<Bill Cunningham 940527.1000>
To Bill P and Rick,
Thanks for your responses. I understand what you mean and agree that we spend
most of our time controlling for mundane things, not reduction of uncertainty.
And I agree fully that I'm concerned with some higher level perception.
It might be useful to explain that my interest is almost exclusively at the
higher level, and give an example. From this, you may at least understand
why I am very conscious of controlling for reduced uncertainty.
The problem I am trying to serve is one of third party epistemology. The
protagonist seeks to develop/improve internal model of some part of the
world either too remote or to widespread to sense directly. The only
recourse is to rely on previous model, updated by data from humans or machines
that can be sent to the area of concern. This an imperfect reporting system,
leaving lots of uncertainty about what is really happening. Ultimately,
the protagonist will have to make decisions committing resources aimed
at changing his/her perception of the remote scene.
Take the LA earthquake as an example. Put yourself in the position of
the following individuals, in turn: A fire captain in LA up to
his ass in alligators; a state official (governor, National Guard, other)
in Sacramento who must muster and channel resources to where they are
needed; a FEMA official in Washington who must also muster and channel
resources, but on a different scale and on different criteria. They all
have a common problem, "What's going on that I can/should change and
what the hell do I do next?"
I watched the news coverage of this event with considerable interest and
frustration at the fragmented reporting and repeated 30 second "high
viz sound bites" that did nothing to build a cohesive picture. There was
no cohesive picture until well after dark in LA and late at night on the
East Coast. And not just for me as an interested bystander. It was
perfectly clear that the above named officials never had a timely and cohesive
picture of their own domain, although each had reporting sources in addition
to the news media. When the fire captain was interviewed (great timing, press)
he didn't know how many simultaneous problems over which to spread his
resources. The National Guard official in Sacremento knew he had to
mobilize, but didn't know which potential marshalling points were/could
be operational and the state of road network to support whatever he had to
do. The FEMA official had an easier job in terms of what to do next, but
would eventually need to know how much money/field agents to dispurse.
These are real people facing a real problem. They control for whatever they
control. Without argument, they do vast amounts of low level controlling.
But attention/control is shifted at some level(s) and on some basis. Control
of higher order perceptual variables seems on the mark, but that implies
some sort of conditional branching and leads back to the same question.
As for a model of selecting low level controlled variables on the basis
of a high level control for uncertainty, all the mechanics of enquiry
fit. Our fire captain might control for picking up microphone, squeezing
switch, the muscle movements for asking "How many fires in my precinct?"
On hearing "Two reported" when he can see five, he might now control
for other means of enquiry. He surely has basis for modifying his model
of the reporting mechanism (like maybe the phones are not working).None
of this requires a new hierarchical model. It merely says that the fire
captain has several choices of action at a very high level of abstraction,
and that the probable outcome of each choice is dependent on the validity
of his model of the situation. So his intermediate course of action may be
to control for reducing any uncertainty he may have about his model.
I now have to control for getting out a weekly report.
Have a great holiday weekend,
Bill C
.=====================================================================
A belated report on shadow.net. I forwarded the offending announcement to
my son, whose job begets knowledge of internet security problems. Hedid
not locate the shadow.net host before it disappeared, but did determine that
it was operating as a blind mailbox for 20-30 individual users who would
log in remotely. He seemed to think that Mr. Thigh Cream was one of the
clients of an unsavory business rather than the prime hacker.
PROFS: mon1(cunningb) Internet: cunningb@monroe-emh1.army.mil
Phone (804) 727-3472/DSN 680-3472. FAX ext 3694/2562
*Eschew Obfuscation*