[From Bruce Gregory (970313.1355 EST)]
Our science education media group has produced three one-hour
videos with the title _Minds of Our Own_ which should be on
Public Televsion in the U.S. this coming fall. We were watching
a portion of one this morning dealing with people's ability to
make simple predictions about what they would see in a mirror.
Example, most people believe that to see more of yourself in a
mirror you should get further away from the mirror. They are
amazed when they actually try this. I realized that this is a
very nice example of an area of life in which people clearly do
_not_ use a model to make predictions. You might argue that they
do use a model, but not a very good one. In any case, the model
or lack thereof can play no useful roll in exercising control
when it comes to seeing where you want to see using a mirror. Yet
people have no trouble adjusting mirrors to see where they want
to see. Where mirrors are concerned, people exercise control but
apparently have no need to develop predictive models to aid that
control.
Bruce Gregory
[Hans Blom, 970318d]
(Bruce Gregory (970313.1355 EST))
We were watching a portion of one this morning dealing with people's
ability to make simple predictions about what they would see in a
mirror.
I take it that the "ability to make predictions" is intimately
connected with having a model (of what to expect).
Example, most people believe that to see more of yourself in a
mirror you should get further away from the mirror. They are amazed
when they actually try this. I realized that this is a very nice
example of an area of life in which people clearly do _not_ use a
model to make predictions.
What _do_ people use to make a prediction?
You might argue that they do use a model, but not a very good one.
I would argue that, indeed...
In any case, the model or lack thereof can play no useful roll in
exercising control when it comes to seeing where you want to see
using a mirror.
Do not their expectation guide their behavior?
Yet people have no trouble adjusting mirrors to see where they want
to see.
Many (all?) of our models are in error and/or contain errors or gaps.
Only by _doing_ (and thinking, some say) we learn and improve our
models. Making accurate predictions is possibly the most difficult
thing there is, especially when you do something for the first time.
Yet we routinely indulge in making predictions -- in other words:
having expectations. In fact, we may become pretty violent when some
of our more potent expectations are thwarted. To our luck, evolution
has done most of our learning _for_ us...
It seems to me that you confuse "having a model" with "having a
_correct_ model". It appears to me that we routinely use
"expectations" to guide our actions. Equally routinely we are
disappointed. Or surprised...
Greetings,
Hans
[From Bruce Gregory (970318.1020 EST)]
Hans Blom, 970318d
me:
>In any case, the model or lack thereof can play no useful roll in
>exercising control when it comes to seeing where you want to see
>using a mirror.
Hans:
Do not their expectation guide their behavior?
As far as I can tell their expectations are minimal and are not
refined by practice. Think of how you adjust a rear-view mirror
when you get into a car that has been driven by someone else. I
suggest that most people (MCT mavens excluded) anticipate that
moving the mirror will alter what they see out the back window.
I'd be willing to bet that they move the mirror, see what
happens, and correct as necessary until they see what they want
to see. I'd even bet that physicists do not invoke a model to
accomplish this task. (Again MCT people doubtless do. 
Bruce Gregory
[Hans Blom, 970318e]
(Bruce Gregory (970318.1020 EST))
I suggest that most people (MCT mavens excluded) anticipate that
moving the mirror will alter what they see out the back window.
How come most people (MCT mavens possibly excluded do anticipate?
What are anticipations? Where do they come from? What generates them?
And, most importantly, how come those anticipations influence
subsequent behavior?
Greetings,
Hans
[From Bruce Gregory 9970318.1555 EST)]
Hans Blom, 970318e
How come most people (MCT mavens possibly excluded do anticipate?
What are anticipations? Where do they come from? What generates them?
And, most importantly, how come those anticipations influence
subsequent behavior?
All interesting questions, but, as far as I can see, not relevant
to the point I am making, which is whether or not people make
use of model of the behavior of mirrors any more sophisticated
than the idea that mirrors show you an image of what is in front
of them and if you move the mirror around the image changes.
Bruce Gregory
