The taste of lemonade (was Re: Winter school ...)

[MartinT]

[Martin Taylor 2016.09.

[From Rick Marken (2016.09.26.0950)]

Of course it is true that not all perceptions correspond to

something “out there”. A pool of water on a hot road cannot provide
water, as you find when you get closer to it, and nobody can see the
monster that lives under the child’s bed. But it’s interesting you
should bring up the taste of lemonade example, since I use it in the
opposite way in something I am writing. Here’s the current draft
version:
-----------Extract-------
* We put a glass to our lips and perceive the taste of the liquid
as “lemonade�. Something about the way that liquid affects our
perceptual apparatus creates the perception of “lemonade�. But
unless we act to influence the environment, we cannot know whether
“lemonade� is the consequence of drinking liquid, or is dependent
on some “real� properties of the liquid. So we try putting a
liquid from a different source into the glass, and find that it
does not taste of “lemonade�. This would not be the case if
“lemonade� were purely a construction of our perceptual apparatus
that happens when it is exposed to liquid. There is something
special about the “lemonade�-tasting liquid that is not found in
the other liquid.* * To pursue the external reality of the “lemonade� taste, we
might try various actions and see whether they affect the
“lemonade� perception. We might try adding different substances,
such as salt or turmeric, and we would find that the liquid tastes
less of “lemonade�, but if we squeeze a lemon into it it might
taste more of “lemonade�. We might try extracting substances from
the liquid by filtering or distilling it. In other words, we do
what scientists are supposed to do. We experiment. Eventually, we
are in a position to say something about what function of physical
and chemical variables in the liquid make us perceive “lemonade�.
When a liquid has those physical and chemical properties, we can
say that, for us though perhaps not for anyone else, the liquid
has the taste of lemonade.* * The taste of lemonade is not a perception that one can
compare with anyone else’s perception of that taste, other than to
have them drink some of the liquid and say something like: “What
does it taste like to you?� It seems ethereal, not really “out
there�. But compare it with a perception such as the position of a
glass on a table. One can act to move the glass to some other
place, and if the action succeeds in changing one’s perception of
the location of the glass, the glass has some “out there� reality.
One cannot determine whether someone else sees the glass on the
table other than by a related test. If someone else acts to move
the glass back to its original position, one may assume that for
the other person it also has some “out there� reality independent
of you both.* * Can one do something similar with the taste of lemonade, and
thereby determine whether the taste has some “out there� reality
independent of the taster? Yes, one can. The analogy to moving the
glass is to change the ingredients of the liquid in the same way
as we did to determine whether for us a particular organization of
what appeared to be “out there� corresponded with the taste of
lemonade. When we change the ingredients, we ask whether the
resulting liquid tastes like lemonade to the other person. If it
turns out that the same physico-chemical organization that
produces “lemonade� for us also produces “lemonade� for the other,
and that organizations that for us do not also do not for the
other, we can say that this organization of ingredients is the
“real world� construct that corresponds to the perception of
“lemonade� taste. “Lemonade taste� is indeed “out there� in the
liquid.*

------End Extract------

I suppose I should add that the "taste" that is "out there" has much

the same relation to the liquid that carries it as does the
beautiful shape of a Michelangelo sculpture to the marble that
carries it. The sculpture is not the marble, and the liquid is not
the taste.

In the bit of your post I quoted, the word "actually" implies that

you have a God-like omniscience as to what is or is not “out there”.
I do not, and can base my judgments only on the effects of my tests
on my own and others’ (stated) perceptions. If manipulations of the
environment have the same kind of influence on my perceptions as
others say they do on theirs, to me that’s a good indication that
the perceptions correspond to something “out there”. It’s better, of
course, if I can use some test other than relying on what they say,
but what they say is a good start.

Martin

···

          Martin

Taylor (2016.09.24.12.08)–

          MT: Bill may have rejected the idea that the nervous

system operates by encoding and decoding messages, but I
think he had a particular view of what those terms imply.
Every perceptual function decodes the input from the
environment and from imagination.

          RM: I don't think this is the correct way to view the

operation of perceptual functions in the PCT model. In
PCT, perceptual functions “construct” perceptual variables
from sensory input. These constructed perceptual variables
may or may not correspond to entities or variables that
are actually “out there” in the environment. An example of
a perceptual variable that corresponds to something that
is not “out there” but completely constructed by a
perceptual input function is, of course, Bill’s example in
B:CP of the taste of lemonade.

[rsmarken]

[From Rick Marken (2016.09.27.1345)]

···

Martin Taylor 2016.09.

MT: Of course it is true that not all perceptions correspond to

something “out there”…

Â

MT: Â But it's interesting you

should bring up the taste of lemonade example, since I use it in the
opposite way in something I am writing. Here’s the current draft
version:Â

-----------Extract-------
* We put a glass to our lips and perceive the taste of the liquid
as “lemonadeâ€?..*

RM: Yes, you do use it in the opposite way Powers does. I’ll stick with Bill’s way. You are trying to argue that the taste of lemonade is a perception that corresponds to an entity in external physical reality. Powers (B:CP, 2nd Ed, p. 112) uses it to show the opposite. In PCT, the taste of lemonade is a perceptual variable constructed as a vector combination of “…intensity signals generated by sugar and acid (together with some oil smells)”. Powers goes on: “However unitary and real this vector [perception] seems, there is no physical entity corresponding to it”.Â

RM: This is the constructivist view of perception and it is at the heart of an understanding of behavior in terms of control of perception. It is certainly at the heart of the hierarchical model of control, where the perceptions at each level of the hierarchy are constructed from lower level perceptions that are themselves constructed from still lower level perceptions.Â

RM: I see two problems with your non-constructivist view of perception. First, it leads to a conception of perceptions as static objects rather than variables. If the taste of lemonade is an external physical entity, then it is either there or it is not (which, I suppose, is a binary variable, but not a variable that can be controlled very well). In PCT, the perceptions that are constructed, like the taste of lemonade, are variables that can be varied by manipulating the lower level perceptions from which they are constructed; the taste of lemonade, for example, can be manipulated by varying the relative amounts of sugar, acid and oils. So this perceptual variable can be brought to a reference state that is “just right” for you.Â

RM: Second, the non-constructivist approach to perception makes it impossible to do the test for the controlled variable (TCV) properly. The TCV is based on the constructivist view of perception. The goal of the TCV is to construct (mathematically, if possible, or using one’s own perceptual systems) a perceptual variable that is equivalent to the perceptual variable that is controlled by the behaving system. The TCV is likely to fail if one is looking for an objective entity that the system is controlling.Â

Â

MT: In the bit of your post I quoted, the word "actually" implies that

you have a God-like omniscience as to what is or is not “out there”.

RM: In PCT, what is “actually” out there in the environment is taken to be what is described in the models of physics and chemistry: forces, masses, photons, atoms, etc.Â

Best

Rick

–
Richard S. MarkenÂ

“The childhood of the human race is far from over. We
have a long way to go before most people will understand that what they do for
others is just as important to their well-being as what they do for
themselves.” – William T. Powers

          MT: Bill may have rejected the idea that the nervous

system operates by encoding and decoding messages, but I
think he had a particular view of what those terms imply.
Every perceptual function decodes the input from the
environment and from imagination.

          RM: I don't think this is the correct way to view the

operation of perceptual functions in the PCT model. In
PCT, perceptual functions “construct” perceptual variables
from sensory input. These constructed perceptual variables
may or may not correspond to entities or variables that
are actually “out there” in the environment. An example of
a perceptual variable that corresponds to something that
is not “out there” but completely constructed by a
perceptual input function is, of course, Bill’s example in
B:CP of the taste of lemonade.

[Abbott]

[From Bruce Abbott (2016.09.27.1810 EDT)]

Rick Marken (2016.09.27.1345) –

Martin Taylor 2016.09.

MT: Bill may have rejected the idea that the nervous system operates by encoding and decoding messages, but I think he had a particular view of what those terms imply. Every perceptual function decodes the input from the environment and from imagination.

RM: I don’t think this is the correct way to view the operation of perceptual functions in the PCT model. In PCT, perceptual functions “construct” perceptual variables from sensory input. These constructed perceptual variables may or may not correspond to entities or variables that are actually “out there” in the environment. An example of a perceptual variable that corresponds to something that is not “out there” but completely constructed by a perceptual input function is, of course, Bill’s example in B:CP of the taste of lemonade.

MT: Of course it is true that not all perceptions correspond to something “out there”…

MT: But it’s interesting you should bring up the taste of lemonade example, since I use it in the opposite way in something I am writing. Here’s the current draft version:

-----------Extract-------
We put a glass to our lips and perceive the taste of the liquid as “lemonade�..

RM: Yes, you do use it in the opposite way Powers does. I’ll stick with Bill’s way. You are trying to argue that the taste of lemonade is a perception that corresponds to an entity in external physical reality. Powers (B:CP, 2nd Ed, p. 112) uses it to show the opposite. In PCT, the taste of lemonade is a perceptual variable constructed as a vector combination of “…intensity signals generated by sugar and acid (together with some oil smells)”. Powers goes on: “However unitary and real this vector [perception] seems, there is no physical entity corresponding to it”.

BA: An intensity signal generated by sugar would not be perceived as sweet, nor by acid, sour, nor by some oils, a particular set of smells. These consciously perceived qualities depend on those signals activating neurons in the relevant sensory cortices. We have plenty of neural signals, arising from sensory receptors, that produce no conscious perceptions – and no qualia – €“ at all. (Baroreceptors in the aorta are one example; we have no direct sense of blood pressure despite the presence of sensors for it and associated neural circuitry of a blood-pressure control system.)

BA: Beyond that, I think you misunderstand what Martin is saying, and base your criticism on that misunderstanding. Martin is not saying that there is a single entity “out there� in external physical reality that corresponds to the taste of lemonade. He is saying that there is a particular combination of factors “out there� that together stimulate a particular suite of sensory receptors in the right ranges of intensities to produce what we recognize as the taste of lemonade. Thus, whether I experience this combination of sensory experiences or you do, we both recognize and label this experience as “the taste of lemonade.� Furthermore, we could use the test for the controlled variable to determine what needs to be present physically in a glass of water for its taste to be perceived by the drinker as that of lemonade. (Anything that will stimulate the right sensory receptors, in the right proportions of intensities, will do so, even if chemically it does not correspond to what we normally label as lemonade.)

Bruce

[rsmarken]

[From Rick Marken (2016.09.27.1620)]

···

 Bruce Abbott (2016.09.27.1810 EDT)–

RM: …You are trying to argue that the taste of lemonade is a perception that corresponds to an entity in external physical reality. Powers (B:CP, 2nd Ed, p. 112) uses it to show the opposite. In PCT, the taste of lemonade is a perceptual variable constructed as a vector combination of “…intensity signals generated by sugar and acid (together with some oil smells)”. Powers goes on: “However unitary and real this vector [perception] seems, there is no physical entity corresponding to it”.Â

Â

BA:Â An intensity signal generated by sugar would not be perceived as sweet, nor by acid, sour, nor by some oils, a particular set of smells.Â

These consciously perceived qualities depend on those signals activating neurons in the relevant sensory cortices.Â

RM: I know. Neither I (nor Bill in the quoted passage) said anything about how these signals would be perceived consciously. All Bill is saying is that the taste of lemonade is (in theory) constructed from a set of intensity signals whose sources are sugar, acid and some oils.Â

 BA: Beyond that, I think you misunderstand what Martin is saying, and base your criticism on that misunderstanding. Martin is not saying that there is a single entity “out thereâ€? in external physical reality that corresponds to the taste of lemonade. He is saying that there is a particular combination of factors “out thereâ€? that together stimulate a particular suite of sensory receptors in the right ranges of intensities to produce what we recognize as the taste of lemonade.Â

 RM: Great. So he’s actually a constructivist since the “particular suite” of factors (I assume you are referring to environmental variables or their sensory consequences) that makes up the perception must be defined by the nervous system of the person perceiving that particular suite as “the taste of lemonade”.Â

BA: Thus, whether I experience this combination of sensory experiences or you do, we both recognize and label this experience as “the taste of lemonade.â€?Â

RM: But that’s only if your nervous system and mine construct that perception in the same way. Or is he saying that  it is the “particular suite” of environmental variables (factors) itself that determines that it will be perceived as the taste of lemonade. If it’s the latter (which it seems to be) then Martin is a Gibsonian rather than a constructivist. A constructivist is just a Gibsonian who has learned how to model perceptual systems (like my PhD adviser).

Â

BA: Furthermore, we could use the test for the controlled variable to determine what needs to be present physically in a glass of water for its taste to be perceived by the drinker as that of lemonade.Â

RM: I think you might still have trouble because if his view is Gibsonian – if he actually believes that a “particular suite” of factors determines the perception – then he might have trouble accounting for the fact that a different suite of factors tastes like lemonade to different people. This could result from a difference in perceptual functions or, more likely, in differences in the reference for the perception (I might like my suite sweeter than you like yours, for example).

Â

BA: (Anything that will stimulate the right sensory receptors, in the right proportions of intensities, will do so, even if chemically it does not correspond to what we normally label as lemonade.)

RM: Yes, it sounds like Martin is just an inch away from the PCT view. All he has to realize is that it’s the perceptual function that determines the “right” Â sensory receptors to stimulate and how much they should be stimulated. The perceptual function constructs the perception (the taste of lemonade) from the “particular suite” of sensory stimulation. Â

Best

Rick

Â

Bruce

Â

Â

Â

Â

–
Richard S. MarkenÂ

“The childhood of the human race is far from over. We
have a long way to go before most people will understand that what they do for
others is just as important to their well-being as what they do for
themselves.” – William T. Powers

[MartinT]

[Martin Taylor 2016.09.27.23.14]

[From Rick Marken (2016.09.27.1345)]

I suppose that really depends on what you mean by a "physical

entity". Since all we ever have is perception, and no controllable
perception is dependent on the input from only one sensor – one
retinal cone, one auditory hair cell, one taste bud or one anything
else – every perception that we control has the same issue; it is
constructed, whether the perception is of the location of a cursor
relative to a target or the politics of a political party. Every
perception is as real as our ability to control it or a perception
to which it contributes. Either you say that all such variables on
which you can do a Test for the Controlled Variable have an
existence in the outer world or you say that none of them do.
There’s no in-between in which some controllable perceptions
correspond to something “out there” and some do not. Even what Bill
called “intensities” are constructed, since they depend on contrasts
between the outputs of some set of sensory receptors as compared to
what they recently were, and as compared to what their neighbours’
outputs are and were. Are those perceived intensities “out there”?

Yes, exactly. That is what I am saying. It is true of every

perception other than the nerve impulses from individual sensors –
a retinal rod or cone, a hair cell, or a taste bud. Moreover, I say
that for a perception to correspond to something “out there” it must
either itself be controllable or it must be part of the input to a
controllable perception.

What non-constructivist view?

Why? "All perceptions are dynamic relations among the input

variables to the corresponding perceptual functions." [Martin Taylor
2016.09.26.15.15] and…

      CG.

Still, the notion of “system concepts� at level 11 remains
problematic…unless it’s tied to doing. [From
Chad Green (2016.09.26.1448 EST)]

  [Martin Taylor 2016.09.26.15.15] But they are ALL necessarily tied

to doing. See if you can find J.G.Taylor’s book in a local library
(The Behavioral Basis of Perception, Yale UP 1962). It’s the same
in PCT. A perception that isn’t useful in control (i.e. isn’t a
controlled perception or a component of the feedback loop of a
controlled perception) is likely to be reorganized out of
existence. Only by “doing” do perceptions (specific perceptual
functions) continue to exist, though they may have a fleeting life
in transit to a form useful when controlled.

  [MT] At least that's something I have believed for more than half

a century, long before I encountered PCT.

  [MT] As I said, the way I see it (having looked only at the

Wikipedia articles on Enactivism and Enactive interfaces),
enactivism looks as though someone read J.G.Taylor’s work and saw
that it was good, while PCT provides the mechanism that is
missing.[Martin Taylor 2016.09.26.15.15]

Parenthetically, I think it worth mentioning once again that I

developed in 1973 a three-level control model to explain the
difference between the tactile perceptions induced by active touch
(haptic perception, resulting in an object “out there” being
perceived) versus passive touch, when the perception is that a body
part is being touched. (Taylor, Lederman and Gibson “The tactile
perception of texture”, Chapter 12 in Carterette and Friedman (Eds)
"Handbool of Perception III: Biology of Perceptual Systems, Academic
Press 1973)

Why? Is an intensity either there or not there? Is a relationship?

Is a configuration (which is what the taste of lemonade is, after
all)? A category perception may be binary, but even that is
questionable.

Glad you agree.

I'm not sure whether I agree, because I don't know what a

non-constructivist approach is. Nor do I understand the last
sentence. Apart from that, I’m glad you agree with what I said in my
“taste of lemonade” extract. Your words seem to say that you
disagree, but you do a pretty good job of paraphrasing my argument.

My argument for the "out-thereness" of the taste of lemonade is

precisely that you CAN do the TCV on it, and moreover, it can be
collectively controlled, so that the precise mix of ingredients that
most people agree is most lemonade-y can drift over time, just as
can the popular meaning of a word or any other collectively
controlled variable. Idiosyncratic perceptions that are not subject
to the TCV, because no external manipulation can influence them,
can’t be proven to be “out there”. Such perceptions are not
controllable through the external environment, though they may be
controllable in imagination.

All of which are just as much constructed perceptions as is the

“taste of lemonade”. How is a photon – which can be seen as a
particle or a wave or something in between, and which can be
mysteriously entangled with another photon or even an atom so you
can’t know where “it” is, if there is an “it” – be “actually” out
there, more than can the environmental correlate of the directly
perceived taste of lemonade? For a long time, the concept of “atom”
was considered to be a nice fiction that made the numbers come out
right. And as I remember, Bill actually used “mass” in a long ago
CSGnet thread as an example of something that is not “out there” any
more than the taste of lemonade.

But it really doesn't matter what Bill said. He was a wonderful

guide and critic, but I don’t think he ever claimed to be always
correct. Personally, I have always treated his words as like those
of the great Niels Bohr, who apparently said to each new class “You
should treat every one of my statements as a question” or words to
that effect. I hope people treat my words the same way.

Martin

···

Martin Taylor 2016.09.

            MT: Of course it is true that not all perceptions

correspond to something “out there”…

Â

            MT: Â But it's interesting you

should bring up the taste of lemonade example, since I
use it in the opposite way in something I am writing.
Here’s the current draft version:Â

-----------Extract-------
* We put a glass to our lips and perceive the taste of
the liquid as “lemonade�..*

          RM: Yes, you do use it in the opposite way Powers does.

I’ll stick with Bill’s way. You are trying to argue that
the taste of lemonade is a perception that corresponds to
an entity in external physical reality. Powers (B:CP, 2nd
Ed, p. 112) uses it to show the opposite. In PCT, the
taste of lemonade is a perceptual variable constructed as
a vector combination of “…intensity signals generated by
sugar and acid (together with some oil smells)”. Powers
goes on: “However unitary and real this vector
[perception] seems, there is no physical entity
corresponding to it”.

                      MT: Bill

may have rejected the idea that the nervous
system operates by encoding and decoding
messages, but I think he had a particular view
of what those terms imply. Every perceptual
function decodes the input from the
environment and from imagination.

                      RM: I don't think this is the correct way

to view the operation of perceptual functions
in the PCT model. In PCT, perceptual functions
“construct” perceptual variables from sensory
input. These constructed perceptual variables
may or may not correspond to entities or
variables that are actually “out there” in the
environment. An example of a perceptual
variable that corresponds to something that is
not “out there” but completely constructed by
a perceptual input function is, of course,
Bill’s example in B:CP of the taste of
lemonade.

          RM: This is the constructivist view of perception and

it is at the heart of an understanding of behavior in
terms of control of perception. It is certainly at the
heart of the hierarchical model of control, where the
perceptions at each level of the hierarchy are constructed
from lower level perceptions that are themselves
constructed from still lower level perceptions.

          RM: I see two problems with your non-constructivist

view of perception.

          First, it leads to a conception of perceptions as

static objects rather than variables.

          [RM] If the taste of lemonade is an external physical

entity, then it is either there or it is not (which, I
suppose, is a binary variable, but not a variable that can
be controlled very well).

          [RM] In PCT, the perceptions that are constructed,

like the taste of lemonade, are variables that can be
varied by manipulating the lower level perceptions from
which they are constructed; the taste of lemonade, for
example, can be manipulated by varying the relative
amounts of sugar, acid and oils. So this perceptual
variable can be brought to a reference state that is “just
right” for you.

          RM: Second, the non-constructivist approach to

perception makes it impossible to do the test for the
controlled variable (TCV) properly. The TCV is based on
the constructivist view of perception. The goal of the TCV
is to construct (mathematically, if possible, or
using one’s own perceptual systems) a perceptual variable
that is equivalent to the perceptual variable that is
controlled by the behaving system. The TCV is likely to
fail if one is looking for an objective entity that the
system is controlling.

Â

            MT: In the bit of your post I

quoted, the word “actually” implies that you have a
God-like omniscience as to what is or is not “out
there”.

          RM: In PCT, what is "actually" out there in the

environment is taken to be what is described in the models
of physics and chemistry: forces, masses, photons, atoms,
etc.

[rsmarken]

[From Rick Marken (2016.09.28.1405)]

···

Martin Taylor (2016.09.27.23.14) –

MT: I suppose that really depends on what you mean by a "physical

entity". Since all we ever have is perception, and no controllable
perception is dependent on the input from only one sensor – one
retinal cone, one auditory hair cell, one taste bud or one anything
else – every perception that we control has the same issue; it is
constructed, whether the perception is of the location of a cursor
relative to a target or the politics of a political party.

RM: Exactly. And PCT just adds that, in theory, the raw materials for this construction are the sensory effects of environmental variables; and the construction is done by neural networks in in the nervous system that we call perceptual input functions.

MT: Every

perception is as real as our ability to control it or a perception
to which it contributes. Either you say that all such variables on
which you can do a Test for the Controlled Variable have an
existence in the outer world or you say that none of them do.

RM: I don’t think Bill’s point was that there is no outer would basis of perception. It was that there is not necessarily an entity in the outer world (per the models of physics and chemistry) that corresponds to that perception. The taste of lemonade does depend on outer world variables – concentrations of various molecules – but the particular concentrations of molecules we enjoy as lemonade is not a particular entity in the real world. It’s like humidity; we can perceive humidity and measure it’s outer world basis (in terms of temperature and water vapor concentration); but humidity doesn’t exist as a entity out there; just temperature (molecular motion) and water vapor (concentration of H2O).

MT: There's no in-between in which some controllable perceptions

correspond to something “out there” and some do not. Even what Bill
called “intensities” are constructed, since they depend on contrasts
between the outputs of some set of sensory receptors as compared to
what they recently were, and as compared to what their neighbours’
outputs are and were. Are those perceived intensities “out there”?

RM: I would say that they depend on variables that are out there; but even intensity perceptions are functions of these variables.

MT: Yes, exactly. That is what I am saying. It is true of every

perception other than the nerve impulses from individual sensors –
a retinal rod or cone, a hair cell, or a taste bud. Moreover, I say
that for a perception to correspond to something “out there” it must
either itself be controllable or it must be part of the input to a
controllable perception.

RM: I don’t see why a perception must be controllable in order for it to correspond to something out there (by “correspond to” I presume you mean “function of”; a perception corresponds only to the function of external variables computed by the input function). I think there are a lot of things we can perceive – that are functions of external variables – that we can’t (or don’t) control. What is surely true is that in order to be controllable a perception must be a function of variables outside the control system – variables that the system itself can affect.

RM: I’d love to see a reprint, if you have one. Otherwise I think I can get it myself.

MT: My argument for the "out-thereness" of the taste of lemonade is

precisely that you CAN do the TCV on it,

RM: OK, great.

MT: But it really doesn't matter what Bill said. He was a wonderful

guide and critic, but I don’t think he ever claimed to be always
correct.

RM: What Bill said that matters most is his theory and the evidence he mustered to support it. Indeed, he never claimed to be correct; he developed PCT as a proposal regarding how behavior works and he wanted (indeed., implored) people to subject the theory to rigorous test so that the theory could be corrected if necessary. What I object to is all the suggested “improvements” to Bill’s theory that are made without any scientific evidence that suggests that such improvements are necessary. Powers, like Newton, developed the theory; now what we need are tests of the theory. That’s the way to carry on Bill’s legacy; not by revising his theory because it does’t suit your taste but by putting it to the test.

Best

Rick

–
Richard S. Marken

“The childhood of the human race is far from over. We
have a long way to go before most people will understand that what they do for
others is just as important to their well-being as what they do for
themselves.” – William T. Powers

          RM: In PCT, the

taste of lemonade is a perceptual variable constructed as
a vector combination of “…intensity signals generated by
sugar and acid (together with some oil smells)”. Powers
goes on: “However unitary and real this vector
[perception] seems, there is no physical entity
corresponding to it”.

          RM: This is the constructivist view of perception and

it is at the heart of an understanding of behavior in
terms of control of perception. It is certainly at the
heart of the hierarchical model of control, where the
perceptions at each level of the hierarchy are constructed
from lower level perceptions that are themselves
constructed from still lower level perceptions.

MT: Parenthetically, I think it worth mentioning once again that I

developed in 1973 a three-level control model to explain the
difference between the tactile perceptions induced by active touch
(haptic perception, resulting in an object “out there” being
perceived) versus passive touch, when the perception is that a body
part is being touched. (Taylor, Lederman and Gibson “The tactile
perception of texture”, Chapter 12 in Carterette and Friedman (Eds)
"Handbool of Perception III: Biology of Perceptual Systems, Academic
Press 1973)