Pilots and such...

[From Fred Nickols (991123.1642)] --

Bruce Gregory was kind enough to respond to what I suspect was a stupid
posting on my part but I'm left a little confused. Bear with me while I
retrace some steps...

Earlier, I wrote...

So, it seems to me that in BPT/PCT we're dealing with the same three
categories of reference conditions: explicit, implicit and
tacit.

And...

We can
probably extend these categories to include controlled
perceptions as well.

[from Bruce Gregory (991122.1036 EST)]

Reference conditions are simply neural signals. I don't see how they can
be explicit, implicit, or tacit.>

And, regarding controlled perceptions...

Same problem. Perceptions are neural signals.

Fred from the earlier posting...

But we also have "the test," a way of determining the
perceptions that are
being controlled in a given situation. Consequently, just as
a good task
analyst or performance analyst or knowledge engineer can often make
implicit knowledge explicit, we can often make visible
certain controlled
perceptions and, by implication, the underlying reference conditions.
Moreover, just as a task analyst or performance analyst or knowledge
engineer would be distrustful of what an expert performer
says is the basis
of that performance, we, too, are not inclined to quickly
accept a person's
own views of their reference conditions.

Bruce

That's unfortunate. If you don't "know" what perceptions you are
controlling, I'd hate to ride an airplane that you are piloting. You may
have difficulty telling me what perceptions you are controlling, but
"thinking out loud" is likely to give me some good clues.

If reference conditions and controlled perceptions are indeed "neural
signals" then I don't and probably can't "know" them in the sense that I
can articulate them. That's what I meant by tacit and that is what tacit
means so far as I know. I've never piloted a plane but I doubt a pilot is
any better equipped to describe the neural signals he's referencing or
controlling any better than I am. What he or she can probably do is make
mention of those aspects of what we would ordinarily call "the environment"
to which he is attending and for which we have a language to describe
(e.g., the attitude and altitude of the aircraft, the instrument panel, the
noise of the engines and so on) and for which he or she can also describe
preferred states, which no doubt vary with conditions.

That said, I also suspect there is a great deal involved in piloting an
aircraft that is indeed "tacit knowledge" -- that is, an accomplished pilot
is very good at doing some things but is unable to articulate the basis of
doing them. I agree with you that "thinking out loud" is likely to give an
observer some good clues but these can also be misleading. The annals of
human performance technology are filled with tales of master performers who
were possessed of mistaken accounts of their own performances.

Anyway, my earlier questions have been answered so thanks to those who
responded. Now, I have another question and let's use the pilot example to
pose it:

If the pilot's reference conditions are neural signals and his or her
controlled perceptions are also neural signals, and if the aim or intention
of the pilot is to maintain an altitude of 15,000 feet, what would be the
proper PCT term for the statement (in writing or uttered aloud) "maintain
an altitude of 15,000 feet" and what would be the proper PCT term for an
altimeter reading of 15,000 feet?

···

--

Fred Nickols
The Distance Consulting Company
"Assistance at A Distance"
http://home.att.net/~nickols/distance.htm
nickols@worldnet.att.net
(609) 490-0095

[From Bruce Nevin (991123.1612)]

Fred Nickols (991123.1642) --

···

At 05:04 PM 11/23/1999 -0500, Fred Nickols wrote:

I think the distinction between explicit, implicit, and tacit has to do
with awareness, not different kinds of perceptions. A knowledge engineer
uses this distinction to help categorize the problems of building a
representation of a domain. Explicit knowledge is what she knows how to
represent in a knowledge representation (KR) program; implicit knowledge is
what she can deduce from explicit knowledge; tacit knowledge is what she
can't represent in a program. This is explained as being due to domain
experts being unable to articulate what they are doing. There is some
justice in this, since they see themselves as reformulating expert
self-descriptions in the terms required by their KR software.

Knowledge representation when I was studying such things was an attempt to
apply cognitive psychology in AI, on the assumption that actions are
planned by comparing perceptual input with cognitive maps of the world that
are built out of perceptual inputs plus memory plus logic.

If instead they identify control variables, gather behavioral data, and
model control of those variables, a different sort of breakdown might
emerge. But that might delay delivery of a product.

  Bruce Nevin

[From Bruce Gregory (991123.1632 EST)]

Fred Nickols (991123.1642)

If the pilot's reference conditions are neural signals and his or her
controlled perceptions are also neural signals, and if the aim or

intention

of the pilot is to maintain an altitude of 15,000 feet, what would be the
proper PCT term for the statement (in writing or uttered aloud) "maintain
an altitude of 15,000 feet" and what would be the proper PCT term for an
altimeter reading of 15,000 feet?

The way I think of it is "perceive the altimeter reading to be 15,000 ft".
The pilot maintains this perception using the elevators and the power
setting, but what the actions, the goal is to perceive the altimeter needle
pointing at 15,000.

Bruce Gregory