[From Adam Matic 2014.03.17 1910 cet]
This is how Bill puts it in LCSIII, chapter 4, page 77, last paragraph:
“Tracking is just one form of control: it’s done by controlling a visual relationship, one part of which varies independently (the target), and the other part of which is affected by the participant’s actions.”
Adam
[Martin Taylor 2014.03.19.14.09]
Let's be sure we are talking about the same thing. Most of the time
when a symbol such as Qo or q.o is used, it is used to represent the
value of the output variable from an elementary control unit. Is
this what you mean?
If it is, then do you agree that the HPCT structure is connected
like this, an example in which someone is ringing a doorbell:
I guess that’s enough questions for this message.
By the way, this lower diagram was explicitly approved by Bill
Powers.
Martin
[From Adam Matic 2014.03.17 1830]
> Martin Taylor 2014.03.19.12.45
How would you model“varying [one variable] in order to control [a second
variable]” without actually controlling [one variable]?
It might work open loop, but as Bill was never tired of
reminding us, open-loop doesn’t work very well if there
are disturbances. In multi-level control, varying one
variable in order to control another is usually done by
sending a continuously changing reference value to the
controller of the second variable.
AM:
In every control loop, Qo is varied, not controlled.So, I would say every control loop controls p by varying
Qo, or several Qo’s.No open loop.
[Martin Taylor 2014.03.19.14.30]
[From Adam Matic 2014.03.17 1910 cet]
This is how Bill puts it in LCSIII, chapter 4, page 77, last paragraph:
"Tracking is just one form of control: it's done by controlling a visual relationship, one part of which varies independently (the target), and the other part of which is affected by the participant's actions."
Adam
Right, he mentions one uncontrolled perception and one controlled perception, which is characteristic of pursuit tracking. And how does that paragraph continue? I leave that to the reader.
He's contrasting the example of pursuit tracking you did in Chapter 4 against what you did in Chapter 2, controlling shape, orientation or position, is he not? In Chapter 2 you were asked to choose one of those aspects and keep it at a reference value you chose. That's compensatory tracking. It doesn't mean that the concept or even the functional structure of control is different.
Martin
[From Adam Matic 2014.03.19]
As far as diagrams go, sure, it looks fine.
Adam
On Wed, Mar 19, 2014 at 7:29 PM, Martin Taylor mmt-csg@mmtaylor.net wrote:
[Martin Taylor 2014.03.19.14.09]
Let's be sure we are talking about the same thing. Most of the timewhen a symbol such as Qo or q.o is used, it is used to represent the
value of the output variable from an elementary control unit. Is
this what you mean?If it is, then do you agree that the HPCT structure is connectedlike this, an example in which someone is ringing a doorbell:
I guess that's enough questions for this message. By the way, this lower diagram was explicitly approved by BillPowers.
Martin
[From Adam Matic 2014.03.17 1830]
> Martin Taylor 2014.03.19.12.45
How would you model“varying [one variable] in order to control [a second
variable]” without actually controlling [one variable]?
It might work open loop, but as Bill was never tired of
reminding us, open-loop doesn’t work very well if there
are disturbances. In multi-level control, varying one
variable in order to control another is usually done by
sending a continuously changing reference value to the
controller of the second variable.
AM:
In every control loop, Qo is varied, not controlled.
So, I would say every control loop controls p by varying
Qo, or several Qo’s.
No open loop.
[From Adam Matic 2014.03.19]
Nope. He does not mention one uncontrolled perception and one controlled perception. He mentions only one controlled perception and that is the visual relationship. There are no other perceptions in this tracking model.
In chapter 2, controlled perceptions were shape, position or color, all visual variables perceived and controlled by the user. User controlled them by varying cursor position on the side of the screen.
On Wed, Mar 19, 2014 at 7:37 PM, Martin Taylor mmt-csg@mmtaylor.net wrote:
[Martin Taylor 2014.03.19.14.30]
[From Adam Matic 2014.03.17 1910 cet]
This is how Bill puts it in LCSIII, chapter 4, page 77, last paragraph:
“Tracking is just one form of control: it’s done by controlling a visual relationship, one part of which varies independently (the target), and the other part of which is affected by the participant’s actions.”
Adam
Right, he mentions one uncontrolled perception and one controlled perception, which is characteristic of pursuit tracking. And how does that paragraph continue? I leave that to the reader.
He’s contrasting the example of pursuit tracking you did in Chapter 4 against what you did in Chapter 2, controlling shape, orientation or position, is he not? In Chapter 2 you were asked to choose one of those aspects and keep it at a reference value you chose. That’s compensatory tracking. It doesn’t mean that the concept or even the functional structure of control is different.
Martin
[Martin Taylor 2014.03.19.16.14]
Not even the perception of the cursor or of the target? If those are
not perceived, where does the perception of the relationship come
from? Martin
[From Adam Matic 2014.03.19]
Nope. He does not mention one uncontrolled perception andone controlled perception. He mentions only one controlled
perception and that is the visual relationship. There are no
other perceptions in this tracking model.
On Wed, Mar 19, 2014 at 7:37 PM, MartinTaylor mmt-csg@mmtaylor.net
wrote:[MartinTaylor 2014.03.19.14.30]
[From Adam Matic 2014.03.17 1910 cet]
This is how Bill puts it in LCSIII, chapter 4, page77, last paragraph:
"Tracking is just one form of control: it's done bycontrolling a visual relationship, one part of which
varies independently (the target), and the other part
of which is affected by the participant’s actions."Adam
Right, he mentions one uncontrolled perception and onecontrolled perception, which is characteristic of pursuit
tracking. And how does that paragraph continue? I leave that
to the reader.Martin
Hi Adam, Martin and Bruce,
I’m sorry to interrupt. I’d like to ask
one question ? Can we conclude from Bill’s generic diagram that every
perceptual signal is controlled (goes through comparator) ?
Best,
Boris
From: Control Systems Group Network (CSGnet)
[mailto:CSGNET@LISTSERV.ILLINOIS.EDU] On Behalf
Of Martin Taylor
Sent: Wednesday, March 19, 2014
9:17 PM
To: CSGNET@LISTSERV.ILLINOIS.EDU
Subject: Re: Measuring
Input-Output Characteristics of Components of a Closed Loop: Redux
[ Martin
Taylor 2014.03.19.16.14]
[From Adam Matic
2014.03.19]
Nope. He does not mention one uncontrolled perception
and one controlled perception. He mentions only one controlled perception and
that is the visual relationship. There are no other perceptions in this
tracking model.
Not even the perception of the cursor or of the target? If those are not
perceived, where does the perception of the relationship come from?
Martin
On Wed, Mar 19, 2014 at 7:37 PM, Martin Taylor mmt-csg@mmtaylor.net wrote:
[ Martin
Taylor 2014.03.19.14.30]
[From Adam Matic
2014.03.17 1910 cet]This is how Bill puts it in LCSIII, chapter 4, page 77, last paragraph:
“Tracking is just one form of control: it’s done by controlling a visual
relationship, one part of which varies independently (the target), and the
other part of which is affected by the participant’s actions.”Adam
Right, he mentions one uncontrolled perception and one
controlled perception, which is characteristic of pursuit tracking. And how
does that paragraph continue? I leave that to the reader.
Martin
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2014.0.4336 / Virus Database: 3722/7217 - Release Date: 03/19/14
[From Adam Matic 2014.03.19 0940cet]
Martin Taylor 2014.03.19.16.14
Not even the perception of the cursor or of the target? If those are not perceived, where does the perception of the relationship come from?
AM:
This is getting terribly nitpicky, but I really don't think 'cursor position' and 'target position' are perceived by this system in track analyse. In the model's input function they are environmental variables that are used to simulate a visual system and are converted directly to our p signal, the visual relationship between visual objects of cursor and target.
In the human visual system, a whole complex process is happening between environmental cursor position on the screen, target position on the screen and the neural representation of the difference between them.
Positions of objects in the visual system are probably relative, but relative to what? The center of fovea? To other objects? It could be that the visual system is consecutively moving target perception to the center of fovea, and then cursor position to the center of fovea and somehow creating the representation of the distance, then calculating the difference. We are modeling just this difference in the track analyse system.
Adam
[From Adam Matic 2014.03.19 2200 cet]
Hi Adam, Martin and Bruce,
I’m sorry to interrupt. I’d like to ask
one question ? Can we conclude from Bill’s generic diagram that every
perceptual signal is controlled (goes through comparator) ?
AM:
I think I speak for all of us when I say that - no - we can’t conclude that every perceptual signal is controlled. There might be some perceptual signals that are not controlled. My hunch is that lower levels don’t contain such signals, perhaps some higher levels do. There is a weird connection between attention, perception and choosing what to control.
What does it mean mechanically that a perception is not controlled is also not very clear to me. It might be that loop gain is zero.
Adam
[Martin Taylor 2014.03.19.16.53]
No problem about interrupting. The more the merrier.
But they constitute almost
all the perceptual signals you have.
Martin
Hi Adam, Martin andBruce,
I'm sorry tointerrupt. I’d like to ask
one question ? Can we conclude from Bill’s generic
diagram that every
perceptual signal is controlled (goes through
comparator) ?
Best,
Boris
` To answer
your question: No. Bill’s generic diagram refers to
perceptions that are being controlled, and they are a
very small proportion of all the perceptions you have
available at any moment. Other perceptions may be
compared with reference values while not currently
being controlled. I hypothesize that unacceptably
large error signals from these ones are related to
attention and to changes in the set of perceptions
that are being controlled. We might call those
excursions of error signals in uncontrolled
perceptions “alerting” signals, but they aren’t in any
control hierarchy that represents what is being
controlled “right now”.
But by far the largest set of perceptions are neither
being controlled nor being compared with reference
values. They may or may not be available to conscious
awareness and they may or may not `` contribute
to the inputs of perceptual functions that generate perceptions
of the first two kinds.`
From: Control Systems
Group Network (CSGnet)
Wednesday, March 19, 2014
9:17 PM
Re: Measuring
Input-Output Characteristics of Components of a Closed
Loop: Redux
[Martin Taylor
2014.03.19.16.14]
[From Adam Matic2014.03.19]
Nope.He does not mention one uncontrolled perception
and one controlled perception. He mentions only
one controlled perception and
that is the visual relationship. There are no
other perceptions in this
tracking model.
Not even the perception of the cursor or of the target?If those are not
perceived, where does the perception of the relationship
come from?Martin
On Wed,Mar 19, 2014 at 7:37 PM, Martin Taylor
<mmt-csg@mmtaylor.net >
wrote:
[Martin Taylor
2014.03.19.14.30]
[From Adam Matic2014.03.17 1910 cet]
This is how Bill puts it in LCSIII, chapter 4,page 77, last paragraph:
"Tracking is just one form of control: it'sdone by controlling a visual
relationship, one part of which varies
independently (the target), and the
other part of which is affected by the
participant’s actions."Adam
Right,he mentions one uncontrolled perception and one
controlled perception, which is characteristic of
pursuit tracking. And how
does that paragraph continue? I leave that to the
reader.Martin
No virusfound in this message.
Checked by AVG - www.avg.com
Version: 2014.0.4336 / Virus Database: 3722/7217 - Release
Date: 03/19/14
mailto:CSGNET@LISTSERV.ILLINOIS.EDU** On Behalf
Of** Martin Taylor
Sent:
**To:**CSGNET@LISTSERV.ILLINOIS.EDU
Subject:
Hi Martin,
MT :
To answer your question: No. Bill's generic diagram refers to perceptions that are being controlled, and they are a very small proportion of all the perceptions you have available at any moment.
``
HB :
Can you specify which are controlled perceptions
in Bill’s diagram in respect to input function ? And which are all the
perceptions that are available at any moment in respect to input function ?
MT :
But by far the largest set of perceptions are neither being controlled nor being compared with reference values.
``
HB :
Where could these perceptual signals go ? Beside the Bill's diagram ?
``
Best,
``
Boris
``
From: Control Systems Group Network (CSGnet)
[mailto:CSGNET@LISTSERV.ILLINOIS.EDU] On
Behalf Of Martin
Taylor
Sent: Wednesday, March 19, 2014
10:05 PM
To: CSGNET@LISTSERV.ILLINOIS.EDU
Subject: Re: Measuring
Input-Output Characteristics of Components of a Closed Loop: Redux
[ Martin
Taylor 2014.03.19.16.53]
Hi Adam, Martin and Bruce,
I’m sorry to interrupt. I’d like to ask
one question ? Can we conclude from Bill’s generic diagram that every
perceptual signal is controlled (goes through comparator) ?
Best,
Boris
No problem about interrupting. The
more the merrier.
To answer your question: No. Bill's generic diagram refers to perceptions that are being controlled, and they are a very small proportion of all the perceptions you have available at any moment. Other perceptions may be compared with reference values while not currently being controlled. I hypothesize that unacceptably large error signals from these ones are related to attention and to changes in the set of perceptions that are being controlled. We might call those excursions of error signals in uncontrolled perceptions "alerting" signals, but they aren't in any control hierarchy that represents what is being controlled "right now".
But by far the largest set of perceptions are neither being controlled nor being compared with reference values. They may or may not be available to conscious awareness and they may or may not ``contribute to the inputs of perceptual functions that generate perceptions of the first two kinds. But they constitute almost all the perceptual signals
you have.
Martin
From: Control Systems Group
Network (CSGnet) [mailto:CSGNET@LISTSERV.ILLINOIS.EDU]
On Behalf Of Martin Taylor
Sent: Wednesday, March 19, 2014
9:17 PM
To: CSGNET@LISTSERV.ILLINOIS.EDU
Subject: Re: Measuring
Input-Output Characteristics of Components of a Closed Loop: Redux
[Martin Taylor
2014.03.19.16.14]
[From Adam Matic
2014.03.19]
Nope. He does not
mention one uncontrolled perception and one controlled perception. He mentions
only one controlled perception and that is the visual relationship. There are
no other perceptions in this tracking model.
Not even the perception of the cursor or of the target? If those are not
perceived, where does the perception of the relationship come from?
Martin
On Wed, Mar 19, 2014 at 7:37 PM, Martin Taylor mmt-csg@mmtaylor.net wrote:
[Martin Taylor
2014.03.19.14.30]
[From Adam Matic
2014.03.17 1910 cet]This is how Bill puts it in LCSIII, chapter 4, page 77, last paragraph:
“Tracking is just one form of control: it’s done by controlling a visual
relationship, one part of which varies independently (the target), and the
other part of which is affected by the participant’s actions.”Adam
Right, he mentions one
uncontrolled perception and one controlled perception, which is characteristic
of pursuit tracking. And how does that paragraph continue? I leave that to the
reader.
Martin
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2014.0.4336 / Virus Database: 3722/7217 - Release Date: 03/19/14
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2014.0.4336 / Virus Database: 3722/7217 - Release Date: 03/19/14
Hi Adam,
AM :
I think I speak for all of us when I say that - no - we can’t conclude
that every perceptual signal is controlled. There might be some perceptual
signals that are not controlled.
HB :
You seemed not to be sure J, but still your first answer was
quite categorical
AM :
My hunch is that lower levels don’t contain such signals, perhaps some
higher levels do. There is a weird connection between attention, perception and
choosing what to control.
HB :
Did I understood you right, that lower level does not contain
“uncontrolled perception”, and perhaps some higher levels do ? Is the
emphasize on “perhaps” or you are sure about that, because your first
answer was categorical “no”, what I interpreted that in any kind of
Bill’s diagram, we can’t conclude " that every perceptual signal is
controlled…But according to what you said we can conclude from Bill’s diagrams
that perceptions on lower levels are all controlled ?
AM :
What does it mean mechanically that a perception is not controlled is
also not very clear to me. It might be that loop gain is zero.
HB :
I don’t quite understand what you meant by “mechanically”,
but physiologically I’m pretty sure that there is no “uncontrolled
perception” on lower level, so I can somehow confirm your
“hypothesis” about lower level control units and Bill’s good model of
perceptual control.
Your “hunch” about loop gain zero could be interesting. Does
that mean that in any Bill’s diagram on any level, there could be
“uncontroled perceptions” ?
Best,
Boris
From: Control Systems Group
Network (CSGnet) [mailto:CSGNET@LISTSERV.ILLINOIS.EDU] On Behalf Of Adam Matic
Sent: Wednesday, March 19, 2014
10:01 PM
To: CSGNET@LISTSERV.ILLINOIS.EDU
Subject: Re: Measuring
Input-Output Characteristics of Components of a Closed Loop: Redux
[From Adam Matic
2014.03.19 2200 cet]
Hi Adam, Martin and Bruce,
I’m sorry to interrupt. I’d like to ask one question ? Can we
conclude from Bill’s generic diagram that every perceptual signal is controlled
(goes through comparator) ?
AM:
I think I speak for all of us when I say that - no - we can’t conclude that
every perceptual signal is controlled. There might be some perceptual signals
that are not controlled. My hunch is that lower levels don’t contain such
signals, perhaps some higher levels do. There is a weird connection between
attention, perception and choosing what to control.
What does it mean
mechanically that a perception is not controlled is also not very clear to me.
It might be that loop gain is zero.
Adam
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2014.0.4336 / Virus Database: 3722/7217 - Release Date: 03/19/14
[From Adam Matic 2014.03.19. 2320cet]
AM:
I think I speak for all of us when I say that - no - we can’t conclude that every perceptual signal is controlled. There might be some perceptual signals that are not controlled.
HB :
You seemed not to be sure J, but still your first answer was quite categorical.
AM
That is what I am saying. I can’t be sure that there are or there aren’t any uncontrolled perceptions. I haven’t been convinced by arguments that there are, but something is still fishy.
HB :
Did I understood you right, that lower level does not contain “uncontrolled perception”, and perhaps some higher levels do ? Is the emphasize on “perhaps” or you are sure about that, because your first answer was categorical “no”, what I interpreted that in any kind of Bill’s diagram, we can’t conclude " that every perceptual signal is controlled…But according to what you said we can conclude from Bill’s diagrams that perceptions on lower levels are all controlled ?
AM:
I wouldn’t go as far as claim that there are certainly no uncontrolled perceptions, just that I haven’t seen them modeled, and I don’t see what place they could have in the hierarchy.
As Rick said a couple posts ago, if we somehow disturb some perception and try to find out more about it, it will be a controlled perception.
HB :
I don’t quite understand what you meant by “mechanically”, but physiologically I’m pretty sure that there is no “uncontrolled perception” on lower level, so I can somehow confirm your “hypothesis” about lower level control units and Bill’s good model of perceptual control.
Your “hunch” about loop gain zero could be interesting. Does that mean that in any Bill’s diagram on any level, there could be “uncontroled perceptions” ?
AM:
If uncontrolled really means zero gain and if it is possible that higher levels can vary gain of functions of lower levels, then I suppose so, yes, but there are too many assumptions in here.
Adam
[Martin Taylor 2014.03.19.17.17]
You haven't explained how the hundred million retinal signals are
all controlled, or the million signals in the optic nerve, each of
which is a fluctuating intensity signal, though not necessarily of a
particular point on the retina. If you want to take it to a slightly
higher level, the lowest level of the Powers hierarchy, how many
patches of different intensity can you see around you at this
moment? Ten thousand? A hundred thousand? How many of those are you
controlling independently at this moment? To me the evidence is that
the proportion of perceptions that are controlled gets higher, the
further up the hierarchy one considers. At the bottom level, I doubt
you are controlling anywhere near one in a thousand. At the top,
there may be only one or two perceptions available to control, and
you may be controlling all of them all the time.
Yes. Different people have different hypotheses. And you should add
“consciousness” into that mix.
Why should a perceptual signal be the controlled variable in any
loop at all? Why might it not just be a signal that is registered
and remembered, or compared with another – is that face now I see
on the street one of the people I met yesterday? The perception of
the face cannot be controlled. I may control whether I look at the
face, but I cannot control what face it is when I look at it.
However, the relationship between that face and the remembered ones
from yesterday may be controlled.
Try this diagram. It may help.
The white backgrounds represent components of currently active
Elementary Control Units (ECUs) for which the control loops are
completed through lower levels and through the outer world. P10,
P11, and P12 are perceptual functions that create values used in the
perceptual function P2, which creates the upper level controlled
perception. The perceptions produced by P10 and P11 are
uncontrolled, and that of P2 is controlled entirely by varying the
reference level for the controlled perception of P12. The output O12
(working through more levels of perceptual control, has many large
and small direct effects on the outer world, which propagate to
influence lots of things in the world, some of which also affect the
sensory systems, others of which are just side-effects. In the outer world, the muscular effects ultimately induced by O12,
as well as disturbances not included in the figure, influence things
in the world, which in turn influence sensors. Variations in the
sensor outputs, having passed through various levels of analysis
such as retinal ganglion cells and cortical perceptual functions,
cause variations in the inputs to P10, P11, and P12, and hence to
their output perceptual signals that form the inputs to P2. No
action is ever taken to influence P10 or P11, but the change in
value of the output of P2 results in changes to the error of the P2
control unit, thus changing the reference value of the P12 control
unit, and so on around the loop. P10 and P11 are simply uncontrolled
perceptions that contribute to a controlled perception, the one
created by the P2 perceptual function.
Martin
[From Adam Matic 2014.03.19 2200 cet]
HiAdam, Martin and Bruce,
I'm sorry tointerrupt. I’d like to ask
one question ? Can we conclude from Bill’s
generic diagram that every
perceptual signal is controlled (goes through
comparator) ?
AM:
I think I speak for all of us when I say that - no - wecan’t conclude that every perceptual signal is
controlled. There might be some perceptual signals that
are not controlled. My hunch is that lower levels don’t
contain such signals, perhaps some higher levels do.
There is a weird connection between attention,perception and choosing what to control.
What does it mean mechanically that a perception isnot controlled is also not very clear to me. It might be
that loop gain is zero.
[Martin Taylor 2014.03.19.17.06]
It's an interesting question, and one that might be amenable to
experiment. Even if the structure of HPCT is exactly correct, the
nature of the different levels is not cast in stone, and Bill
himself sometimes switched them around when modelling.
However, I don't see it as relevant to the question I asked, so I
repeat my question, unchanged: “Not even the perception of the
cursor or of the target? If those are not perceived, where does the
perception of the relationship come from”.
I disagree with your assertion that the TrackAnalyze model implies
that the modelled human system has direct access to the
environmental variable we call “target position”. The only evidence
we have of the existence of “target” is a perception of it. In the
model, as in many PCT models, the perceptual function is elided by
being taken to be a unity multiplier, which the software omits. The
same goes for the cursor position perception. In neither case is the
state of the environment being modelled as having some
non-perceptual access to the brain. The only access we have to
properties of the environment is our perception of them.
Anyway, did you look at my two models? I can definitely assure you
that I, as modeller, did not consider that the target position got
into the brain without going through a perceptual system, even if I
took that perceptual system as simply providing a true value for the
target position. And the two models both treat the target position
as an uncontrolled variable.
Martin
[From Adam Matic 2014.03.19 0940cet]
Martin Taylor2014.03.19.16.14
Not even the perception of the cursor or
of the target? If those are not perceived, where does
the perception of the relationship come from?
AM:
This is getting terribly nitpicky, but I really don't
think ‘cursor position’ and ‘target position’ are
perceived by this system in track analyse. In the model’s
input function they are environmental variables that are
used to simulate a visual system and are converted
directly to our p signal, the visual relationship between
visual objects of cursor and target.
In the human visual system, a whole complex process is
happening between environmental cursor position on the
screen, target position on the screen and the neural
representation of the difference between them.
Positions of objects in the visual system are probably
relative, but relative to what? The center of fovea? To
other objects? It could be that the visual system is
consecutively moving target perception to the center of
fovea, and then cursor position to the center of fovea and
somehow creating the representation of the distance, then
calculating the difference. We are modeling just this
difference in the track analyse system.
Adam
[From Adam Matic 2014.03.20 0750 cet]
AM: … no - we can’t conclude that every perceptual signal is controlled. There might be some perceptual signals that are not controlled. My hunch is that lower levels don’t contain such signals, perhaps some higher levels do.
MT: You haven’t explained how the hundred million retinal signals are all controlled.
AM:
They are controlled by their respective control systems. It’s not me who is controlling retinal intensities, but there are probably retinal intensity control systems that do that. I am also not controlling tensions of the tendons connected to my muscles, or arm angles, or arm reach and elevation. There are specific control systems that do that.
What I’m getting at might be connected to the source of all this talking and not much understanding. It might not be, but let’s see.
As Bruce pointed out, and I have had to re-read a few times, he was talking about variables in an experimental situation that the subject can control and other variables that the subject can not control. I take it that you, Martin, have the same view. From the point of view of the experimenter, target position is in a tracking task a variable that the subject does not have control over, while cursor position is something he has control over.
I was struggling to explain my view about controlled variables from the point of view inside the control system. What makes this confusing is that when some first level’s controlled perception is sent ‘upward’, is an environmental variable to the second order perceptual function. It is combined with other environmental variables to create an new perceived and controlled variable on level two. This level two system varies his output, which is not necessarily reference for the mentioned first level system, but could be, and could be reference for another first level system. In the first case, we could say that level two is controlling the system on level one, or manipulating level one, or varying it’s reference. In the second case, it’s controlling a different level one system, and we can’t really say what it is doing to our first level system.
Do you see the difference in use and meaning of the expression “to control”?
How this discussion started is with evaluating psychophysical research. I agree that it is interesting and important that psychophisics found absolute thresholds for different sound frequencies and especially considering this was 150 years ago. Abs. threshold research in any modality is probably a good place to look for possible controlled variables.
Next up was the differential threshold research. I can’t deny the usefulness of finding the logarithmic relationship between reported loudness and physical measures of stimulus intensity, since, as noted, these findings were used to construct the decibel scale of loudness, and many other things.
It was noted that this logarithmic relationship is the same whether the subject has control of the stimulus, or does not have control over it; leading to a discussion about uncontrolled perceptions.
I was pointing out that it is the subject’s auditory system that has control over what it is perceiving, and therefore, sound perception is at all times a controlled variable - from the point of view of the auditory system. I don’t know exactly how exactly this is done, perhaps trough muscles in the middle ear.
So, what exactly this curve made from differential thresholds represents in terms of control systems is an open question.
I hope I have managed to understand the point of view you guys take.
On Thu, Mar 20, 2014 at 4:41 AM, Martin Taylor mmt-csg@mmtaylor.net wrote:
[Martin Taylor 2014.03.19.17.06]
It's an interesting question, and one that might be amenable toexperiment. Even if the structure of HPCT is exactly correct, the
nature of the different levels is not cast in stone, and Bill
himself sometimes switched them around when modelling.However, I don't see it as relevant to the question I asked, so Irepeat my question, unchanged: “Not even the perception of the
cursor or of the target? If those are not perceived, where does the
perception of the relationship come from”.I disagree with your assertion that the TrackAnalyze model impliesthat the modelled human system has direct access to the
environmental variable we call “target position”. The only evidence
we have of the existence of “target” is a perception of it. In the
model, as in many PCT models, the perceptual function is elided by
being taken to be a unity multiplier, which the software omits. The
same goes for the cursor position perception. In neither case is the
state of the environment being modelled as having some
non-perceptual access to the brain. The only access we have to
properties of the environment is our perception of them.Anyway, did you look at my two models? I can definitely assure youthat I, as modeller, did not consider that the target position got
into the brain without going through a perceptual system, even if I
took that perceptual system as simply providing a true value for the
target position. And the two models both treat the target position
as an uncontrolled variable.Martin
[From Adam Matic 2014.03.19 0940cet]
Martin Taylor2014.03.19.16.14
Not even the perception of the cursor or
of the target? If those are not perceived, where does
the perception of the relationship come from?
AM:
This is getting terribly nitpicky, but I really don't
think ‘cursor position’ and ‘target position’ are
perceived by this system in track analyse. In the model’s
input function they are environmental variables that are
used to simulate a visual system and are converted
directly to our p signal, the visual relationship between
visual objects of cursor and target.
In the human visual system, a whole complex process is
happening between environmental cursor position on the
screen, target position on the screen and the neural
representation of the difference between them.
Positions of objects in the visual system are probably
relative, but relative to what? The center of fovea? To
other objects? It could be that the visual system is
consecutively moving target perception to the center of
fovea, and then cursor position to the center of fovea and
somehow creating the representation of the distance, then
calculating the difference. We are modeling just this
difference in the track analyse system.
Adam
Hi Martin, thank you for this diagram, this is extremely helpful. I would add that, what I am sure we all agree, which is that p10 and p11 could still be controlled, even if they send their signals upwards, by various means including whatever higher level systems set their reference values. But it is also indeed possible that they are not. I am thinking however that if p10 and p11 are not currently controlled, this means that the control systems relating to them are in passive observation mode, and it may be the case that after the memory of these perceptions are stored, this still creates error which is then acted upon when controlled mode is restored at this level in the hierachy.
So perceptions that form higher level perceptions can be stored in memory incidentally whilst being uncontrolled, but still will need to be controlled if they create error in their own system.
Imagine I am in a room with a someone who is pulling a critical face and a loud tone of voice. I am in a meeting with them so I am focusing on listening to the content of what they say to be a good colleague (very high level perception) - yet p10 (loud voice) and p11 (critical expression) are signals that are sent upward to form my perception of the content of what they are saying. I am not trying to make them less critical right now. But after the meeting I get intrusions of their face and their loud voice popping into my head. I would suggest that these lower level perceptions have been stored in memory and now I wish to control them. This is driven ultimately by a higher level system, but will need to involve p10 and p11 if I am to successfully make them less critical next time I see them and I can be sure this is happening. So, I enter imagination mode to run through some programs of how I could do this, then do it later when I see them, in controlled mode.
So, is it fair to say that a hierarchy CAN attempt to control any perception within the hierarchy, but it may not be controlling that specific perception at any one time, even if it does contribute to higher level perceptions that are being controlled at that time.
I think that is fairly consistent with Martin, Adam and Rick somehow, by making the conditions of control very specific, or am I wrong?
Warren
On Thu, Mar 20, 2014 at 3:31 AM, Martin Taylor mmt-csg@mmtaylor.net wrote:
[Martin Taylor 2014.03.19.17.17]
[From Adam Matic 2014.03.19 2200 cet]
You haven't explained how the hundred million retinal signals areall controlled, or the million signals in the optic nerve, each of
which is a fluctuating intensity signal, though not necessarily of a
particular point on the retina. If you want to take it to a slightly
higher level, the lowest level of the Powers hierarchy, how many
patches of different intensity can you see around you at this
moment? Ten thousand? A hundred thousand? How many of those are you
controlling independently at this moment? To me the evidence is that
the proportion of perceptions that are controlled gets higher, the
further up the hierarchy one considers. At the bottom level, I doubt
you are controlling anywhere near one in a thousand. At the top,
there may be only one or two perceptions available to control, and
you may be controlling all of them all the time.
Yes. Different people have different hypotheses. And you should add“consciousness” into that mix.
Why should a perceptual signal be the controlled variable in anyloop at all? Why might it not just be a signal that is registered
and remembered, or compared with another – is that face now I see
on the street one of the people I met yesterday? The perception of
the face cannot be controlled. I may control whether I look at the
face, but I cannot control what face it is when I look at it.
However, the relationship between that face and the remembered ones
from yesterday may be controlled.Try this diagram. It may help. The white backgrounds represent components of currently activeElementary Control Units (ECUs) for which the control loops are
completed through lower levels and through the outer world. P10,
P11, and P12 are perceptual functions that create values used in the
perceptual function P2, which creates the upper level controlled
perception. The perceptions produced by P10 and P11 are
uncontrolled, and that of P2 is controlled entirely by varying the
reference level for the controlled perception of P12. The output O12
(working through more levels of perceptual control, has many large
and small direct effects on the outer world, which propagate to
influence lots of things in the world, some of which also affect the
sensory systems, others of which are just side-effects.In the outer world, the muscular effects ultimately induced by O12,as well as disturbances not included in the figure, influence things
in the world, which in turn influence sensors. Variations in the
sensor outputs, having passed through various levels of analysis
such as retinal ganglion cells and cortical perceptual functions,
cause variations in the inputs to P10, P11, and P12, and hence to
their output perceptual signals that form the inputs to P2. No
action is ever taken to influence P10 or P11, but the change in
value of the output of P2 results in changes to the error of the P2
control unit, thus changing the reference value of the P12 control
unit, and so on around the loop. P10 and P11 are simply uncontrolled
perceptions that contribute to a controlled perception, the one
created by the P2 perceptual function.Martin
–
Dr Warren Mansell
Reader in Psychology
Cognitive Behavioural Therapist & Chartered Clinical Psychologist
School of Psychological Sciences
Coupland I
University of Manchester
Oxford Road
Manchester M13 9PL
Email: warren.mansell@manchester.ac.uk
Tel: +44 (0) 161 275 8589
Website: http://www.psych-sci.manchester.ac.uk/staff/131406
See teamstrial.net for further information on our trial of CBT for Bipolar Disorders in NW England
The highly acclaimed therapy manual on A Transdiagnostic Approach to CBT using Method of Levels is available now.
Check www.pctweb.org for further information on Perceptual Control Theory
Hi
Adam, Martin and Bruce,
I'm sorry tointerrupt. I’d like to ask
one question ? Can we conclude from Bill’s
generic diagram that every
perceptual signal is controlled (goes through
comparator) ?
AM:
I think I speak for all of us when I say that - no - we
can’t conclude that every perceptual signal is
controlled. There might be some perceptual signals that
are not controlled. My hunch is that lower levels don’t
contain such signals, perhaps some higher levels do.
There is a weird connection between attention,
perception and choosing what to control.
What does it mean mechanically that a perception is
not controlled is also not very clear to me. It might be
that loop gain is zero.
[From Bruce Abbott (2014.03.20.0900 EDT)]
Rick Marken (2014.03.19.0940)–
Bruce Abbott (2014.03.19.1005 EDT)
BA: …You can learn something about the input function of the variable you are controlling in ChooseControl (namely, that you can perceive that aspect of the object). If you concluded from this that, generally speaking, you can’t learn something about an input function of a variable unless the variable is under control, then you would be committing a logical fallacy known as “affirming the consequent.” In fact you can learn about input functions of variables that are not under control, as I’ve illustrated previously with psychophysical examples.
RM: I agree that we can learn about the input functions that produce perceptions whether or not those perceptions are currently under control. But I must have missed your illustration of how this is done in psychophysical experiments. I can’t think of any psychophysical procedure that doesn’t involve asking the subject to either directly control the perceptual variable under study (method of adjustment, limits, etc) or directly control a perception that is disturbed by the perceptual variable under study (magnitude estimation). I guess I’m thinking that it’s not possible to learn about the characteristics of input function of any perceptual variable in psychophysical experiments without that variable being under control; if it’s just a disturbance to a variable that’s under control then you have the behavioral illusion problem (as you do have in magnitude estimation experiments; see http://www.mindreadings.com/BehavioralIllusion.pdf).
BA: I guess you did miss it, so here it is again.
This procedure will do the job:
· Present tone at a given frequency and intensity
· Have participant raise finger if tone is detected, lower it if not
· Change to a new frequency and/or intensity
· Repeat
The participant controls a relationship between perceiving/not perceiving the tone and raising/not raising the finger.
The participant has no control over the presentation of the tone – onset time, duration, frequency and intensity are controlled by the audiologist.
Thus, the variable that the audiologist assesses (intensity threshold) is not the variable that the participant controls.
Presenting/not presenting the tone does potentially disturb the participant’s relationship perception, but so what? That’s a different perception from the one the audiologist is assessing. Clearly, in this case the behavioral illusion is irrelevant to the audiologist’s analysis.
Bruce
[From Bruce Abbott (2014.03.20.0930 EDT)]
Adam Matic 2014.03.19 –
Bruce Abbott (2014.03.19.1220 EDT)
Some demos do compute a perception that is not controlled by the program. For example, the tracking demo does compute the perception of the target position as a function of the disturbance waveform that is being applied to target position. Target position is determined S-R fashion from these computations; the program contains no control system for target position.
AM:
This is the way I see it:
The controlled variable is the difference, so neither the ‘cursor position’ nor the ‘target position’ are controlled variables. Cursor position is an aspect of the environment and is the output function that is varied to control the perceptual variable.
BA previously: With respect to your last sentence above, it is not true that “we are never in control of the environment.” If that were true, we wouldn’t last long. Although the only things I can know about the environment are those presented to me by my perceptual systems. However, if jumping back onto the (perceived) curb to avoid (the perception of) being hit by a (perceived) car did not prevent me from actually being hit by the very real car behind those perceptions, I’d soon be dead. It is true that that’s an article of faith (perhaps the only reality that exists is the one in our minds – er, my mind, since you would exist only in my mind), but I wouldn’t be willing to test that proposition by jumping in front of a speeding car. The phrase “it’s all perception” refers to what we actually experience, as opposed to directly experiencing Reality. It is not a denial that Reality exists.
AM:
We do affect the environment. We do vary it. We are just not in control of it, if we take the strict meaning of “to control”.
BA: “The environment” is too broad a term. Control is always over some aspect, a particular variable, not “the environment” in general.
BA: To control a variable (even an environmental one) is to produce an action whose effects feed back onto the same variable, in such a way as to reduce the error between the current state of that variable and the control system’s reference value for that variable (negative feedback). That’s the strict meaning of “to control.” And we control environmental variables in this way all the time. That’s what behavior is for.
Where you are getting hung up, I suspect, is in the difference between an environmental variable and our perception of it. The actions of a control system typically affect an environmental variable; this variable is perceived via sensory and related perceptual mechanisms. Our actions affect the environmental variable, which affect our perceptions of its value. Those perceptions are compared to our internal reference; the difference (error) drives our output, the physical actions that in turn feed back to affect the environmental variable.
Although it is technically correct to say that we control our perceptions, you should not take this to mean that we do not control the environmental variables of which our perceptions are a function.
Consider, for example, an automobile’s cruise control. A sensor senses the car’s speed and generates a voltage proportional to that speed. The voltage represents the cruise control’s perception of the car’s speed. It compares that speed to a voltage representing the driver’s set point or reference speed. The difference between the two drives the output, which adjusts the engine’s throttle setting. This adjustment increases or decreases power to the wheels so as to bring the car’s sensed speed toward the set point. In consequence, the car’s actual speed on the road is adjusted to match the reference value that the driver set.
If the system is working properly, it will control the car’s actual speed on the road, not just the cruise control’s perception of that speed, and it does so by making physical adjustments to environmental variables – throttle position, and thus engine power and road speed.
Bruce
[From Bruce Abbott (2014.03.20.0940 EDT)]
Adam Matic 2014.03.17 1910 cet –
AM: This is how Bill puts it in LCSIII, chapter 4, page 77, last paragraph:
AM: “Tracking is just one form of control: it’s done by controlling a visual relationship, one part of which varies independently (the target), and the other part of which is affected by the participant’s actions.”
BA: Bill is right, but you are misunderstanding him. The part that varies independently (the target) is not under control; the other part (the cursor) is affected by the participant’s actions. A variable is not under control unless the actions affect that variable. The visual relationship between target and cursor is controlled by means of controlling the position of the cursor.
To perceive the visual relationship between target and cursor, one must perceive both visual objects. Yet only the cursor is under the participant’s direct control.
Bruce