On the perceptual hierarchy

[From Bruce Abbott (2017.11.21.1805 EST)}

The hierarchy of perceptual control described by Bill Powers in B:CP and elsewhere is as much a hierarchy of perceptions as of control. At the bottom level of this hierarchy are, according to PCT, intensities, which arise directly from sensory endings. Bill notes that “any one first-order perceptual signal can vary only along one dimension – frequency. This means that however the associated input function is stimulated, only one dimension of that stimulation can affect the perceptual signal: the intensity of stimulation.” He goes on to note that “a first-order perceptual signal reflects only what happens at the sensory ending: the source of the stimulation is completely undefined and unsensed.” He states that the “information is one-dimensional, indicating only how much effect is occurring.” And from this he concludes that the only thing one can tell based on such a sensory input is its intensity; the nature of the input – whether visual, auditory, etc., is not communicated.

One could argue just as well that in PCT, all perceptual signals are one-dimensional, varying only in the amplitude of the neural currents that represent them. Nevertheless, we usually can and do distinguish them, even when they arise directly from the stimulation of sensory receptors. We are able to do this because of where the signals are analyzed in the brain. Auditory signals, for example, are routed from the cochlea of the ear (where the so-called hair cells respond to pressure variations within a certain range of frequencies and stimulate associated auditory neurons) to the cochlear nucleus in the brainstem, to the thalamus, and from there to the auditory cortex located in the temporal lobes. Our conscious experience of this stimulation arises there in the auditory cortex, where we perceive it as sound. Route the same stream of impulses to, say, the visual cortex, and it will be experienced as light.

According to Bill, it is only when two or more intensities are combined that the product is a definite sensation. I don’t see how the process of combination could yield an input identifiable as a specific kind of sensation (such as light or sound), because the resulting perception is just another neural signal varying in intensity.

To further complicate things, what our sensory receptors respond to is not necessarily simply the intensity of the stimulation. For example, receptors that respond to skin temperature respond more strongly to change in temperature than to temperature per se. The output of the retina of the eye depends strongly on change of intensity, so much so that stabilizing an image on the retina quickly produces a loss of vision (although fragments of the image may reappear and disappear over time). Thus, to suggest that the outputs of first-order sensors represent simple intensities seems oversimplified.

What I worry about is that the simple hierarchical structure imagine by Bill and colleagues doesn’t fit well with what we currently know about the perceptual systems. However, the evidence does support the view that perception is hierarchical. Retinal output via the ganglion cells is already the product of significant processing of the photoreceptor inputs, including contrast enhancement and the formation of color-opponent signals. Input to the visual cortex area V1 is analyzed to produce signals representing the angles of edges or lines passing across specific regions of each retina, and these signals are processed “upstream,” yielding the perception of objects and so on. Interestingly, movement within the visual field is analyzed by a specific cortical area; in the rare cases in which this area has ceased to function, individuals report seeing the world as a series of still images without motion. (One would think that perception of motion would depend only on “noticing” that the position of an object has changed, relative to others, but apparently not!)

It is worth noting that control systems do not need to “know” the nature of the sources of their inputs – they just compare the resulting neural signal to the reference signal and take action to correct the difference. Furthermore, control may not involve any consciousness of the stimulation. We have no conscious sensation of blood pressure, for example, yet there are “baroreceptors” located in the aorta leading into the heart, which provide inputs to the blood-pressure regulating systems. It is just as true, however, that we may consciously aware of what we are controlling, what the reference for that perception is, and what actions we are performing in order to exert that control.

Comments?

Bruce

One could argue just as well that in PCT,
all perceptual signals are one-dimensional, varying only in
the amplitude of the neural currents that represent them.
Nevertheless, we usually can and do distinguish them, even
when they arise directly from the stimulation of sensory
receptors. We are able to do this because of where the
signals are analyzed in the brain.

[From Bruce Abbott (2017.11.22.1055 EST)]

Rupert Young (2017.11.22 11.50 )]

(Bruce Abbott (2017.11.21.1805 EST)}

RY: These are all good and valid points, and pretty much agree with everything you said.

One could argue just as well that in PCT, all perceptual signals are one-dimensional, varying only in the amplitude of the neural currents that represent them. Nevertheless, we usually can and do distinguish them, even when they arise directly from the stimulation of sensory receptors. We are able to do this because of where the signals are analyzed in the brain.

RY: But don’t we only distinguish them due to conscious awareness of them; that is, we give the signals labels as part of conscious meaning. I’d suggest that it is not necessary to distinguish signals in this way for a system to control successfully [you say this below]. In other words, a living system without consciousness (e.g. fly) can control signals successfully irrespective of where they originate. Though there may be an additional role for consciousness that provides other capabilities.

According to Bill, it is only when two or more intensities are combined that the product is a definite sensation. I don’t see how the process of combination could yield an input identifiable as a specific kind of sensation (such as light or sound), because the resulting perception is just another neural signal varying in intensity.

RY: Why does it need to be identifiable as a specific kind of sensation?

BA: Logically it would not seem necessary, yet my own experiences come in the form of distinctive kinds of sensation – the qualia of experience. I suspect that the same holds for you. Despite the apparent lack of logical necessity, sensations are indeed identifiable as specific kinds of sensation, at least at the cortical level where somehow those neural signals somehow generate conscious experience. So why does the brain take the trouble of labeling different kinds of sensations as qualitatively different experiences? (How this happens is a total mystery, one of the “hard” problems of psychology. Answering “why” may be less difficult.)

BA: I speculate that it has to do with how conscious experience is experienced – as a Cartesian Theater in which perceptions arising from the various sense modalities appear simultaneously. To prevent the various modalities being confused, they must appear in consciousness as distinctive kinds of experience. Because they are so distinctive, I’m never going to confuse sensations arising from my eyes with those from my ears or nose. What we experience is the brain’s model of reality, both external to the body and to some degree also internal. Perhaps by keeping perceptions within that representation qualitatively distinctive, it allows the brain to efficiently deal with the aspects of the environment that are represented in experience.

To further complicate things, what our sensory receptors respond to is not necessarily simply the intensity of the stimulation. For example, receptors that respond to skin temperature respond more strongly to change in temperature than to temperature per se. The output of the retina of the eye depends strongly on change of intensity, so much so that stabilizing an image on the retina quickly produces a loss of vision (although fragments of the image may reappear and disappear over time). Thus, to suggest that the outputs of first-order sensors represent simple intensities seems oversimplified.

RM: I agree, but the hierarchy is a simplification to be taken as a rough guideline of the type of things that may be controlled. I suspect the reality is much messier, and you can have change (transitions) at the lower level, and you can have a sequence of events and and an event of sequences etc etc.

What I worry about is that the simple hierarchical structure imagine by Bill and colleagues doesn’t fit well with what we currently know about the perceptual systems. However, the evidence does support the view that perception is hierarchical. Retinal output via the ganglion cells is already the product of significant processing of the photoreceptor inputs, including contrast enhancement and the formation of color-opponent signals. Input to the visual cortex area V1 is analyzed to produce signals representing the angles of edges or lines passing across specific regions of each retina, and these signals are processed “upstream,” yielding the perception of objects and so on. Interestingly, movement within the visual field is analyzed by a specific cortical area; in the rare cases in which this area has ceased to function, individuals report seeing the world as a series of still images without motion. (One would think that perception of motion would depend only on “noticing” that the position of an object has changed, relative to others, but apparently not!)

It is worth noting that control systems do not need to “know” the nature of the sources of their inputs – they just compare the resulting neural signal to the reference signal and take action to correct the difference.

RM: Yep, its signals all the way down.

Furthermore, control may not involve any consciousness of the stimulation.

RM: Yep, this is what I meant above.

Bruce

Hi Bruce, I think this is very possible indeed and a conclusion I had considered before too:Â

BA: I speculate that it has to do with how conscious experience is experienced – as a Cartesian Theaterr in which perceptions arising from the various sense modalities appear simultaneously. To prevent the various modalities being confused, they must appear in consciousness as distinctive kinds of experience. Because they are so distinctive, I’m never going to confuse sensations arising from my eyes with those from my ears or nose. What we experience is the brain’s model of reality, both external to the body and to some degree also internal. Perhaps by keeping perceptions within that representation qualitatively distinctive, it allows the brain to efficiently deal with the aspects of the environment that are represented in experience. Â

···

On Wed, Nov 22, 2017 at 3:57 PM, Bruce Abbott bbabbott@frontier.com wrote:

[From Bruce Abbott (2017.11.22.1055 EST)]

Rupert Young (2017.11.22 11.50 )]

(Bruce Abbott (2017.11.21.1805 EST)}

RY: These are all good and valid points, and pretty much agree with everything you said.

One could argue just as well that in PCT, all perceptual signals are one-dimensional, varying only in the amplitude of the neural currents that represent them. Nevertheless, we usually can and do distinguish them, even when they arise directly from the stimulation of sensory receptors. We are able to do this because of where the signals are analyzed in the brain.

RY: But don’t we only distinguish them due to conscious awareness of them; that is, we give the signals labels as part of conscious meaning. I’d suggest that it is not necessary to distinguish signals in this way for a system to control successfully [you say this below]. In other words, a living system without consciousness (e.g. fly) can control signals successfully irrespective of where they originate. Though there may be an additional role for consciousness that provides other capabilities.

According to Bill, it is only when two or more intensities are combined that the product is a definite sensation. I don’t see how the process of combination could yield an input identifiable as a specific kind of sensation (such as light or sound), because the resulting perception is just another neural signal varying in intensity.

RY: Why does it need to be identifiable as a specific kind of sensation?

Â

BA: Logically it would not seem necessary, yet my own experiences come in the form of distinctive kinds of sensation – the qqualia of experience. I suspect that the same holds for you. Despite the apparent lack of logical necessity, sensations are indeed identifiable as specific kinds of sensation, at least at the cortical level where somehow those neural signals somehow generate conscious experience. So why does the brain take the trouble of labeling different kinds of sensations as qualitatively different experiences? (How this happens is a total mystery, one of the “hardâ€? problems of psychology. Answering “whyâ€? may be less difficult.)

Â

BA: I speculate that it has to do with how conscious experience is experienced – as a Cartesian Theater in which perceptions arising ffrom the various sense modalities appear simultaneously. To prevent the various modalities being confused, they must appear in consciousness as distinctive kinds of experience. Because they are so distinctive, I’m never going to confuse sensations arising from my eyes with those from my ears or nose. What we experience is the brain’s model of reality, both external to the body and to some degree also internal. Perhaps by keeping perceptions within that representation qualitatively distinctive, it allows the brain to efficiently deal with the aspects of the environment that are represented in experience. Â

To further complicate things, what our sensory receptors respond to is not necessarily simply the intensity of the stimulation. For example, receptors that respond to skin temperature respond more strongly to change in temperature than to temperature per se. The output of the retina of the eye depends strongly on change of intensity, so much so that stabilizing an image on the retina quickly produces a loss of vision (although fragments of the image may reappear and disappear over time). Thus, to suggest that the outputs of first-order sensors represent simple intensities seems oversimplified.

RM: I agree, but the hierarchy is a simplification to be taken as a rough guideline of the type of things that may be controlled. I suspect the reality is much messier, and you can have change (transitions) at the lower level, and you can have a sequence of events and and an event of sequences etc etc.

What I worry about is that the simple hierarchical structure imagine by Bill and colleagues doesn’t fit well with what we currently know about the perceptual systems. However, the evidence does support the view that perception is hierarchical. Retinal output via the ganglion cells is already the product of significant processing of the photoreceptor inputs, including contrast enhancement and the formation of color-opponent signals. Input to the visual cortex area V1 is analyzed to produce signals representing the angles of edges or lines passing across specific regions of each retina, and these signals are processed “upstream,â€? yielding the perception of objects and so on. Interestingly, movement within the visual field is analyzed by a specific cortical area; in the rare cases in which this area has ceased to function, individuals report seeing the world as a series of still images without motion. (One would think that perception of motion would depend only on “noticingâ€? that the position of an object has changed, relative to others, but apparently not!)

Â

It is worth noting that control systems do not need to “knowâ€? the nature of the sources of their inputs – they just commpare the resulting neural signal to the reference signal and take action to correct the difference.

RM: Yep, its signals all the way down.

Furthermore, control may not involve any consciousness of the stimulation.Â

RM: Yep, this is what I meant above.

Bruce

Dr Warren Mansell
Reader in Clinical Psychology

School of Health Sciences
2nd Floor Zochonis Building
University of Manchester
Oxford Road
Manchester M13 9PL
Email: warren.mansell@manchester.ac.uk
Â
Tel: +44 (0) 161 275 8589
Â
Website: http://www.psych-sci.manchester.ac.uk/staff/131406
Â
Advanced notice of a new transdiagnostic therapy manual, authored by Carey, Mansell & Tai - Principles-Based Counselling and Psychotherapy: A Method of Levels Approach

Available Now

Check www.pctweb.org for further information on Perceptual Control Theory

[From Bruce
Abbott (2017.11.22.1055 EST)]

Rupert Young (2017.11.22 11.50 )]

(Bruce Abbott (2017.11.21.1805 EST)}

        BA:  I

speculate that it has to do with how conscious experience is
experienced – as a Cartesian Theater in which perceptions
arising from the various sense modalities appear
simultaneously. To prevent the various modalities being
confused, they must appear in consciousness as distinctive
kinds of experience. Because they are so distinctive, I’m
never going to confuse sensations arising from my eyes with
those from my ears or nose.

[From Rupert Young (2017.11.23 10.50 )]

  Bruce, I often find your reply posts quite difficult to parse as

some parts of the previous message are indented and some not, and
some parts are labelled with poster and some not (and labelled RM
rather than RY :slight_smile: ), and you include the whole message but only
respond to bits, so I find them quite difficult to follow . Maybe
it’s just me.

(

If there is any modal confusion wouldn’t that result in error
leading to reorganisation, of the modalities? When a system first
organises it will organise its connections in such a way that its
output affects its input; it doesn’t need to “know” about
modalities, does it? So, I don’t see any reason why modal confusion
would occur in the first place. Modality only has meaning in the
context of conscious experience.
Rupert

···

Anyway …

Bruce Abbott (2017.11.22.1055 EST)]

        BA:  I

speculate that it has to do with how conscious experience is
experienced – as a Cartesian Theater in which perceptions
arising from the various sense modalities appear
simultaneously. To prevent the various modalities being
confused, they must appear in consciousness as distinctive
kinds of experience. Because they are so distinctive, I’m
never going to confuse sensations arising from my eyes with
those from my ears or nose. What we experience is the
brain’s model of reality, both external to the body and to
some degree also internal. Perhaps by keeping perceptions
within that representation qualitatively distinctive, it
allows the brain to efficiently deal with the aspects of the
environment that are represented in experience.

[Martin Taylor 2017.11.23.08.39]

[From Rupert Young (2017.11.23 10.50 )]

Bruce, I often find your reply posts quite difficult to parse as some parts of the previous message are indented and some not, and some parts are labelled with poster and some not (and labelled RM rather than RY :slight_smile: ), and you include the whole message but only respond to bits, so I find them quite difficult to follow . Maybe it's just me.

No, it's not just you.

Anyway ...

...
If there is any modal confusion wouldn't that result in error leading to reorganisation, of the modalities?

Why would it result in error? Error is the difference between a perceptual value and its reference value. How many modalities contribute inputs to the perceptual function seems to me to be irrelevant to whether the system has reorganized to produce actions that move the perception toward its reference value.

When a system first organises it will organise its connections in such a way that its output affects its input; it doesn't need to "know" about modalities, does it?

I wouldn't have thought it would in general, though there may be cases in which it might. In those cases, the identity of the modality would contribute to some perceptual input function to create a controllable perception.

So, I don't see any reason why modal confusion would occur in the first place. Modality only has meaning in the context of conscious experience.

I fail to see the connection between modal confusion, such as occurs in synaesthesia or the kinds of momentary experiences I mentioned in [Martin Taylor 2017.11.22.11.03], and either error or your conjecture that modality has meaning only to consciousness. Maybe you could expand on those ideas?

Martin

[From Bruce Abbott (2017.11.24.1240 EST)]

[From Rupert Young (2017.11.23 10.50 )]

RY, definitely not RM :->: Bruce, I often find your reply posts quite difficult to parse as some parts of the previous message are indented and some not, and some parts are labelled with poster and some not (and labelled RM rather than RY :slight_smile: ), and you include the whole message but only respond to bits, so I find them quite difficult to follow . Maybe it’s just me.

BA: I try to be consistent, but evidently I am failing badly. But I must say that part of the problem is that I am responding to other posts that don’t all use the same style, and I haven’t bothered to change the quoted text. In some posts I receive, responses to previous posts are simply marked with the initials of the originator, in others, the text is indented, and in yet others, previous material is preceded with a < on every line.

BA: In my own writing, I have used indented text when lifting material from a source, such as a web page or book, an initials to indicate the writer of previous post material. In general I’ve also deleted material from the post to which I am currently responding, where I did not have a comment. However, I did not do that in the post that is the subject of this complaint.

BA: Keeping things straight was easier in the old Eudora days, because that program always used carets to mark material from the post being replied to. Outlook doesn’t – or if it does, I haven’t figured out how to set that up.

RY: Anyway …

(Bruce Abbott (2017.11.22.1055 EST)]

BA: I speculate that it has to do with how conscious experience is experienced – as a Cartesian Theater in which perceptions arising from the various sense modalities appear simultaneously. To prevent the various modalities being confused, they must appear in consciousness as distinctive kinds of experience. Because they are so distinctive, I’m never going to confuse sensations arising from my eyes with those from my ears or nose. What we experience is the brain’s model of reality, both external to the body and to some degree also internal. Perhaps by keeping perceptions within that representation qualitatively distinctive, it allows the brain to efficiently deal with the aspects of the environment that are represented in experience.

RY: If there is any modal confusion wouldn’t that result in error leading to reorganisation, of the modalities? When a system first organises it will organise its connections in such a way that its output affects its input; it doesn’t need to “know” about modalities, does it? So, I don’t see any reason why modal confusion would occur in the first place. Modality only has meaning in the context of conscious experience.

BA: Phenomena phirst! We do in fact have conscious experience (or at least I claim to, and infer that you, other human beings, and probably many species of animal do also). The fact that control systems are capable of functioning perfectly well without consciousnses makes its presence all the more mysterious. For it to have evolved and been conserved over thousands of generations, it must confer significant advantages with respect to survival and reproductive success.

BA: Bill Powers speculated that at conscious attention is somehow involved at directing reorganization to the systems in which there is persistent error, but as a formal theory, PCT ignores consciousness. Leaving out consciousness reduces human beings to a vast collection of hierarchically organized control systems. I suspect that most folks find that picture less than appealing.

BA: Perhaps a key to the role of consciousness can be found in the way in which we experience it. Not only to we perceive the world around us as a coherent picture in which all the senses participate, we also perceive ourselves at the center of it. That is, the brain constructs a model of reality that includes ourselves in the middle of it as perceivers and doers. When we are first learning a new skill, we consciously experiment, trying this and that until we succeed in performing at least adequately – conscious perception and thought allow us to consciously construct control systems and improve them, instead of relying strictly on random reorganization to find a solution. In this way, consciousness plays the role of supervisor in “supervised” learning. As we become skilled, the control system’s structure stabilizes the activity can be performed without conscious intervention – we say that it has become a “habit.” The system may then undergo further refinement as reorganization gradually fine-tunes the system.

BA: When I taught motivation in my classes, I liked to point out that physiological control systems such as those regulating blood pressure or heart rate generally operate without conscious attention. However, certain states do produce conscious experience, such as the state of feeling chilled, hungry, or in pain. I suggested that we have conscious experience of such states because these are conditions in which such perceptions allow us to consciously learn control strategies that are often more effective than merely relying on automatic unconscious mechanisms. In other words, we have these perceptions in consciousness because there are ways in which we can deal with them via our actions. Thus, when I’m cold, for example, I don’t have to rely on shivering (which generates extra heat) and the constriction of capillaries near the skin surface (which reduces heat loss through the skin) to keep us from freezing to death. Experiencing the cold, we learn what works to reduce that unpleasant experience. We put on a coat, run in place, cozy up to a fire, turn up the thermostat, or drink a hot drink, among other possibilities. We construct control systems that use behavior as the means to improve control over such variables, and which employ different means toward the same end under different circumstances, depending on what the environment makes available.

BA: It appears to me that in conditions such as hunger, thirst, feeling hot or cold, and so on, what we experience is not only labeled – hunger feels different than heat or cold or pain – but that the discomfort associated with these state suggests that we are experiencing them as errors. By providing a distinctive feeling in consciousness, these error states not only indicate error, but they signal in which control systems the error exists – those regulating feeding/nutrient level, deficits in hydration, deviations from reference body temperature, stress on damaged muscles, and so on. This allows us to select the appropriate means to deal with such errors (as learned through previous experience).

BA: Comments?

Bruce

[From Bruce
Abbott (2017.11.24.1240 EST)]

[From Rupert Young (2017.11.23 10.50 )]

      RY, definitely not RM :->:
      Bruce, I often find your reply posts quite difficult to

parse as some parts of the previous message are indented and
some not, and some parts are labelled with poster and some not
(and labelled RM rather than RY :slight_smile: ), and you include the
whole message but only respond to bits, so I find them quite
difficult to follow . Maybe it’s just me.

        BA: I try to be consistent,

but evidently I am failing badly. But I must say that part
of the problem is that I am responding to other posts that
don’t all use the same style, and I haven’t bothered to
change the quoted text. In some posts I receive, responses
to previous posts are simply marked with the initials of the
originator, in others, the text is indented, and in yet
others, previous material is preceded with a < on every
line.

···

RY: Anyway …

( Bruce
Abbott (2017.11.22.1055 EST)]

        BA: Perhaps

a key to the role of consciousness can be found in the way
in which we experience it. Not only to we perceive the
world around us as a coherent picture in which all the
senses participate, we also perceive ourselves at the center
of it. That is, the brain constructs a model of reality
that includes ourselves in the middle of it as perceivers
and doers. When we are first learning a new skill, we
consciously experiment, trying this and that until we
succeed in performing at least adequately – conscious
perception and thought allow us to consciously construct
control systems and improve them, instead of relying
strictly on random reorganization to find a solution. In
this way, consciousness plays the role of supervisor in
“supervised” learning. As we become skilled, the control
system’s structure stabilizes the activity can be performed
without conscious intervention – we say that it has become a
“habit.” The system may then undergo further refinement as
reorganization gradually fine-tunes the system.

        BA: When I

taught motivation in my classes, I liked to point out that
physiological control systems such as those regulating blood
pressure or heart rate generally operate without conscious
attention. However, certain states do produce conscious
experience, such as the state of feeling chilled, hungry, or
in pain. I suggested that we have conscious experience of
such states because these are conditions in which such
perceptions allow us to consciously learn control strategies
that are often more effective than merely relying on
automatic unconscious mechanisms. In other words, we have
these perceptions in consciousness because there are ways in
which we can deal with them via our actions. Thus, when I’m
cold, for example, I don’t have to rely on shivering (which
generates extra heat) and the constriction of capillaries
near the skin surface (which reduces heat loss through the
skin) to keep us from freezing to death. Experiencing the
cold, we learn what works to reduce that unpleasant
experience. We put on a coat, run in place, cozy up to a
fire, turn up the thermostat, or drink a hot drink, among
other possibilities. We construct control systems that use
behavior as the means to improve control over such
variables, and which employ different means toward the same
end under different circumstances, depending on what the
environment makes available.

        BA:  It

appears to me that in conditions such as hunger, thirst,
feeling hot or cold, and so on, what we experience is not
only labeled – hunger feels different than heat or cold or
pain – but that the discomfort associated with these state
suggests that we are experiencing them as errors. By
providing a distinctive feeling in consciousness, these
error states not only indicate error, but they signal in
which control systems the error exists – those regulating
feeding/nutrient level, deficits in hydration, deviations
from reference body temperature, stress on damaged muscles,
and so on. This allows us to select the appropriate means
to deal with such errors (as learned through previous
experience).

        BA:

Comments?

[From Rupert Young (2017.11.25 11.55 )]

(Martin Taylor 2017.11.23.08.39]

...
If there is any modal confusion wouldn't that result in error leading to reorganisation, of the modalities?

Why would it result in error? Error is the difference between a perceptual value and its reference value. How many modalities contribute inputs to the perceptual function seems to me to be irrelevant to whether the system has reorganized to produce actions that move the perception toward its reference value.

Suppose you have something increasing pressure on your chest. But due to modal confusion you perceive that pressure as light. Turning a dimmer switch is going to have no effect on the pressure so the error will not reduce but increase.

When a system first organises it will organise its connections in such a way that its output affects its input; it doesn't need to "know" about modalities, does it?

I wouldn't have thought it would in general, though there may be cases in which it might. In those cases, the identity of the modality would contribute to some perceptual input function to create a controllable perception.

Isn't that the same as what I said?

So, I don't see any reason why modal confusion would occur in the first place. Modality only has meaning in the context of conscious experience.

I fail to see the connection between modal confusion, such as occurs in synaesthesia or the kinds of momentary experiences I mentioned in [Martin Taylor 2017.11.22.11.03], and either error or your conjecture that modality has meaning only to consciousness. Maybe you could expand on those ideas?

The momentary experiences you mentioned don't have an effect on intrinsic error so they are not an issue in the context of reorganisation. If such modal confusion did have a negative effect on intrinsic error the organisation that led to them would be organised out of existence. So, perhaps, the only modal confusion that would develop is that that would not have an effect on intrinsic error.

When we talk about modalities we are just putting labels on signals. To an organism which is not conscious this is not relevant; it is all about the control of signals that maintain low intrinsic error. If the intrinsic error increases the organism will reorganise to control different signals (or control the same signals better) that do reduce that error. It doesn't matter to the organism what those modalities are.

However, there may be additional benefits to consciousness, for better learning associated with directed reorganisation along the lines Bruce outlined. I'm just saying that I don't see the necessity for consciousness for (all) organisms to control successfully, and so, the labelling of the origins of the signals.

Rupert

[From Rupert Young (2017.11.25 12.20 )]

(

Yes, hopefully this won’t be an issue if we move to a proper forum.
I like the idea of Hmm, I wonder if this relates to my point about reorganisation
taking place in sub-hierarchies due to the error at the top of the
hierarchy, rather than intrinsic error. So if you are learning to
play the guitar, and experience the error of being rubbish, you
direct efforts at places within that guitar-playing sub-hierarchy,
such as the lower level of finger positions, that will contribute to
reducing the overall error within that sub-hierarchy.
Rupert

···

Bruce Abbott (2017.11.24.1240 EST)]

        BA: But I must say that part

of the problem is that I am responding to other posts that
don’t all use the same style, …

        BA: Perhaps

a key to the role of consciousness can be found in the way
in which we experience it. Not only to we perceive the
world around us as a coherent picture in which all the
senses participate, we also perceive ourselves at the center
of it. That is, the brain constructs a model of reality
that includes ourselves in the middle of it as perceivers
and doers. When we are first learning a new skill, we
consciously experiment, trying this and that until we
succeed in performing at least adequately – conscious
perception and thought allow us to consciously construct
control systems and improve them, instead of relying
strictly on random reorganization to find a solution. In
this way, consciousness plays the role of supervisor in
“supervised” learning. As we become skilled, the control
system’s structure stabilizes the activity can be performed
without conscious intervention – we say that it has become a
“habit.” The system may then undergo further refinement as
reorganization gradually fine-tunes the system.

  “supervised”

learning coming from within.

        BA:  It

appears to me that in conditions such as hunger, thirst,
feeling hot or cold, and so on, what we experience is not
only labeled – hunger feels different than heat or cold or
pain – but that the discomfort associated with these state
suggests that we are experiencing them as errors. By
providing a distinctive feeling in consciousness, these
error states not only indicate error, but they signal in
which control systems the error exists – those regulating
feeding/nutrient level, deficits in hydration, deviations
from reference body temperature, stress on damaged muscles,
and so on. This allows us to select the appropriate means
to deal with such errors (as learned through previous
experience).

[Martin Taylor 2017.11.25.11.05]

I think we are talking at cross purposes, because I agree with just about everything you say, without disagreeing with what I wrote that you seem to disagree with.

[From Rupert Young (2017.11.25 11.55 )]

(Martin Taylor 2017.11.23.08.39]

...
If there is any modal confusion wouldn't that result in error leading to reorganisation, of the modalities?

Why would it result in error? Error is the difference between a perceptual value and its reference value. How many modalities contribute inputs to the perceptual function seems to me to be irrelevant to whether the system has reorganized to produce actions that move the perception toward its reference value.

Suppose you have something increasing pressure on your chest. But due to modal confusion you perceive that pressure as light. Turning a dimmer switch is going to have no effect on the pressure so the error will not reduce but increase.

True. That does not affect the fact that you would be conscious of changing light levels when you lift the weight off your chest.

When a system first organises it will organise its connections in such a way that its output affects its input; it doesn't need to "know" about modalities, does it?

I wouldn't have thought it would in general, though there may be cases in which it might. In those cases, the identity of the modality would contribute to some perceptual input function to create a controllable perception.

Isn't that the same as what I said?

I hope so.

So, I don't see any reason why modal confusion would occur in the first place. Modality only has meaning in the context of conscious experience.

I fail to see the connection between modal confusion, such as occurs in synaesthesia or the kinds of momentary experiences I mentioned in [Martin Taylor 2017.11.22.11.03], and either error or your conjecture that modality has meaning only to consciousness. Maybe you could expand on those ideas?

The momentary experiences you mentioned don't have an effect on intrinsic error so they are not an issue in the context of reorganisation. If such modal confusion did have a negative effect on intrinsic error the organisation that led to them would be organised out of existence. So, perhaps, the only modal confusion that would develop is that that would not have an effect on intrinsic error.

When we talk about modalities we are just putting labels on signals. To an organism which is not conscious this is not relevant; it is all about the control of signals that maintain low intrinsic error. If the intrinsic error increases the organism will reorganise to control different signals (or control the same signals better) that do reduce that error. It doesn't matter to the organism what those modalities are.

However, there may be additional benefits to consciousness, for better learning associated with directed reorganisation along the lines Bruce outlined. I'm just saying that I don't see the necessity for consciousness for (all) organisms to control successfully, and so, the labelling of the origins of the signals.

Rupert

Exactly! I wouldn't change a word.

I think the cross-purposes might be the distinction between "perception" and "conscious perception as". Over the years, there have been many different suggestions about the role of conscious perception as opposed to perception in control. MoL is based on the idea that you can learn to be conscious of just about any perception you control, and of the reference values you have for those perceptions. You can learn to say (if you are a ballerina, for example) I want my knee to bend and relax over this trajectory, but I keep following that trajectory instead. The ballerina reorganizes, (if she is to become a prima ballerina).

Other ideas about conscious perception have often centred around error, choice, and alerting (so does MoL, which is based on reducing errors produced by conflict). So far as I know, nobody has data that would allow us to refine the role of consciousness within PCT, though it is obviously an important problem. Whether consciousness is necessary for complex control or an epiphenomenon or something that is needed to provide a scaffold for building a hierarchy are all questions that may some day be resolved.

Consider your chest pressure example. Does it matter for control whether it has the same qualia as light intensity? No it does not. Does it matter whether you use a dimmer switch to relieve the pressure? Yes it does. What allows you to determine whether to use a dimmer switch or to remove the weight when you perceive increased light levels beyond what you want? That's a question open to experiment, but if there isn't any conscious perceptual difference, maybe you just have to try one and then the other, just as you must if you don't know whether a door swings inward or outward when you want to pass.

For theory, consciousness is essentially terra incognita. In practice, it's the only terra that is cognita. It's all we know.

Martin

···

[From Rupert Young (2017.11.26 21.20 )]

(Martin Taylor 2017.11.25.11.05]

I think we are talking at cross purposes, because I agree with just about everything you say, without disagreeing with what I wrote that you seem to disagree with.

Well, I'm afraid we'll just have to agree to agree on this one!

Rupert