[From Bill Powers (2004.01.12.1136 MST)]
Marc Abrams (2004.01.12.1133)--
> PCT is a model of human and animal organization. You say I don't need
> models of that model. I don't even know what that means.
Oh I see, from a model of 'purposeful human behavior' it has turned into a
model and theory of 'human and animal organization'.
Who said 'purposeful human behavior?' That was Bruce Gregory, wasn't it?
Will it wash my windows as well? I _really_ think your overreaching
here. But I'm certainly in no position to judge such things.
Judging from these comments, I'd say you're right. At least object to what
I, not someone else, said.
So this argument of you/me is shear nonsense. It is not about whose ideas
are superior. It's about where I want to spend my resouces. I have a
limited amount of energy, money, and time and just like you built on
Weiner's model, I'm going to build on yours so I guess you could either be
flattered or feel like I'm trying to 'steal' something from you and be
_real_ paranoic_ about the whole thing. It's your choice, I have no
control over how you perceive me and what I'm attempting to do.
I agree about the limited resources; I'm feeling stretched right now, too.
Incidentally, I learned about control theory before I ever read Wiener and
Ashby. Wiener and Ashby were not control engineers. Wiener was a
mathematician, and Ashby was a psychiatrist. I learned my control theory
from engineering texts, analog computing, and developing control systems
for medical and scientific uses.
I thought that we accepted models by testing them all and
picking the one that predicts best.
This sounds wonderful but not doable in the real world.
On the contrary, this is how models are selected in all the real sciences.
They are not selected by "believing in them." They either predict well or
predict poorly. It's not hard to pick the one you prefer.
I once gave you a
review paper on Olfaction. It tied our ability to perceive taste and smell
on past experience. The paper had 60 cites. B:CP has 72 just to understand
that this review was not based on a whole bunch of hand waving. It took you
about 1 hour to trash it. Did you test the model as it was proposed? I don't
think so.
There wasn't anything I could see to test that was within my experimental
capabilities, or those of the authors, either.
What model of Memory & Olfaction did you use in determing what
role it plays in the construction of perceptions?
I don't have one outside my general model of how memory is related to
perception.
>Also, isn't it a little early to start believing in ideas that you haven't
worked out yet?
See Bill, this is what I have a _real_ hard time with. If you show me some
data, I'll be more than happy to change a position I hold based on it.
But I don't believe in things first and then set out to prove them. To the
degree that I believe anything in PCT, the belief comes only after I have
proposed a model and shown at least some reason to accept it. I have spent
many years developing concrete demonstrations of the principles of PCT,
demonstrations that anyone can use to verify that the phenomena exist and
work as I propose they do. Even so, what I "believe" is a very limited set
of the things in the demos, and I do not cling to or defend those beliefs.
If challenged, I can go back to the basic principles and observations from
which I derived my proposals, and lay out my reasoning so anyone can check
up on me. It's always possible that I made some mistake that others will
discover, and I try to make it as easy as possible for others to find any
mistakes.
But you have _not_ done that. You tell me that neurologists know nothing
about the nervous system like you do. What data do you have to back up
your claim?
When I get that book, I'll cite chapter and verse. Just one hint. How many
studies have you seen in which a neurologist varied the frequencies of
impulses in a set of signals coming into the dendrites of a neuron, and
measured the output frequency of the neuron's firings? I will predict that
the approximate number is zero. Yet this is the only way to determine the
nature of the analog computation performed by the neuron.
I'm open. Skeptical, but open. You claim your hiearchal model represents the
physiological reality of our bodies. _A_ hierarchy, _maybe_. _Your_
hierarchy, kind of doubtful. Not impossible, but I just don't see it. Now,
just because I don't hold a firm unyielding view on something doesn't mean I
can't believe in certain concepts, or _not_ believe in others.
You seem to think that belief is a virtue. In science, it's a vice that one
is supposed to try to get rid of, especially when trying to develop
something new. Belief shapes our perceptions more strongly than anything
else: given alternatives, we choose the perception that matches what we
believe. Think of what that does do your ability to discover anything
contrary to your belief!
You seem to be saying that science consists of stating beliefs and then
trying to prove that you're right.
No, that's what you seem to be practicing.
>Do the beliefs come before the proofs?
I hope so.
If you don't see the complete contradiction between your last two
statements, I give up. You seem to think that a belief in a explanation
should come before proving that the explanation is correct. I say you
should first show that the explanation is correct (as nearly as you can)
and then decide what degree of belief, if any, to invest in it.
Well, this goes nowhere. Let's move on to other things.
Best,
Bill P.