Biased reorganization (was Re: Magician's new clothes)

CSGnet Archive CSG2016
1

[Martin Taylor 2016.06.29.09.07]

These two comments lead me into PCT speculative territory, because I

think they are both likely to be correct, and that leads me to ask
myself whether maybe reorganization tumbles are sometimes, usually,
or always biased rather than being in purely random directions. The
deception described by Rick is an example of the general demagogic
practice of providing an easy answer for “the sea of troubles and by
[doing what the demagogue want] ending them.” But what if the
concept of “easy answer” could bias internal reorganization? Might
tumbles have “easy” and “hard” directions. For example, an “easy”
direction might involve only one or two levels and require using
existing lower-level machinery in a new way, whereas a “hard”
direction might be a multi-level change that required construction
of new controlled perceptions at several levels – seeing the world
in a new way (as seems to happen when one catches on to what PCT is
really all about).
I don’t have any idea of a mechanism. It’s just a suggestion that I
might disown tomorrow, or that might lead somewhere some day.
Martin

···

(Vyv Huddy [1152.29.06.2016])

              VH:

Thanks for your reply Rick. All makes good sense. Your
comments reminded me of a Bill Powers quote that Tim Carey
mentioned at the April MOL workshops “t he childhood of
the human race is far from over. We have a long way to go before most people
will understand that what they do for others is just as
important to their well-being as what they do for themselves.”

From:
Richard Marken 28 June 2016 18:11
Re: Magician’s new clothes

[From Rick Marken (2016.06.28.1010)]

                        Vyv

Huddy (26:06:16.2028)

Rick Marken (2016.06.26.840)

                                    RM:

I think what happened [with the
Brexit vote] was, to a large
extent, control by deception;

                                  VH: I wonder if

there’s a role for
reorganisation.

                  RM:

Yes. But control by deception takes advantage of
control organizations that are fairly stable –
that aren’t reorgnizing. So if you know that
people are controlling for being “free” then you
can get them to vote for you by saying that you
will remove the rules that have constrained their
freedom. It’s just like in the rubber band demo
(or operant conditioning); set things up so that
the only way people can (or think they can)
control for what they want is by doing what you
want them to do.

                  VH:                          Many people

are experiencing massive material challenges,
there are many food banks, high rates of
unemployment, poor housing etc. For these people
I think it was reorganisation to “change the
status quo” (the opposite of usual behaviour in
referendums). But unfortunately this group don’t
realise that this is going to increase error
not decrease it. Unlike an e coli they can’t
tumble again and get chance to regain control.

                    RM: Yes, I guess it could

be seen as reorganization; taking a random swing
at change. But I like the control by deception
idea better because the vote to exit the EU was
not really a random change in a control system.
It was specifically what some people wanted
voters to do and they got many of them to do it
– vote for exit – by lying to them about what
it would accomplish.

rsmarken@gmail.com
Sent:
**To:**csgnet@lists.illinois.edu
Subject:

2

(Vyv Huddy [1730.02.07.2016])

···

VH: I’ve been playing around with the LCS III demos today and was reminded of your comment Martin.

[Martin Taylor 2016.06.29.09.07]

MT: But what if the concept of “easy answer” could bias internal reorganization? Might tumbles have “easy” and “hard” directions. For example, an “easy” direction might
involve only one or two levels and require using existing lower-level machinery in a new way, whereas a “hard” direction might be a multi-level change that required construction of new controlled perceptions at several levels – seeing the world in a new way
(as seems to happen when one catches on to what PCT is really all about).

VH: When you say “hard” or “easy” it sounds like you are saying there is some prior knowledge of the consequence of reorganisation before it happens? It might be worse in
some places than others?

VH: Also isn’t subjective ease or difficulty always something that reflects control happening in the present? The exception could be where control is happening in the imagination mode? This would involve imagine future
(or past) states of variables? Such as the voting example I posted last week.

I don’t have any idea of a mechanism. It’s just a suggestion that I might disown tomorrow, or that might lead somewhere some day.

Martin

3

[Martin Taylor 2106.07.02.13.31]

··· 
          (Vyv Huddy

[1730.02.07.2016])

          VH: I've been playing around

with the LCS III demos today and was reminded of your
comment Martin.

          [Martin Taylor

2016.06.29.09.07]

  Sorry. My words were obviously misleading. I had

never thought of a controlled (or uncontrolled) perception of
“easy” or “hard”. I was thinking more of the situation of a
raindrop falling on rough terrain. There’s an “easy” downhill
direction for the water to go and a “hard” uphill direction. If
the drop falls on a flat surface, there’s no “easy” or “hard”
direction. All directions are equally easy or hard. That’s also
the case with e-coli reorganization as currently described.

  The analogy I see in reorganization is that to change one linkage

is easy, while a coordinated change in several linkages is hard.
If either would be equally effective in regaining or improving
control, would the reorganization mechanism tend to find the
“easy” change? Classic e-coli reorganization changes all the
linkages randomly at every tumble, and the direction of change
that is then followed has no preference for such “easy” solutions,
solutions that would be found by an external system designer
trying to make his system work better. Almost all invention
consists of making simple modifications of existing products, as a
bicycle came from adding pedals to a velocipede, or of combining
existing modules in new ways. Invention preferentially finds
“easy” paths to improvement, and I suspect that evolution is
likely to have produced a mechanism that does the same.

  I still don't have a mechanism in mind, and although I didn't

disown the idea the next day (nor have I yet), I may still do so.

  Martin

            MT: But what if the concept of

“easy answer” could bias internal reorganization? Might
tumbles have “easy” and “hard” directions. For example,
an “easy” direction might involve only one or two levels
and require using existing lower-level machinery in a
new way, whereas a “hard” direction might be a
multi-level change that required construction of new
controlled perceptions at several levels – seeing the
world in a new way (as seems to happen when one catches
on to what PCT is really all about).

            VH: When you say

“hard” or “easy” it sounds like
you are saying there is some prior knowledge of the
consequence of reorganisation before it happens? It might be worse in some places
than others?

4

[Martin Taylor 2016.07.03.10.01]

My original message on biased reorganization [Martin Taylor

2016.06.29.09.07] was intended to offer an analogy between the
demagogue’s offer of an easy way to fix problems that is exemplified
by your "

···

A Postscript to …

[Martin Taylor 2106.07.02.13.31]

  some prior knowledge of the consequence of

reorganisation before it happens?" and the
possibility of biased reorganization.

  In the case of the demagogue, the "easy answer" is, as Rick

pointed out, usually deceptive. But easy answers aren’t always
deceptive. You may be wrestling with a problem, whether it be
drying the laundry, solving a differential equation, or solving
the issues of social inequality, and someone suggests “Why don’t
you try it this way. It ought to be much easier than what you seem
to be trying.” Such “easy answers” do presume to provide " some prior knowledge of the consequence of
reorganisation before it happens", except
that no reorganization is required.

  "Trying it this way" uses no novel control linkages, but it does

imply a controlled perception of linkage change. You give up
trying to hang the laundry on an inadequate number of chairs
arranged about the lawn, but hang them on a line in the sun.
That’s a perceptually controlled linkage change, sending the
output of “perceived clothes dampness” controller to different
lower level control units. In the language of atenfels, the
perceptual controls involved in putting the clothes over
chairbacks are atenfels that go from active to potential, while
those involved with setting up and using a clothesline go from
potential to active.

  Changing the atenfels used in control of a perception (using

“prior knowledge of the consequences”, even if that “knowledge”
turns out to be wrong)
seemed to me to have much in common with reorganization
that produces new structure in the hierarchy. Both potentially
offer “easy” and “hard” directions of change. The demagogue offers
“prior knowledge of consequences” that are probably deceptive, but
are used by the gullible to change the atenfels for controlling
perceptions they have been trying unsuccessfully to control (“our
country’s greatness”, for example). The helpful teacher offers
“prior knowledge of consequences” that is probably reasonably
reliable. The well-meaning friend offers prior knowledge that has
a dubious likelihood of turning out to be correct.

  The reorganization system has no obvious way of finding easy

directions in preference to “Rube Goldberg” (Heath Robinson)
solutions to problems of poor control, but I have faith that
evolution is likely to have provided us with some mechanism to do
it.

  Martin

(Vyv Huddy [1730.02.07.2016])

            VH: I've been

playing around with the LCS III demos today and was
reminded of your comment Martin.

            [Martin Taylor

2016.06.29.09.07]

              MT: But what if the concept

of “easy answer” could bias internal reorganization?
Might tumbles have “easy” and “hard” directions. For
example, an “easy” direction might involve only one or
two levels and require using existing lower-level
machinery in a new way, whereas a “hard” direction
might be a multi-level change that required
construction of new controlled perceptions at several
levels – seeing the world in a new way (as seems to
happen when one catches on to what PCT is really all
about).

              VH: When you

say “hard” or “easy” it sounds like
you are saying there is some prior knowledge of the
consequence of reorganisation before it happens? It might be worse in some places
than others?

    Sorry. My words were obviously misleading. I had

never thought of a controlled (or uncontrolled) perception of
“easy” or “hard”. I was thinking more of the situation of a
raindrop falling on rough terrain. There’s an “easy” downhill
direction for the water to go and a “hard” uphill direction. If
the drop falls on a flat surface, there’s no “easy” or “hard”
direction. All directions are equally easy or hard. That’s also
the case with e-coli reorganization as currently described.

    The analogy I see in reorganization is that to change one

linkage is easy, while a coordinated change in several linkages
is hard. If either would be equally effective in regaining or
improving control, would the reorganization mechanism tend to
find the “easy” change? Classic e-coli reorganization changes
all the linkages randomly at every tumble, and the direction of
change that is then followed has no preference for such “easy”
solutions, solutions that would be found by an external system
designer trying to make his system work better. Almost all
invention consists of making simple modifications of existing
products, as a bicycle came from adding pedals to a velocipede,
or of combining existing modules in new ways. Invention
preferentially finds “easy” paths to improvement, and I suspect
that evolution is likely to have produced a mechanism that does
the same.

    I still don't have a mechanism in mind, and although I didn't

disown the idea the next day (nor have I yet), I may still do
so.

    Martin
5

[Vyv Huddy 18.08.04.07.2016]

···

[Martin Taylor 2016.07.03.10.01]

MT: In the case of the demagogue, the “easy answer” is, as Rick pointed out, usually deceptive. But easy answers aren’t always deceptive. You may be wrestling with a problem, whether it be drying the laundry, solving a differential equation,
or solving the issues of social inequality, and someone suggests “Why don’t you try it this way. It ought to be much easier than what you seem to be trying.” Such “easy answers” do presume to provide " some prior knowledge of the consequence of
reorganisation before it happens", except that no reorganization is required.

VH: So an “easy answer” would be control via existing organisation of CSs rather than reorganisation?

MT: “Trying it this way” uses no novel control linkages, but it does imply a controlled perception of linkage change. You give up trying to hang the laundry on an inadequate number of chairs arranged about the lawn, but hang them on a line in the sun. That’s
a perceptually controlled linkage change, sending the output of “perceived clothes dampness” controller to different lower level control units. In the language of atenfels, the perceptual controls involved in putting the clothes over chairbacks are atenfels

VH: I’m going to have to get my head around the atenfels concept so will read your other posts Martin in more detail. Hard to keep up… thanks for your help so far.

that go from active to potential, while those involved with setting up and using a clothesline go from potential to active.

Changing the atenfels used in control of a perception (using “prior knowledge of the consequences”, even if that “knowledge” turns out to be wrong) seemed to me to have much in common with
reorganization that produces new structure in the hierarchy. Both potentially offer “easy” and “hard” directions of change. The demagogue offers “prior knowledge of consequences” that are probably deceptive, but are used by the gullible to change the atenfels
for controlling perceptions they have been trying unsuccessfully to control (“our country’s greatness”, for example). The helpful teacher offers “prior knowledge of consequences” that is probably reasonably reliable. The well-meaning friend offers prior knowledge
that has a dubious likelihood of turning out to be correct.

The reorganization system has no obvious way of finding easy directions in preference to “Rube Goldberg” (Heath Robinson) solutions to problems of poor control, but I have faith that evolution is likely to have provided us with some mechanism to do it.

Martin