SMALL PCT SEMINAR-RKC

<Bob Clark (940510.1415 EDT)>

Martin Taylor (940428.1900)Subject: Re: Listening to Prozak

For some months I have been conducting a small-group seminar on
using PCT to develop a human-computer interface.

...

Almost the most difficult thing I do in this group is to get them to
concentrate on the PERCEPTION, NOT THE ACTION. They say: " The user
will want to achieve this perception [fine] so he must do that [not
fine]."

It seems to me that this is mainly a problem in translation from one
language (ordinary) to another (formal PCT terminology). As
hierarchical control systems in operation, people are already using
and discussing control system concepts in everyday language. Indeed,
they use their program and principles levels most of the time.
Working within that same viewpoint seems to improve communication.
Thus I have tried to approach the situation by using familiar terms
to identify the primary concepts, later introducing more precise
terminology and more complete analysis.

I'll be interested in your opinion of the following approach:

I begin with the concept of purposes, or goals. "Reference Levels"
terminology seems unfamiliar and a bit abstract for a lot of people.
"Purposes" seems to communicate.

Next I ask what is needed in order to accomplish your goals. The
answer, nearly always, is "power" or some equivalent. "Output
function," again is unfamiliar and abstract, but that is essentially
what they mean by "power." The formal, technical meaning of "power"
is also unfamiliar. But the need for some way to affect the
environment is clear.

The need for "corrective action" seems to be familiar. But
"corrective action" is meaningless without knowledge of the current
situation. Thus need for "perception" of the relevant aspects of the
environment is accepted. As above, the formal phrase, "feedback
function," is unnecessary.

To use their "power" for "corrective action" it is necessary to
compare the current situation with the corresponding goals. The
resulting difference(s), "errors," guide the selection and adjustment
of the "power."

This completes the basic negative feedback control system and
provides the basic concepts and terminology. The hierarchical
concept, emphasizing "PERCEPTION, NOT THE ACTION," can readily be
added.

I have had very limited opportunity to use this approach -- I'd be
very interested in knowing what you think of it.

Regards, Bob Clark