signal units and units of physical units

[From Bjorn Simonsen (2010.12.03, 13:50)]

I think I have made wrong models, when I have tried to make PCT models with three (or more) levels. And maybe you have done the same mistakes Rick. I think upon your hier.exl.

My thoughts follow what you see in the figure below.

PCSP-I2.GIF

The p, perceptual signals, going from one level to the level above are nerve signals (BCP page 286) p = k(i) * q(i).

The q(o), has the units of physical units per signal unit and I find that OK at Level 1 where the q(o) goes to muscles and organs.

My question is this. Should there more correct og a line from e (error is a nerve signal) from a level to the level below? The r (reference signal) is a nerve signal.

If you Rick go to your hier.exl, you see that your reference signals are from the line O above. Maybe you are correct, then you must tell me the way I think wrong.

bjorn

[From Rick Marken (2010.12.03.1125)]

Bjorn Simonsen (2010.12.03, 13:50)–

My question is this. Should there more correct og a line from e (error is a nerve signal) from a level to the level below? The r (reference signal) is a nerve signal.

Sorry, Bjorn, I don’t understand this.

If you Rick go to your hier.exl, you see that your reference signals are from the line O above. Maybe you are correct, then you must tell me the way I think wrong.

I don’t know if I understand what you are thinking. But let me take a stab at it. The outputs (O cells) at each level in hier.exl are the output signals (efferent neural signals) from each control system at that level. The reference signals (the R cells in the systems below the O cells at each level, except level 1) are the properly combined outputs from the relevant O cells in the next level up; properly combined in the sense that an output is combined into a lower level reference only if the perception from that lower level system contributes to the perception of the higher level system from whence that output comes, and with the opposite sign.

The architecture of my hierarchical model is exactly equivalent to that described in one of Bill’s Byte articles. Ah, I found it at Dag’s site: http://www.livingcontrolsystems.com/enclosures/byte_aug_1979.pdf. Bill describes a two level model, implemented in Basic. I made it a three level model, originally in Lotus 123 with a couple more systems at each level.

Best

Rick

···


Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

[From Bjorn Simonsen (2010. 12.03, 22:55 EU ST)]

[From Rick Marken (2010.12.03.1125)]

I don’t know if I understand what you are thinking. But let me take a stab at it. The outputs

(O cells) at each level in hier.exl are the output signals (efferent neural signals) from each

control system at that level.

If that is correct, you have answered my problem which was not a problem. But…

Let me take an example from your hier.exl.

**System **


1
2
3
4
5
6
Average
Level
*=imagine

Error
Slowing
Gain
R(3,i)
1,00
1,00
1,00
1,00
1,00
1,00
3
P(3,i)
-1,00
-1,00
1,00
1,00
1,00
-1,00
1,000
Error
0,0001
1000
O(3,i)
30,60
26,94
23,05
32,00
39,19
49,31
*=imagine

R(2,i)
10,00
6,00
-3,89
8,95
7,19
10,12
2
P(2,i)
7,99
7,99
8,55
-1,62
-2,65
-3,16
8,358
Error
0,0001
500
O(2,i)
14,77
-6,82
-49,70
48,90
44,02
51,26
*=imagine
R(1,i)
7,95
1,50
4,63
-13,55
0,09
15,99
1
P(1,i)
7,99
2,34
5,07
-12,82
0,31
14,79
0,579
Error
0,01
50
O(1,i)
-2,01
9,86
-21,02
-28,56
-11,16
18,96
System


CV
Q(i)
7,99
2,29
5,07
-12,89
0,43
14,97
3,312

Disturbance
D(i)
10,00
-7,09
26,09
15,62
11,66
-4,24
Behavior

-33,93
**Disturbance **
Cycle
20

Level 2 R(2,6) 10,12. The formula here is I5 - H5 (49.31 - 39.19). Or as you say:

The reference signals (the R cells in the systems below the O cells at each level, except level

  1. are the properly combined outputs from the relevant O cells in the next level up; properly

combined in the sense that an output is combined into a lower level reference only if the

perception from that lower level system contributes to the perception of the higher level system

from whence that output comes, and with the opposite sign.

But if you read BCP page 286, you see

  1. q(o) = k(o)* e and six lines above, k(o) has units of (other) physical units per signal unit. How can q(o) be:

The outputs (O cells) at each level in hier.exl are the output signals (efferent neural signals) from each

control system at that level.

I think e is an efferent neural signal and when you multiply it with k(o) which has units of (other) physical units per signal unit it can’t still be a neural signal.

Of course I know I think the wrong way. and you the correct way. But I can’t see where I think wrong.

bjorn

[Martin Taylor 2010.112.03.17.11]

      [From Bjorn Simonsen (2010.

12.03, 22:55 EU ST)]

[From Rick Marken (2010.12.03.1125)]

      >I don't know if I understand

what you are thinking. But let me take a stab at it. The
outputs

      >(O cells) at each level in

hier.exl are the output signals (efferent neural signals) from
each

control system at that level.

      If that is correct, you have

answered my problem which was not a problem. But…


3. q(o) = k(o)* e and six
lines above, k(o) has units of (other) physical units per
signal unit. How can q(o) be:

      >The outputs (O cells) at each

level in hier.exl are the output signals (efferent neural
signals) from each

control system at that level.

      I think e is an efferent neural

signal and when you multiply it with k(o) which has units of
(other) physical units per signal unit it can’t still be a
neural signal.

      Of course I know I think the

wrong way. and you the correct way. But I can’t see where I
think wrong.

The problem is that the neural signals are pure numbers or have

dimensions sec-1, depending on whether you treat them as signal
values or as pulses per second. The applies to every signal other
than those that cross the barrier defined by the sensor systems and
the muscular outputs. Sensor systems transduce environmental units
such as photons/sec or degrees K into neural pulse rates, but the
transform ratio changes depending on past history and signals in
other transducers, so there’s no defined ratio anywhere. Weight is
sensed as muscle stretch, which is a pure number
metres_stretch/metres_muscle_fibre_length.

When you come to think of the environmental variable that

corresponds to a perceptual function, it’s sometimes hard to say
what physical dimensions apply. It’s easy enough to say that an
acceleration has dimensions metres/sec^2, but what are the
dimensions of “democracy”?

I think that when you are dealing with control systems as entire

loops, it is better to forget the dimensions, and remember only that
if you do want to trace dimensions around the loop, when you get
back to where you started, the dimensionality had better be the
same, because if it isn’t, you’ve done your analysis improperly. In
the environmental feedback path, if you consider that alone, you
must get the dimensions correct, but once you get into the neural
part of the control loops, all you have is either pure numbers or
pulses/sec.

As for the error signal, it is indeed an efferent signal, but in

HPCT, it goes out to only one place, the output function of the
control loop of which it is the error signal. You could (and maybe
should) see the output function as a time-binding transducer of the
error signal that generates the effects on the environment of that
control unit (the environment of any control unit includes all the
lower level control units to which it sends its output signal and
from which it receives inputs to its perceptual input function; the
environment of the organism is what lies outside the sensor/muscle
interface – the same word is used in both senses, and it is
sometimes hard to see what is meant in a particular case).

The question of how output signals from level N control units

combine to create a reference value in any particular level N-1
control unit is as yet an open question. At one point I proposed
that control units have a “Reference Input Function” analogous to
the “Perceptual Input Function”, but I’ve never followed up that
idea to see whether it would work. In B:CP, Bill proposed that the
reference values are remembered values from prior occasions. I like
that idea, because it solves a whole lot of difficult issues with
HPCT. The “Reference Input Function” would be an associative memory
rather than a simple function of its input values. It’s not clear
(to me, anyway), how reorganization would play together with the
associative memory concept, but I think its an area well worth
study.

Martin

[From Rick Marken (2010.12.03.1720)]

Bjorn Simonsen (2010. 12.03, 22:55 EU ST)–

[From Rick Marken (2010.12.03.1125)]

I don’t know if I understand what you are thinking. But let me take a stab at it. The outputs

(O cells) at each level in hier.exl are the output signals (efferent neural signals) from each

control system at that level.

If that is correct, you have answered my problem which was not a problem. But…

Why don’t we just stop at “not a problem”. You do notice that the hierarchy of control systems works, right?

But if you read BCP page 286, you see

  1. q(o) = k(o)* e and six lines above, k(o) has units of (other) physical units per signal unit. How can q(o) be:

The outputs (O cells) at each level in hier.exl are the output signals (efferent neural signals) from each

control system at that level.

I think e is an efferent neural signal and when you multiply it with k(o) which has units of (other) physical units per signal unit it can’t still be a neural signal.

I believe Bill was doing an analysis of a control system that is closed through the environment. Higher level control systems are closed through the environment but their outputs must go down though the lower level systems first. So for control systems above the lowest level, k(o) has units of signal units/signal units rather than physical units/signal units.

Best

Rick

···


Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

[From Bjorn Simonsen (2010. 12.03, 22:55 EU ST)]

[Martin Taylor 2010.112.03.17.11]

I am sorry for answering so late, my day’s form is variable. Today I am fine and I liked to read and reread your comments.

MT

···

The problem is that the neural signals are pure numbers or have dimensions sec-1, depending on whether you treat them as signal values or as pulses per second. The applies to every signal other than those that cross the barrier defined by the sensor systems and the muscular outputs. Sensor systems transduce environmental units such as photons/sec or degrees K into neural pulse rates, but the transform ratio changes depending on past history and signals in other transducers, so there’s no defined ratio anywhere. Weight is sensed as muscle stretch, which is a pure number metres_stretch/metres_muscle_fibre_length.
bjorn

I thought Weight is sensed as muscle shortening, a pure number metres_shortening/metres_muscle_fibre_lengt. I thought muscles only stretch because other muscles are shortening???

MT
When you come to think of the environmental variable that corresponds to a perceptual function, it’s sometimes hard to say what physical dimensions apply. It’s easy enough to say that an acceleration has dimensions metres/sec^2, but what are the dimensions of “democracy”?

bjorn

Maybe I slip out now. I am not sure I think that neither accelaration nor “democracy” are environmental variables. I think I look at an object moving in front of me with an increasing speed. This is a perception. My awareness of this perception start thoughts about the perceptions. These thoughts are inside my brain generated perceptions, they are not sensed. They are imagined. I test these thoughts relative to sensed environmental variables and I find that some thoughts about increased speed are confirmed and my thoughts are described as acceleration thoughts. Giving them the dimension metres/sec^2 is another inside my brain generated thoughts. They are not environmental variables. ----What do you say??

TM
I think that when you are dealing with control systems as entire loops, it is better to forget the dimensions, and remember only that if you do want to trace dimensions around the loop, when you get back to where you started, the dimensionality had better be the same, because if it isn’t, you’ve done your analysis improperly. In the environmental feedback path, if you consider that alone, you must get the dimensions correct, but once you get into the neural part of the control loops, all you have is either pure numbers or pulses/sec.

bjorn

Good words

As for the error signal, it is indeed an efferent signal, but in HPCT, it goes out to only one place, the output function of the control loop of which it is the error signal. You could (and maybe should) see the output function as a time-binding transducer of the error signal that generates the effects on the environment of that control unit (the environment of any control unit includes all the lower level control units to which it sends its output signal and from which it receives inputs to its perceptual input function; the environment of the organism is what lies outside the sensor/muscle interface – the same word is used in both senses, and it is sometimes hard to see what is meant in a particular case).

bjorn
OK, but I am not sure I understand you the same well you understad yourself. I will work with your word “time-binding-transduser”.

MT
The question of how output signals from level N control units combine to create a reference value in any particular level N-1 control unit is as yet an open question. At one point I proposed that control units have a “Reference Input Function” analogous to the “Perceptual Input Function”, but I’ve never followed up that idea to see whether it would work. In B:CP, Bill proposed that the reference values are remembered values from prior occasions. I like that idea, because it solves a whole lot of difficult issues with HPCT. The “Reference Input Function” would be an associative memory rather than a simple function of its input values. It’s not clear (to me, anyway), how reorganization would play together with the associative memory concept, but I think its an area well worth study.

bjorn

I liked this section and I liked special the “Reference input Function”.

ty for good words Martin

[Bjorn Simonsen (2010. 12.06, 11:30 EU ST)]

[From Rick Marken (2010.12.03.1720)]

RM

I believe Bill was doing an analysis of a control system that is closed through the environment. Higher level control systems are closed through the environment but their outputs must go down though the lower level systems first. So for control systems above the lowest level, k(o) has units of signal units/signal units rather than physical units/signal units.
bjorn

Thinking this way is OK for me. ty for comments.

bjorn

Sorry for sending
[From Bjorn Simonsen (2010. 12.03, 22:55 EU ST)]

[Martin Taylor 2010.112.03.17.11]

it should be
[From Bjorn Simonsen (2010. 12.06, 11:20 EU ST)]

[Martin Taylor 2010.112.03.17.11]

bjorn

[Martin Taylor 2010.12.06.10.12]

Something went wrong with the quoting mechanism in your reply. I'll

try to use [MT now] instead of the succession of quoting marks.

        [From Bjorn Simonsen (2010.

12.03, 22:55 EU ST)]

[Martin Taylor 2010.112.03.17.11]

      I am sorry for answering so late,

my day’s form is variable. Today I am fine and I liked to read
and reread your comments.

[MT now] I hope you continue to be fine!

    MT
      The problem is that the neural signals are pure numbers or

have dimensions sec-1, depending on whether you treat them as
signal values or as pulses per second. The applies to every
signal other than those that cross the barrier defined by the
sensor systems and the muscular outputs. Sensor systems
transduce environmental units such as photons/sec or degrees K
into neural pulse rates, but the transform ratio changes
depending on past history and signals in other transducers, so
there’s no defined ratio anywhere. Weight is sensed as muscle
stretch, which is a pure number
metres_stretch/metres_muscle_fibre_length.

bjorn

        I thought Weight is sensed as

muscle shortening, a pure number
metres_shortening/metres_muscle_fibre_lengt. I thought
muscles only stretch because other muscles are shortening???

[MT now] You may well be right. The point is that the units for the

neural system are not the units in the environment, but are either
pure numbers throughout or sec-1 (pulses/second) throughout. What is
sensed is not the same dimension as what is output to the
environment, either. Apart from chemical emissions, the outputs to
the environment are all in the form of mechanical force, even though
the internal physical representation is
delta_muscle_length/muscle_length (a pure number). The outputs to
the environment may cause effects that can be seen. For visible
things, the input is variation of photon arrival rates on different
parts of the retina, and force does not enter the equation.

MT

      When you come to think of the environmental variable that

corresponds to a perceptual function, it’s sometimes hard to
say what physical dimensions apply. It’s easy enough to say
that an acceleration has dimensions metres/sec^2, but what are
the dimensions of “democracy”?

bjorn

        Maybe I slip out now. I am not

sure I think that neither accelaration nor “democracy” are
environmental variables.

[MT now] Are they not? You control your perception of them by acting

on the environment, don’t you? How could you do that if they were
not environmental variables?

I think I look at an object
moving in front of me with an increasing speed. This is a
perception.

[MT now] So if that's a controlled perception, how do you influence

it? By acting on something in the environment, don’t you? Maybe you
physically touch the object, maybe you turn a knob or step on a
brake pedal. Whatever you do affects the environment by applying
some force, with the result that your perception changes of how
rapidly the speed is increasing.

My awareness of this
perception start thoughts about the perceptions. These
thoughts are inside my brain generated perceptions, they are
not sensed. They are imagined. I test these thoughts
relative to sensed environmental variables and I find that
some thoughts about increased speed are confirmed and my thoughts are
described as acceleration thoughts. Giving them the
dimension metres/sec^2 is another inside my brain generated
thoughts. They are not environmental variables. ----What
do you say??

I say your imaginings and thoughts about your perceptions are just

that – imaginings about your perceptions. We have to be careful
when we deal with consciousness. Consciousness is a bit of a
mystery, to put it mildly! How you analyze where your perceptions
come from is not the same as where your perceptions come from, even
if your analysis is correct; “the map is not the territory”.

      As for the error signal, it is indeed an efferent signal, but

in HPCT, it goes out to only one place, the output function of
the control loop of which it is the error signal. You could
(and maybe should) see the output function as a time-binding
transducer of the error signal that generates the effects on
the environment of that control unit (the environment of any
control unit includes all the lower level control units to
which it sends its output signal and from which it receives
inputs to its perceptual input function; the environment of
the organism is what lies outside the sensor/muscle interface
– the same word is used in both senses, and it is sometimes
hard to see what is meant in a particular case).

bjorn

        OK, but I am not sure I understand you the same well you

understad yourself. I will work with your word
“time-binding-transduser”.

[MT now] "Time-binding" means that it may incorporate some of the

history of its input into its present output. An integrator such as
the classic HPCT output function is a time-binding transducer
because its output sums up all the previous input (apart from
leakage). A sensor that adapts to the intensity of input is a
time-binding transducer, because its output for an input of 1 unit
is different if the recent inputs have all been around 10 units than
if the resent inputs have all been around 0.1 units.

ty for good words Martin

I hope to help more than I confuse!

Martin

( Gavin
Ritz 2010.12.07.9.48NZT)

[Martin Taylor
2010.12.06.10.12]

[From Bjorn Simonsen (2010.
12.03, 22:55 EU ST)]

[Martin Taylor
2010.112.03.17.11]

[MT] Are they not? You control your perception of them by acting on the
environment, don’t you? How could you do that if they were not environmental
variables?

Martin

You say one is acting on the environment,
is it not selection of Perceptual Controlled Variables to control perceptions. “Acting
on the environment” sounds like trying to control the environment to me
rather than control perceptions.

Regards

Gavin

···

[Martin Taylor 2010.12.06.16.12]

(Gavin Ritz 2010.12.07.9.48NZT)

[ Martin Taylor
2010.12.06.10.12]

[From Bjorn Simonsen (2010.
12.03, 22:55 EU ST)]

[ Martin Taylor
2010.112.03.17.11]

          [MT] Are they not? You control your perception of them by

acting on the
environment, don’t you? How could you do that if they were
not environmental
variables?

Martin

          You

say one is acting on the environment,
is it not selection of Perceptual Controlled Variables to
control perceptions. “Acting
on the environment” sounds like trying to control the
environment to me
rather than control perceptions.

Regards

Gavin

Your question confuses me. How else do you control a perceptual

signal value other than by acting on the environment?

What do you mean by "selection of Perceptual Controlled Variables to

control perceptions". I can derive no meaning out of that. It’s like
“twas brillig and the slithy toves did gyre an gimbal in the wabe”.
You think it ought to mean something, but what could that something
be?

Martin

( Gavin
Ritz 2010.12.07.10.31NZT)

[Martin Taylor
2010.12.06.16.12]

(Gavin
Ritz 2010.12.07.9.48NZT)

[Martin Taylor
2010.12.06.10.12]

[From Bjorn Simonsen (2010.
12.03, 22:55 EU ST)]

[Martin Taylor
2010.112.03.17.11]

[MT] Are they not? You control your perception of them by acting on the
environment, don’t you? How could you do that if they were not environmental
variables?

Martin

You say
one is acting on the environment, is it not selection of Perceptual Controlled
Variables to control perceptions. “Acting on the environment”
sounds like trying to control the environment to me rather than control
perceptions.

Regards

Gavin

Your question confuses me. How else do you control a perceptual signal
value other than by acting on the environment?

Are you
saying then if you act on the environment you are controlling the environment
or the perception.

Where
does the PCV come into the equation?

Are you
saying the PCV and the environment is one and the same?

What do you mean by “selection of Perceptual Controlled Variables to
control perceptions”. I can derive no meaning out of that. It’s like
“twas brillig and the slithy toves did gyre an gimbal in the wabe”.
You think it ought to mean something, but what could that something be?

[Martin Taylor 2010.12.06.17.45]

(Gavin Ritz 2010.12.07.10.31NZT)

[ Martin Taylor
2010.12.06.16.12]

          (Gavin

Ritz 2010.12.07.9.48NZT)

[ Martin Taylor
2010.12.06.10.12]

[From Bjorn Simonsen (2010.
12.03, 22:55 EU ST)]

[ Martin Taylor
2010.112.03.17.11]

          [MT] Are they not? You control your perception of them by

acting on the
environment, don’t you? How could you do that if they were
not environmental
variables?

Martin

          You say

one is acting on the environment, is it not selection of
Perceptual Controlled
Variables to control perceptions. “Acting on the
environment”
sounds like trying to control the environment to me rather
than control
perceptions.

Regards

Gavin

          Your question

confuses me. How else do you control a perceptual signal
value other than by acting on the environment?

          Are you

saying then if you act on the environment you are
controlling the environment
or the perception.

The perception. It's standard PCT. Perceptions are the only

variables you can control.

          Where

does the PCV come into the equation?

It's the controlled quantity, if by "PCV" you mean the value of the

perceptual signal. If that’s not what you mean, you had better
define PCV.

          Are you

saying the PCV and the environment is one and the same?

The perceptual signal value is a function of data from the sensors

of the environment (possibly along with imagined components). It is
in the brain. The environment is out there. It is not in the brain.

Martin

( Gavin
Ritz 2010.12.07.11.57NZT)

[Martin Taylor
2010.12.06.17.45]

(Gavin
Ritz 2010.12.07.10.31NZT)

[Martin Taylor 2010.12.06.16.12]

(Gavin
Ritz 2010.12.07.9.48NZT)

[Martin Taylor
2010.12.06.10.12]

[From Bjorn Simonsen (2010. 12.03,
22:55 EU ST)]

[Martin Taylor
2010.112.03.17.11]

[MT] Are they not? You control your perception of them by acting on the
environment, don’t you? How could you do that if they were not environmental
variables?

Martin

You say
one is acting on the environment, is it not selection of Perceptual Controlled
Variables to control perceptions. “Acting on the environment”
sounds like trying to control the environment to me rather than control
perceptions.

Regards

Gavin

Your
question confuses me. How else do you control a perceptual signal value other
than by acting on the environment?

Are you saying then if you act on the environment
you are controlling the environment or the perception.

The perception. It’s standard PCT. Perceptions are the only variables you can
control.

Ok so what’s happened to the Perceptual
Controlled Variable?

Where does the PCV come into the equation?

It’s the controlled quantity,

Is the Perceptual Controlled Variable, now
the controlled quantity in your opinion?

if by “PCV” you
mean the value of the perceptual signal. If that’s not what you mean, you had
better define PCV.

Perceptual Controlled Variable (PCV)

Are you saying the PCV and the environment is one
and the same?

The perceptual signal value is a function of data

Data, is
this a new unit, where does this data come from?

from the sensors of the environment

Where
are these “sensors of the environment”? What are they?

(possibly along with imagined components).

So what’s
happened to the PCV? Perceptual Controlled Variable.

It is in the brain.

So is
the PCV now in the brain?

So when
you transduce sound waves (from the environment) to electrical signals in the
neural structure-process are you controlling that.

The environment is out there. It is not in the brain.

Fair
enough.

Gavin

Martin

[Martin Taylor 2010.12.06.23.35]

(Gavin Ritz 2010.12.07.11.57NZT)

[ Martin Taylor
2010.12.06.17.45]

          (Gavin

Ritz 2010.12.07.10.31NZT)

[ Martin
Taylor 2010.12.06.16.12]

          (Gavin

Ritz 2010.12.07.9.48NZT)

[ Martin Taylor
2010.12.06.10.12]

[From Bjorn Simonsen (2010. 12.03,
22:55 EU ST)]

[ Martin Taylor
2010.112.03.17.11]

          [MT] Are they not? You control your perception of them by

acting on the
environment, don’t you? How could you do that if they were
not environmental
variables?

Martin

          You say

one is acting on the environment, is it not selection of
Perceptual Controlled
Variables to control perceptions. “Acting on the
environment”
sounds like trying to control the environment to me rather
than control
perceptions.

Regards

Gavin

          Your

question confuses me. How else do you control a perceptual
signal value other
than by acting on the environment?

          Are you saying then if you act on the

environment
you are controlling the environment or the perception.

          The perception. It's standard PCT. Perceptions are the

only variables you can
control.

          Ok

so what’s happened to the Perceptual
Controlled Variable?

I have heard this name only from you, so I don't know. You tell me.

Where does the PCV come into the equation?

          It's the controlled quantity,
          Is

the Perceptual Controlled Variable, now
the controlled quantity in your opinion?

          if by "PCV" you

mean the value of the perceptual signal. If that’s not
what you mean, you had
better define PCV.

          Perceptual

Controlled Variable (PCV)

it would be nice if you would define PCV, instead of repeating the

expansion of the acronym.

You say it is an acronym for "Perceptual Controlled Variable", but

apparently you don’t mean the perceptual signal value, which is the
controlled variable in Perceptual Control Theory. A definition would
be helpful for understanding.

          Are you saying the PCV and the environment

is one
and the same?

          The

perceptual signal value is a function of data

          Data, is

this a new unit, where does this data come from?

See the next line
          from the

sensors of the environment

          Where

are these “sensors of the environment”? What are they?

Eyes, ears, tastebuds, ...
          (possibly

along with imagined components).

          So what’s

happened to the PCV? Perceptual Controlled Variable.

          It is in the

brain.

          So is

the PCV now in the brain?

The perception that is controlled is of course in the brain. It's

the only place it could possibly be. That’s what Bill P. may have
been the first to notice, half a century ago. It’s why the book was
“Behaviour: the Control of Perception”, not “Perception, the control
of behaviour”. Where do you think it is?

···
          So when

you transduce sound waves (from the environment) to
electrical signals in the
neural structure-process are you controlling that.

          The

environment is out there. It is not in the brain.

          Fair

enough.

Gavin

        Martin

( Gavin
Ritz 2010.12.07.18.44NZT)

[Martin Taylor 2010.12.06.23.35]
(Gavin Ritz 2010.12.07.11.57NZT)

[Martin Taylor 2010.12.06.17.45]

( Gavin
Ritz 2010.12.07.10.31NZT)

[ Martin
Taylor 2010.12.06.16.12]

( Gavin
Ritz 2010.12.07.9.48NZT)

[Martin Taylor 2010.12.06.10.12]

[From Bjorn Simonsen (2010.
12.03, 22:55 EU ST)]

[Martin Taylor 2010.112.03.17.11]

[MT] Are they not? You control your perception of them by acting on the
environment, don’t you? How could you do that if they were not environmental
variables?

The perception. It’s standard PCT. Perceptions are the only variables you can
control.

Ok so
what’s happened to the Perceptual Controlled Variable?

I have heard this name only from you, so I don’t know. You tell me.

It’s the variable the individual
controls for. I certainly didn’t make up this name for the variable.

We just simply select for the Perceptual
Controlled variable.

Not sure how you can say you’ve never
heard of it. You are pulling my leg I think.

My point was I didn’t get your
comment “acting on the environment”. We don’t act on
the environment.

See:

http://www.responsiblethinking.com/tom/PCTFacts/CV.htm

“In the PCT model, but not in any conventional
model, there is a controlled variable between the apparent
cause and its apparent effect as usually observed. The behavior is directed
toward control
of the controlled
variable
; it is not simply a response to the stimulus or a
dependent variable being determined by an independent variable.” Bill Powers.

From the PCT web site here called perceptual variable but the same as the controlled variable. (or Perceptual
Controlled Variable)

What are the meanings of
“perceptual” and “control” in PCT?

“Perception”
in PCT refers to the world as we experience it. The dimensions of our
experience – the colors, shapes, movements, relationships and so on that we
experience as aspect of the world we live in – are called perceptual variables. Perceptual variables
are what we control.
Perceptual variables, such as the color of a house, can
have many different possible values, such as white, yellow, green, etc. Control
refers to the process of bringing a perceptual variable to a pre-selected state and maintaining it in that state, protected from disturbances.
When you decide to paint your house white and then do it (or have someone do it
for you) you are controlling the perceived color of your house

···

[Martin Taylor 2010.12.17.10.45]

(Gavin Ritz 2010.12.07.18.44NZT)

          We

just simply select for the Perceptual
Controlled variable.

Explain how thisidea is consistent with HPCT. I grant there could be

a version of PCT in which this holds, but as I understand HPCT there
is no such selection. There are lots of perceptual variables being
controlled at any moment in HPCT.

          Not

sure how you can say you’ve never
heard of it. You are pulling my leg I think.

Not so. I have always heard "Perceptual signal", "Perception",

“Perceptual signal value” and the like used to talk about the
variable controlled by any single control unit. The word string
“Perceptual Controlled Variable” is redundant, since ALL controlled
variables are perceptual.

Nevertheless, I Googled for "Perceptual Controlled Variable". I

found only one link, a message sent by you 2010.10.20 to the
CHAOPSYC Archives. So it seems that you are the authority for the
fact that the label “Perceptual Controlled Variable” has been used
elsewhere. I quote from your message (grammar exactly as written):

"The organism basically selects for "a Perceptual Controlled

Variable" which is really dictated almost entirely by an inner
signal. So an environmental disturbance doesn’t play such an
enormous role like in a stimulus-response model does."

What I gather from this is that you think that the organism is

selecting one variable to control, and that an environmental
disturbance doesn’t affect behaviour. Neither of those beliefs is
consistent with HPCT, and the second one is inconsistent with any
form of PCT that I can think of.

The first claim is inconsistent with HPCT because in HPCT a whole

multitude of variables is being controlled simultaneously. Why is
the second one inconsistent with PCT? Because in order to keep the
controlled variable (the perception) near its reference value, the
organism must counter the disturbance by acting on the environment
in a way that depends on the disturbance. If the reference value for
a controlled perception is constant, the effect observable from
outside the organism is indistinguishable from stimulus-response. In
PCT that’s called “the behavioural illusion”.

          My

point was I didn’t get your
comment “acting on the environment”. We don’t act on
the environment.

That's interesting. How did you get your message into your computer?

How do you manage to eat? Are you actually a virtual person manifest
as a product of the Web?

See:

http://www.responsiblethinking.com/tom/PCTFacts/CV.htm

          “In the PCT model,

but not in any conventional
model, there is a ** controlled
variable** between the apparent
cause and its apparent effect as usually observed. The
behavior is directed
toward control
of the ** controlled
variable** ; it is not simply a response
to the stimulus or a
dependent variable being determined by an independent
variable.” Bill Powers.

Exactly so. The controlled variable for each individual control

element is called a perception in PCT.

From the PCT web site here called ** perceptual
variable** but the same as the ** controlled
variable**. ** (or
Perceptual
Controlled Variable)**

          What

are the meanings of
“perceptual” and “control” in PCT?

          "Perception"

in PCT refers to the world as we experience it. The
dimensions of our
experience – the colors, shapes, movements, relationships
and so on that we
experience as aspect of the world we live in – are called
perceptual variables. ** Perceptual
variables
are what we control.** ** Perceptual
variables** , such as the color of a house,
can
have many different possible values, such as white,
yellow, green, etc. Control
refers to the process of bringing a ** perceptual
variable** to a pre-selected state and
maintaining it in that state, protected from disturbances.
When you decide to paint your house white and then do it
(or have someone do it
for you) you are controlling the perceived color of your
house

Exactly so. That is a good description of the foundation of PCT.

Bringing a perceptual variable to a pre-selected state is not the
same as selecting for a perceptual controlled variable.

Martin

( Gavin
Ritz 2010.12.08.10.20NZT)

[Martin Taylor
2010.12.17.10.45]

(Gavin
Ritz 2010.12.07.18.44NZT)

Not so. I have always heard “Perceptual signal”,
“Perception”, “Perceptual signal value” and the like used
to talk about the variable controlled by any single control unit. The word
string “Perceptual Controlled Variable” is redundant, since ALL
controlled variables are perceptual.

Okay if that’s what your opinion is,
then the Perceptual Controlled variable is redundant in your opinion.

What I gather from this is that you think that the organism is selecting
one variable to control, and that an environmental disturbance doesn’t affect
behaviour. Neither of those beliefs is consistent with HPCT, and the second one
is inconsistent with any form of PCT that I can think of.

The first claim is inconsistent with HPCT because in HPCT a whole multitude of
variables is being controlled simultaneously.

I never once said that there is only one
PCV, in fact if you look at the earlier thread I mentioned Controlled
variables, so this argument is redundant.

Why is the second one
inconsistent with PCT? Because in order to keep the controlled variable (the
perception)

So are you now saying there is such a
thing as a controlled variable?

My point
was I didn’t get your comment “acting on the environment”.
We don’t act on the environment.

That’s interesting. How did you get your message into your computer? How do you
manage to eat? Are you actually a virtual person manifest as a product of the
Web?

I think
this comment really rounds out your opinion on Perceptual “controlled
variables”
Martin**. Now I see why you get so frustrated with things being said.**

The short answer is I don’t act on
the environment at all. This is what the concept would be if you are using a stimulus-response model.

Martin, I unfortunately have to add that you seem to have missed one of
the really finer points of PCT. I don’t want to dialogue with you about
any other aspect of PCT other than the “Controlled Variable” and what
it means.

I understand your dilemma I battled with
this concept too, once you release your preconceived notion of acting on the environment
and just embrace the concept of the Perceptual
Controlled variable
then your frustration will
be released.

You are having a massive internal battle
on this issue believe me it’s the central problem of explaining PCT to a
wider audience. It’s like saying to a fish the medium you live in is
water.

I have spoken with many other listers (non
PCTERS) about PCT all have exactly the same problem.

Imagine this, don’t think or talk
about acting, don’t talk or
think about behavior these words
only confuse the issue.

Then imagine this you are only a
transducer nothing you take into your body (except food and gas) is actually
going into your body. It’s only a mirage that has been transduced. Your
brain doesn’t see feel, taste, hear the environment it only transduces
that stuff coming in.

Can you imagine this?

Then what would you call that stuff you
have just transduced? (Remember it’s not the environment)

Kind regards

Gavin

···

[Martin Taylor 2010.12.07.17.06]

(Gavin Ritz 2010.12.08.10.20NZT)

I had had a forlorn hope that you might somehow come around to

seeing PCT (perceptual control theory) as being the theory that
living beings control perceptions, but it appears that the concept
of “invincible ignorance” applies.

I resign the attempt.

Martin

( Gavin
Ritz 2010.12.08.11.34NZT)

[Martin Taylor
2010.12.07.17.06]

(Gavin
Ritz 2010.12.08.10.20NZT)

I had had a forlorn hope that you might somehow come around to seeing
PCT (perceptual control theory) as being the theory that living beings control
perceptions, but it appears that the concept of “invincible
ignorance” applies.

Hey
buddy I actually do see it very clearly. And it does explain things very well
in human discourse just not the way you explain it.

I don’t
believe that that I select for an invincible ignorant. I think that an invincible
ignorant is your controlled variable as you selected it. Be careful what you
select it shows your inner world.

It has
been these types of comments on the list that destroys trust between conversers;
you have been part of that Martin. Personal accountability and responsibility is paramount if PCT is
to go forward.

I resign the attempt.

I’m
sorry you feel that way Martin

Regards

Gavin

Martin