interpretations that
came later in this post, you inintentionally have provided a suburb example
of why it is important to be familiar with the literature on an area of
research (if such literature exists) before embarking on a program of
research in that area. Chicks emerge from the egg pretty much exhausted
from the effort to extract themselves from it; they must recover and clean
themselves up before much else can happen. The reason why the above
question has been "overlooked" is simple; the premise (that they do NOT etc.
etc.) is false.
Suppose we postulate that at some age, chicks become able to learn that
controlling one variable can become a means of controlling other
variables of importance to them: in other words, they become able to
learn a new level of control. What is there that a chick is highly
likely to learn to control better than it previously could, by learning
to control the relative position of some large moving object?
The evidence shows that the tendency to follow the first moving object
having certain simple, well-defined characteristics, does not depend on the
sort of learning mechanism offered above; imprinting takes place readily to
moving objects presented in a sterile laboratory box under conditions in
which the chick cannot even come into contact with the object but must
observe it through a window. Instead, the chick comes equipped with a
specific mechanism through which a suitable "object of attachment" is
identified and "locked onto." There is no doubt plenty of opportunity for
learning from here on out, (e.g., by focusing its attention on this object,
the chick learns to identify it by many features over and above those used
by the primitive identification mechanism, such as the particular sound,
markings, and smell of the mother). The chick subsequently may well learn
that it can control other variables by maintaining a specific distance from
the object, but if so, it is not the mechanism through which imprinting
takes place.
When those goslings started following Konrad Lorenz, where did he go?
Did he go into his house, into his study, and start writing? Did he just
go about his business, with a line of goslings following him in an
amusing display of imprinting? Surely, at some point, Konrad must have
led his little train of goslings to where there was some food they could
eat, some water they could drink or swim in. I seem to recall some
snapshots of Konrad sitting belly-deep in a pond with goslings around
him in the water, and of walking across a yard with goslings following
him through the grass. Where else did he lead them? And what good did it
do the goslings to follow him?
My gosh, Bill, you sound like a reinforcement theorist here! Whether you
call the result "conditioned reinforcement" or "reorganization," the tests
used to assess this possibility are essentially the same. As you might
guess, reinforcement theorists were quick to offer this kind of explanation.
So it was extensively tested--and ruled out.
Of course, you are not familiar with this work so as a PCT theorist you will
just have to go out and reprove what is already firmly established.
Hans Bloom has already done a beautiful job of defending the use of labels
in science, so I'll defer to his post on that point.
What I think, of course, has no influence on nature. I could be quite
wrong. But as far as I can see, psychologists and ethologists have by no
means exhausted the possibilities for rational present-time explanations
of the behaviors they observe. How could they have done so? They have
not had the concept of controlled variables, and they have not realized
that all behaviors are produce solely in order to control perceptual
variables.
You could see a lot farther if you had become thoroughly familiar with what
research has been done in this field, and you might have a more charitable
view of what has been accomplished. This is not to say that all the
relevant work on controlled variables has been done, by any means. But your
assertions about what researchers have and have not realized or investigated
are pure speculation on your part even if you do present them as if they
were established fact.
Regards,
Bruce