Choosing

[Martin Taylor 2000.01.24 09:26]

[From Bill Powers (2000.01.22.1234 MST)]

I don't see "choosing" as requiring special treatment as a specific mode of
behavior. ....
It would be like trying to characterize a computer by naming the programs
that can run on it. Is a computer a spreadsheet-operating device? A word
processor? A compiler? A game player? A drawing aid? A decision-maker? It
can, of course, act like all these things, but that's just an accidental
matter of programming. The computer itself does not have any particular
inclination to do any of these things; what it does is more fundamental
than any of them.

Yes, what it does is more fundamental than any of them. That fundamental
thing it does is "choosing." It is the ability to execute

IF (A is true) THEN (Do X) ELSE (Do Y)

that allows the computer to be "a spreadsheet-operating device, A word
processor, A compiler, A game player, A drawing aid, A decision-maker."
Firstly, there has to be a fundamental mechanism that allows a clean
distinction to be made between "A is true" and "A is not true", and
then there has to be a fundamental mechanism that links "A is true"
to "Do X" and "A is not true" to "Do Y."

That, precisely, is "choosing." Without it--no computers that can be
spreadsheet-operating devices, game-players, drawing aids, and so forth.
Babbage found out how to do"choosing" mechanically, or at least
understood its implications, and Ada Lovelace managed to develop it
into the first programs. Before that, there were no computers in which
the input data could determine what the machine was to do, although
there had been many clever automata--analogous to the sequence level of
perceptual control.

That's basically how I see "choosing" in the meaning of considering
alternatives and somehow selecting among them. That's not, as I see it, a
basic property of the brain, but just something any programmable device
could do -- or not do.

We know of no programmable device that can do anything if it does not
have a fundamental mechanism for "choosing." That doesn't make choosing
a basic property of the brain, of course. But if we assert that there
is a region of the perceptual control hierarchy that acts on logical
variables, and in particular that there is a level that perceives
"programs", we are invoking a very close analogy with the programmable
devices about which we do know. To sustain that analogy, we have to
assert that there exists a fundamental mechanism for "choosing."

My earlier questions were not about the fundamental "choosing" ability
of programmable systems, but about how "choosing" integrates with the
normal action of a hierarchy in which the output signals of _all_
control units are simply increases or decreases in signal magnitude.

If there is, as you suggested, an ability in the higher-level units
to select where their output is directed, it would seem that the
mechanism that does the redirection must be distinct from the output
function whose signals are redirected. You seemed to dispute this
statement. Could you clarify how you think the same output function
could do both? Or if it isn't the same output function, how the extra
output functions might be connected within this higher-level unit?

-------------speculation ahead-----------

If we want to pursue the analogy between brain-logical processes and the
silicon ones we understand, the fundamental nature of "choosing" becomes
more deeply embedded.

In the analogue levels of perceptual control, when there is an error in
a control unit, the resulting output can be characterised as "do more/less
of what you were doing." A logical control unit (one with a logical
perceptual signal) could do the same thing, but not if the error signal
is also logical. Computers don't control by "do more/less of the same".
They output "do something else instead"--IF (error) THEN (Do X) ELSE
(Do Y).

Following the analogy, one can imagine the levels of the hierarchy above
category as having output functions that influence switches connected
a "true" or "false" value to the reference inputs of units immediately
below them, rather than having output functions that vary the magnitude
of the signal fed to a fixed (apart from reorganization) set of lower
level reference signals.

As I said, this is pure speculation, and I haven't even tried to think
of the ramifications of such an idea. Analogies of brain function to
mechanical devices have historically been disastrously wrong. But it's
worth keeping in the back of one's mind if we continue to think of there
being "logical" levels of perceptual control. After all, Boole developed
his logic as an attempt to describe how the mind worked, and it may be
fair to reverse the process.

---------end speculation segment--------

Most people don't have a very clear idea of
what they mean by choosing. In many cases, the meaning is simply that you
did something on purpose (chose to create a result) rather than causing it
accidentally, with no particular mention of alternatives.

Most people don't have a clear idea of what "perception" means in PCT,
either. That doesn't invalidate the use of the concept as a fundamental
aspect of the theory, does it?

In other cases,
the alternatives are the important factor, with an unspoken assumption that
you can't have them all at the same time.

Not so "unspoken," I think.

A CT analysis of these two
meanings would show that they are operationally completely different,
involving different processes. And I wouldn't be surprised if there are
others.

Yes, they are quite different. And quite possibly there are others. If
one wants to play word games, English is a lovely playing field. Almost
all common words have myriads of meanings that can be used in the game.
It's lots of fun, and even sometimes leads somewhere. But not often.

···

-----------------------

In the RTP case, I like Erling Jorgensen's interpretation as a kind of
warning or reminder.

Good. And to Chuck Tucker's list of Bruces, I think the name of Bruce G.
might well be added, even though he has given up on CSGnet, as one who
advocated a similar interpretation.

Martin

[From Bill Powers (2000.01.24.1029 MST)]

Martin Taylor 2000.01.24 09:26--

IF (A is true) THEN (Do X) ELSE (Do Y)

This describes a choice-point in a program; in a closed loop, it's the TOTE
unit. If this is what you want to mean by "choosing", then I guess I can go
along with it. I don't thinks it covers all situations in which people use
that term, but if that isn't the concern -- if you're just establishing a
technical term -- that's not a problem. Of course in that case, the meaning
of "I see you've chosen ..." suggested by you, Tucker, the Bruces,
Jorgensen, and apparently almost everyone else but me doesn't apply, does it?

If
one wants to play word games, English is a lovely playing field. Almost
all common words have myriads of meanings that can be used in the game.
It's lots of fun, and even sometimes leads somewhere. But not often.

Not often, I agree.

Best,

Bill P.

I shouldn't do this, really.

But I will. This is too good to pass up.

I heard on the radio this morning an interview with a Russian spokesman,
Mr. Konenko (sp?) re the Russian invasion of Grozny.

"They [the civilians in Grozny] are free people living in a free country,"
he said. The Russian army had given them a choice to leave safely or stay
and risk being killed. "But," expostulated the interviewer, "this is
their city. Why should they have to leave their own homes?" Konenko
replied (utterly ignoring the interviewer's point) that for two months they
had had the opportunity to leave peacefully, so what happens to them now is
their own fault.

-- Richard Kennaway, jrk@sys.uea.ac.uk, http://www.sys.uea.ac.uk/~jrk/
   School of Information Systems, Univ. of East Anglia, Norwich, U.K.

[From Bruce Gregory (991216.0527 EST)]

I shouldn't do this, really.

Conflict is a necessary element in choice.

Bruce Gregory

[From Bill Powers (991216.0832 MDT)]

I shouldn't do this, really.

You're right, Richard, but you're right that it's too good to pass up.

"But," expostulated the interviewer, "this is
their city. Why should they have to leave their own homes?" Konenko
replied (utterly ignoring the interviewer's point) that for two months they
had had the opportunity to leave peacefully, so what happens to them now is
their own fault.

What you're forgetting, Richard, is that the Russians are the Bad Guys, so
when they do that, it is a Bad Thing. However, we are the Good Guys, so
when we do it, it is a Good Thing. Do you understand now?

Best,

Bill P.

[From Bruce Gregory (991216.1217 EST)]

Bill Powers (991216.0832 MDT)

What you're forgetting, Richard, is that the Russians are the
Bad Guys, so
when they do that, it is a Bad Thing. However, we are the
Good Guys, so
when we do it, it is a Good Thing. Do you understand now?

Exactly. When we bomb Yugoslavia it is a Good Thing, because we are the
Good Guys.

Bruce Gregory