Perceptual Cartoon

[MartinT]

[Martin Taylor 2017.07.07.23.08]

Or, as I suggested, maybe it's a question of the complexity of the

figure. As I pointed out, all our figures were very simple.
Personally, I can switch voluntarily between the interpretations of
the figures in the cartoon, but much more easily with the old-young
woman than for the simple duck-rabbit. I find it harder to keep the
latter in the form I want it to take. It switches over to the other
form quite quickly even if I don’t want it to. That also happens,
but not to the same extent, for the woman.
In either case, though, what you are looking at is a line drawing,
not a duck, a rabbit, or a woman of any age. I doubt you would be
able to see any of the real objects switch to the other form.
Somehow I doubt that our middle-class housewife subjects or
professional colleagues were sufficiently involved with the air of
that ethos!
But I think it is an important observation when you are thinking
about the mechanism of the ambiguity effect. Just as Fred and Erling
didn’t see the rabbit until you told them, so our subjects listening
to the repeated words didn’t perceive nonsense if we told than that
everything they heard would be English, but the ones told that
sometimes they might hear nonsense did sometimes hear nonsense.
Don’t you think that might have something to do with the
construction of political opinion: “Everything X tells you is a lie”
repeatedly told to one group, while “X tells it like it is, but Y
lies all the time” told repeatedly to the other just might affect
the way the different groups perceive X and Y, might it not, since
all most people have to go on is a very sketchy outline of the
things X and Y pontificate about – like the sketchy nature of the
sketches in the ambiguous cartoon.
The logic here escapes me. It’s like saying that because a mother
has two sons, neither boy has a mother.
What IS demonstrated is that a particular sensory input can lead to
a variety of different perceptions. But I guess that’s not news.
Martin

···

On 2017/07/7 3:06 PM, Richard Marken
wrote:

[From Rick Marken (2017.07.07.1205)]

Martin Taylor (2017.07.06.14.51)]

                        RM:  Yes, this may be related to

reorganization. But it is an interesting
question; how to we voluntarily switch from
seeing the same environmental variables in
one way (as an old women or a duck) and then
another (young woman and rabbit). Clearly we
are switching between perceptual functions.
How is that done and how does that affect
control? I’ll noodle on this a bit and see
what I can come up with. But I love the
cartoon – an ambiguous figure projecting an
ambiguous figure. It not funny ha ha but it
is funny strange.

            MT: I wonder about the word "voluntarily". How

might one test whether it happens? All you would have to
go on would be subjective experience and what the
subject told you.

          RM: Yes, all I can go on is my subjective experience.

And I can definitely switch from one perception to another
voluntarily.

            MT: When I was working on ambiguous figures in the late

60s and early 70s… nobody ever reported being able to
switch voluntarily
https://www.researchgate.net/publication/225324620_Stochastic_processes_in_reversing_figure_perception.

          RM: Interesting. Maybe it's a skill I developed as a

student of perception.

            In other studies in different

perceptual dimensions (hearing repeated words, beep
rhythm, repetitive motion, Necker Cube, among others),
we found that naive subjects without an academic
background did not necessarily see the forms that
classical psychology suggested they should. The Necker
Cube, for example, might be seen in more than ten
different forms by any one subject. Some of them were
quite intrigued by the way it moved and changed…

RM: Well, this was done during the 60s;-)

          RM: One thing that seems to be demonstrated by the fact

that we can see the same physical reality in (at least)
two different ways is that there is no such thing as an
environmental variable (EV) that corresponds to a
perception. In the young/old woman “illusion” what we know
is the exact same environmental reality – the same lines
and shadings on the paper or screen – can be perceived in
two different ways: as the young woman or the old woman.

[MartinT]

[Marti Taylor 2017.07.09.17.08]

That discussion was a long time ago, while Bill was still alive. I

think that the flip-flop and its cousin, the “polyflop” are
essential components of category perception, which is what the
ambiguity is about. [“Polyflop” is my own word for a circuit based
on the flip-flop that has more than two possibilities, only one of
which produces a significant output at any moment. We used hardware
version of “triflops” – three-way p luflops – back in the 60’s for
running psychoacoustic experiments]
It’s interesting that you would say that the reference for one
inhibits the reference for the other, rather than that the
perception of one inhibits the perception of the other. Is there a
testable reason for that choice?
Yes.
Martin

···

Remarkable. I think I endorse most of
what Rick says here (quoted in full below). A couple of minor
comments are interleaved with the quoted text.

   But I don't see how he squares what he says here with his

separate comment [From Rick Marken (2017.07.07.1205)] and [From
Rick Marken (2017.07.09.0945)] that one environmental state cannot
correspond to two or more different perceptual states.

[From Rick Marken (2017.07.09.1040)]

Erling Jorgensen (2017.07.07 1520 EDT)

                  >RM: One thing that seems to be demonstrated

by the fact that we can see the same physical
reality in (at least) two different ways is that
there is no such thing as an environmental
variable (EV) that corresponds to a perception. In
the young/old woman “illusion” what we know is the
exact same environmental reality – the same lines
and shadings on the paper or screen – can be
perceived in two different ways: as the young
woman or the old woman.

                EJ:   This is what makes me think about two

different reference specifications for the
perceptual configuration of lines and shadings, one
summoning the figure of an old woman, the other
summoning the figure of a young woman (or
alternatively, the rabbit or the duck on the
screen). This seems consistent with your subjective
experience, where you say –

                  >RM: Yes, all I can go on is my subjective

experience. And I can definitely switch from one
perception to another voluntarily.

          RM: The reason I don't think this is the right

explanation is because it assumes that the young and old
woman are different states of the same perceptual variable
– different values of the output of the same perceptual
function. And the same perceptual function – at least as
perceptual functions are conceived of in PCT _- can’t
produce different outputs (the young versus the old woman
perception) given the exact same input (the configuration
of lines in the drawing).

                EJ:  Switching between the two

configurations seems to be a matter of switching the
preferred reference for it. That means going up a
level and, with apparent ease, indicating that some
higher level wants the other configuration to be
perceived.

          RM:I agree that the switching is done by a higher level

system but I think what is switched is the perceptual
input to that higher level system. Apparently the higher
level system can only produce its required perception
using either the output of the young woman or old woman
perceptual function, but not both at the same time. I
think this switching could, indeed, be done by switching
the references the higher level system sends to lower
level systems. But it seems like what is being switched
are that the references that are being sent to two
different control systems with two different perceptual
functions, one perceiving and controlling for young woman
and the other perceiving and and controlling for old
woman. These references would act in opposition so that
when one specifies the young woman perception the other
specifies no old woman perception and vice versa.

                  EJ:  Now, being able

to see a given configuration in the first place
seems to require reorganization. For instance, I
initially only saw a) the young woman, b) the old
woman, and c) the duck projected on the screen.
The rabbit escaped me! It is only when you called
attention (seemingly a reorganization phenomenon?)
to ambiguity or reversibility also being on the
screen that I then saw the rabbit.

          RM: Yes, that's a kind of reorganization. But a very

simple one, in the sense that you didn’t need to learn to
perceive in a new way; I 'm pretty sure you were already
able to perceive rabbit caricatures (Bugs Bunny, for
example). I think you didn’t notice the rabbit perception
because it was being suppressed by the higher level system
that could not perceive something that included the duck
and rabbit perception at the same time.

EJ: I think it was in Bill Powers’ article " A
Cybernetic Approach to the Assessment of Children "
that he proposed the order of developing a control
system. If I recall accurately, the perceptual
function has to come first. Until that is
constructed, you just don’t see it (as Fred & I
didn’t, with the reversible figure on the screen.)
Once the perception becomes possible, then one can
call from memory a reference specification for how
one wants to see the perception again.

          RM: Yes, but as I said, I don't think this illusion

required reorganizing (building) new perceptual functions.
I think you already had the required “rabbit caricature”
perceptual function built; it was just being suppressed. I
think what my hint did was allow a reorganization in the
sense that the higher level system suppressing the rabbit
perception was able to unsuppress it long enough to see
that it could be incorporated into the higher level
perception by suppressing the duck.

          RM:This is all just speculation of course. But I do

think that whatever the correct explanation of this
illusion is, it must involve separate perceptual functions
for the two different ways of perceiving the same
environmental situation – the lines and shadings in the
cartoon. Since the same environment can result in two
different perceptions, this illusion is a clear
demonstration that there is no such thing as a complex
environmental variable that corresponds to the perceptions
we control. In PCT, perceptions (perceptual variables) are
functions of environmental variables. Different
functions of the same environmental variables produce
different perceptual variables.

                EJ:  Having seen each figure once, it seems

relatively easy for me to switch to its
alternative. I still notice some tendency for each
set to resolve into one OR the other, so that makes
me wonder about Martin’s proposal for a flip-flop
way to model the respective perceptual input
functions.

RM: I didn’t catch Martin’s flip-flop model

          but my own guess about why the reversals happen does

imply a flip-flip type of circuit organization, where
activation of the reference for the perception controlled
by one control system suppresses the reference for the
perception controlled by the other.

                EJ: But the initial reorganization seemed more

effort-ful. I don’t think it was exactly random
reorganization and selective retention, because it
was guided by your suggestion of where to look. So
it seems others can have some role in proposing
possible perceptual-references to look for. And if
I still had trouble, I suppose you might have said
something like, ‘What happens when you consider that
the animal is looking the other way?’ That seems to
be what a teacher does, in general terms: propose
ways to view things. They can’t guarantee that the
student’s perceptions will arrive there, only point
them in a certain direction.

RM: Yes, good teaching technique!

                EJ:  Nice find, for this doubly ambiguous

figure!

RM: Thanks. Seemed like a rather relevant cartoon.

[MartinT]

[From Fred
Nickols (2017.07.08.0520 ET)]

Â

        One minor

nit, Martin. Rick didn’t tell me there was a rabbit in the
cartoon. He asked me if I noticed anything interesting
about the project image. That led me to look at it and tilt
my head. That’s when I saw the rabbit.

Â

Fred

Â

[MartinT]

[Martin Taylor 2017.07.11.09.41]

This is all getting very weird.

From Bruce Nevin (2017.07.10. ET)]

Rick Marken (2017.07.10.1130)–

Why?

We are talking only about a theoretical construct, a simple control

loop, are we not?

![BasicControlDiagram1.jpg|659x473](upload://qLqerwT0zME3AjoC3R6MCnpv24Z.jpeg)

The question that seems to have been at issue is simply whether what

I have called “Sensory input = s”, known as “q.i” by Powers, is the
same variable as the variable I have labelled “Perception = p”. The
answer seems clear fro inspection of the diagram.

If the "Perceptual function" "xP" is a simple pass-through that

imposes no delay, then p = s (a.k.a. “q.i”). Otherwise p≠s.

Is any further complication warranted?

Martin

···

                      Fred Nickols

(2017.07.10.1316 ET)

                    FN: Why is the observer’s

perception of q.i. a fact and my perception of
it is theoretical?

              RM: Because the observer

is not you. So your controlled perception is a theory
for the observer; it’s a fact for you.

            BN> And the state of the

environment is a theory represented for you by the fact
of your perception, and it is a theory represented for
the other party in the transaction of the TCV by the
fact of their perception.

BN> Some of the confusion
is that q.i is represented as though it is in the
environment. To diagram the location of q.i requires the
experimenter to be included in the diagram.

[MartinT]

[Martin Taylor 2017.07.12.17.12]

[From Bruce Nevin (2017.07.12.16:17 ET)]

            Rick Marken

(2017.07.12.1245) –

    So you and Martin agree that we know about the environment apart

from our perceptions?

      No, that's unfair. Rather, I think that in both cases the

diagram is a theoretical statement about control, about
control-system structure of the organism, and about the
environment. About the environment the diagram is a
theoretical assertion that our perceptions are in fact the
realities that we perceive

That's not my theoretical assertion, as I have pointed out on quite

a few occasions over the years and as recently as elsewhere today.
My assertion is more subtle. It is that either we are solipsists
(meaning I am imagining all of you as well as my own body) or there
is a Real World (RW) or which we can know only what it presents to
our sensory organs.

The possibilities for that RW are unbounded. Maybe hordes of

nano-elves keep track of our muscle twitches (if muscles exist) and
pass messages to other nano-elves that tweak our sensors so as to
mimic the effects of our muscles on the unreal world we perceive to
be out there. Unlikely, but possible. It’s much more likely that
when we push something and we feel and see it move, there is
something in RW that we are acting on and perceiving changing its
location as we act.

When we perceive by way of our senses something that we can control

by our actions, of whatever kind the perception and the action might
be, the best bet is that it exists out there, in the sense that
there is an RW entity of which what we see out there is a projection
through our perceptual functions. RW doesn’t allow us to perceive
just anything for which we might have produced a perceptual
function. Nor does RW allow a variable (a perception) produced by a
perceptual function to take on any old arbitrary value. The value
you get out of a perceptual function is what that function produces
from the sensory inputs allowed by RW (possibly with some
augmentation from imagination). The more imagination comes into
play, the less likely it is that we are perceiving and controlling
some real property of RW.

I have an aphorism: "The Perceptual Function proposes; the Real

World disposes." Our “reality” isn’t all there is, nor is it
necessarily accurate about what it produces, but it is what RW
allows.

      (e.g some of the collectively controlled perceptions of

physics, labeled here v1, v2 … vn). We don’t think of it as
a theoretical statement because it’s just common sense. As a
theoretical statement that’s fine, as long as we are clear
about what we are saying. Consistent with that, then, you
should be able to accept Martin’s CEV as being in the
environment in the same sense.

Yes, so I think. In a parallel thread last year, I never could

understand what gave some variables privileged status as being in
the environment, while others were not allowed to be. To me they are
all in the same boat, as projections of RW through the filter of the
perceptual functions. This is equally true at all levels of
perception, though the contribution of imagination is likely to
increase, and the “reality” of what is perceived likely to decrease,
the higher up the hierarchy you look.

Martin

[MartinT]

[Martin Taylor 2017.07.12.23.52]

Yes. I was looking at it from the angle that the sensory data made

available by RW is the limit to the data on which perceptual
functions and hierarchies can be built, whereas you are looking at
it from the viewpoint of an already built hierarchy of control
units. Both viewpoints are valid, and complementary.
I also think we are in agreement.
Martin

···

On 2017/07/12 8:32 PM, Bruce Nevin
wrote:

[Bruce Nevin (2017.07.12.20:31 ET)]

Martin Taylor 2017.07.12.17.12–

MMT> there is a Real
World (RW) of which we can know only what it presents to
our sensory organs

        I would say, rather, of

which we can know only what our perceptual hierarchies make
of what our interactions with it present to our sensory
organs.

…

        I think we're in agreement

and I’m preaching to the choir. I’ll shut up now and attend
to home matters. Pardon the blather.

/Bruce

[rsmarken]

[From Rick Marken (2017.07.04.1150)]

RM: Some perceptual fun from the New Yorker for the perception freaks out there. This has nothing to do with PCT…or does it?

BestÂ

Rick

BestÂ

Rick

···

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.04.1305)]

···

Fred Nickols (2017.07.04.1542 ET)

Â

FN: I find it interesting that the old woman seems focused on her hands and the young woman seems focused on the image being projected on the wall. I rather suspect they are controlling for different things.

RM: Did you notice anything interesting about the image projected on the wall!

BestÂ

RickÂ

Â

Fred Nickols

Â

From: Richard Marken [mailto:rsmarken@gmail.com]
Sent: Tuesday, July 4, 2017 2:52 PM
To: csgnet@lists.illinois.edu
Cc: Richard Marken rsmarken@gmail.com
Subject: Perceptual Cartoon

Â

[From Rick Marken (2017.07.04.1150)]

Â

RM: Some perceptual fun from the New Yorker for the perception freaks out there. This has nothing to do with PCT…or does it?

Â

BestÂ

Â

Rick

Â

Â

BestÂ

Â

Rick

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.05.0545)]

[From Fred Nickols (2017.07.05.0537 ET)]

FN: The projected image doesn’t appear to be a faithful reproduction of the position of the woman’s fingers but I don’t expect precision in a cartoon so I don’t find the projected image to be particularly interesting. What do you have in mind?

RM: It is a reversible figure too!Â

Best

Rick

···

Â

Fred Nickols

Â

From: Richard Marken [mailto:rsmarken@gmail.com]

Sent: Tuesday, July 4, 2017 4:04 PM
To: csgnet@lists.illinois.edu
Subject: Re: Perceptual Cartoon

Â

[From Rick Marken (2017.07.04.1305)]

Â

Fred Nickols (2017.07.04.1542 ET)

Â

FN: I find it interesting that the old woman seems focused on her hands and the young woman seems focused on the image being projected on the wall. I rather suspect they are controlling for different things.

Â

RM: Did you notice anything interesting about the image projected on the wall!

Â

BestÂ

Â

RickÂ

Â

Fred Nickols

Â

From: Richard Marken [mailto:rsmarken@gmail.com]
Sent: Tuesday, July 4, 2017 2:52 PM
To: csgnet@lists.illinois.edu
Cc: Richard Marken rsmarken@gmail.com
Subject: Perceptual Cartoon

Â

[From Rick Marken (2017.07.04.1150)]

Â

RM: Some perceptual fun from the New Yorker for the perception freaks out there. This has nothing to do with PCT…or does it?

Â

BestÂ

Â

Rick

Â

Â

BestÂ

Â

Rick

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

Â

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.06.0835)]

···

Fred Nickols (2017.07.05.0905 ET)

Â

FN: When I was an instructor at the Navy’s instructor training school, I used to use the old/young woman picture to illustrate the concept of “setâ€? – a predisposition to view things a certaain way. As I think about it, that concept ought to be relevant to PCT in some way. Is it perhaps a small instance of “reorganizationâ€? that enabled me to see the rabbit – with a little promppting of course.

RM: Â Yes, this may be related to reorganization. But it is an interesting question; how to we voluntarily switch from seeing the same environmental variables in one way (as an old women or a duck) and then another (young woman and rabbit). Clearly we are switching between perceptual functions. How is that done and how does that affect control? I’ll noodle on this a bit and see what I can come up with. But I love the cartoon – an ambiguous figure projecting an ambiguous figure. It not funny ha ha but it is funny strange.Â

BestÂ

Rick

Â

Â

Fred

Â

From: Richard Marken [mailto:rsmarken@gmail.com]
Sent: Wednesday, July 5, 2017 8:45 AM

To: csgnet@lists.illinois.edu
Subject: Re: Perceptual Cartoon

Â

[From Rick Marken (2017.07.05.0545)]

Â

[From Fred Nickols (2017.07.05.0537 ET)]

FN: The projected image doesn’t appear to be a faithful reproduction of the position of the woman’s fingers but I don’t expect precision in a cartoon so I don’t find the projected image to be particularly interesting. What do you have in mind?

Â

RM: It is a reversible figure too!Â

Â

Best

Â

Rick

Â

Fred Nickols

Â

From: Richard Marken [mailto:rsmarken@gmail.com]

Sent: Tuesday, July 4, 2017 4:04 PM
To: csgnet@lists.illinois.edu
Subject: Re: Perceptual Cartoon

Â

[From Rick Marken (2017.07.04.1305)]

Â

Fred Nickols (2017.07.04.1542 ET)

Â

FN: I find it interesting that the old woman seems focused on her hands and the young woman seems focused on the image being projected on the wall. I rather suspect they are controlling for different things.

Â

RM: Did you notice anything interesting about the image projected on the wall!

Â

BestÂ

Â

RickÂ

Â

Fred Nickols

Â

From: Richard Marken [mailto:rsmarken@gmail.com]
Sent: Tuesday, July 4, 2017 2:52 PM
To: csgnet@lists.illinois.edu
Cc: Richard Marken rsmarken@gmail.com
Subject: Perceptual Cartoon

Â

[From Rick Marken (2017.07.04.1150)]

Â

RM: Some perceptual fun from the New Yorker for the perception freaks out there. This has nothing to do with PCT…or does it?

Â

BestÂ

Â

Rick

Â

Â

BestÂ

Â

Rick

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

Â

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.07.1205)]

···

Martin Taylor (2017.07.06.14.51)]

MT: I wonder about the word "voluntarily". How might one test whether it

happens? All you would have to go on would be subjective experience
and what the subject told you.

 RM: Yes, all I can go on is my subjective experience. And I can definitely switch from one perception to another voluntarily.Â

MT: When I was working on ambiguous figures in the late 60s and early

70s… nobody
ever reported being able to switch voluntarily
https://www.researchgate.net/publication/225324620_Stochastic_processes_in_reversing_figure_perception.

RM: Interesting. Maybe it’s a skill I developed as a student of perception.Â

In other studies in different perceptual dimensions (hearing

repeated words, beep rhythm, repetitive motion, Necker Cube, among
others), we found that naive subjects without an academic background
did not necessarily see the forms that classical psychology
suggested they should. The Necker Cube, for example, might be seen
in more than ten different forms by any one subject. Some of them
were quite intrigued by the way it moved and changed…

RM: Well, this was done during the 60s;-)Â

RM: One thing that seems to be demonstrated by the fact that we can see the same physical reality in (at least) two different ways is that there is no such thing as an environmental variable (EV) that corresponds to a perception. In the young/old woman “illusion” what we know is the exact same environmental reality – the same lines and shadings on the paper or screen – can be perceived in two different ways: as the young woman or the old woman.Â

RM: The idea of an EV seems to require that the same environment be configured as a young woman when you perceive a young woman and as an old woman when you perceive an old woman. But the environment is exactly the same in the two cases. The difference between the perceptions (according to PCT) is that the same environment is seen via different perceptual functions; a perceptual function that produces a perception of a young woman in one case and a perceptual function that produces a perception of an old woman in the other.Â

RM: These perceptual functions determine which aspects of the same environment will be used – and how they will be used – to construct a perception. In order for an observer to determine what another organism is perceiving the observer must be able to perceive the environment via a perceptual function that constructs perceptions in the same way as the those in the organism under observation. This is what is done when an observer tries to determine what perceptual variables another person is controlling. For example, in the coin game the observer (E) is trying to perceive the arrangement of the coins in the same way S does. E can tell that s/he is perceiving the coins in the same way as S when all disturbances to that perception are countered.Â

RM: Both E and S will experience the perception that is found to be controlled as being “out there” in the environment. For example, if S is trying to perceive an ordered sequence of dates on the coins, this sequence of dates on the coins will appear to both E and S as being “out there” in the environment. But, in fact, that perception exists only in the brains of E and S. The physical basis of these perceptions is, indeed, out in the environment of both E and S but the perception of coins arranged according to an ordered sequence of dates is not. And as in the case of the ambiguous figure, the exact same physical reality could be perceived quite differently – as a line of coins, for example. So there is no CEV that corresponds to the perceptual variable being controlled.Â

BestÂ

Rick

One very interesting finding was that for figures such as the Necker

cube and the others mentioned above, if you plotted T, the number of
different transitions (Form i to Form j) ,against N, the number of
different forms reported so far, the fit was in most cases
extremely close to T = kN(N-1), sometimes so close that it was hard
to distinguish the dots representing data from the curve
representing the function. The curve is what you would expect if a
transition from Form i was equally likely to go to any previously
observed form, if it didn’t go to a new one.

We used this result in another experiment that demonstrated that

what people perceived (at least in this experiment) depended on what
they expected to perceive – something that is PCT-relevant.

In this later experiment, we asked subjects to listen to a sound

pattern repeated many times at a fixed interval. The sound patterns
were words that were clipped from a word or phrase of 8 syllables
(one I remember was “We ate some bananas today”). The clips were 4,
2, or 1 syllable long, always being of real words, but no matter
what the length of the clip, the repetition interval was the same.

We told all the subjects that the sounds they heard would be

changing very subtly, and gave them examples of clear changes (“We
aint so mad Anna today”). Some subjects were told that however the
sound changed, it would always be real words, while the other was
given an example including some nonsense gabble and told that some
of what they hear would be real words, and some wouldn’t.

Both groups of subjects again produced data that fitted the T =

kN(N-1) curve. But the “real words” group produced substantially
fewer forms and transitions than the “could be nonsense” group.

Question: *        Did the "real words" subjects

perceive some nonsense and simply not report it, or did they not
perceive any nonsense*?

Answer: If they had heard nonsense but didn't report it because the

experimenters had said they shouldn’t, they would have heard
transition sequences such as “real word i” → “nonsense k” →
“real word j”, which they would have reported as “realword i” →
“real word j”, and this would mean that the number of transitions
they reported would have been almost as many as the number reported
by the other group, which was not the case.

We concluded that in this ambiguous listening situation, you

perceive what you expect to perceive. By extrapolation (perhaps
unwarranted), I suggest that this is true of much of what we see,
hear, or (in my CSGnet expeerience particularly) read.

If you want to see a couple of the papers, they are here:

https://www.researchgate.net/publication/9572813_Verbal_transformations_and_an_effect_of_instructional_bias_on_perception
and
https://www.researchgate.net/publication/9478058_Transformations_of_perception_with_prolonged_observation

I didn't re-read them when writing the above, so you may find some

minor details differ.

Martin

–

          RM: Â Yes, this may be related to reorganization. But it

is an interesting question; how to we voluntarily switch
from seeing the same environmental variables in one way
(as an old women or a duck) and then another (young woman
and rabbit). Clearly we are switching between perceptual
functions. How is that done and how does that affect
control? I’ll noodle on this a bit and see what I can come
up with. But I love the cartoon – an ambiguous figure
projecting an ambiguous figure. It not funny ha ha but it
is funny strange.Â

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.09.0945)]

···

Martin Taylor (2017.07.06.14.51)–

MT: I wonder about the word "voluntarily". How might one test whether it

happens? All you would have to go on would be subjective experience
and what the subject told you.

RM: I agree that it would be tough to test. But I can definitely do the reversals voluntarily.Â

MT: When I was working on ambiguous figures in the late 60s and early

70s, most of the time neither experimenters nor subjects reported a
consistent ability to see one or the other figure.

RM: I don’t see how you could be familiar with (and, indeed, studied) the phenomenon of ambiguous figures and at the same time believe that there are environmental correlates of controlled perceptions called CEVs. Ambiguous figures seem to illustrate clearly that such CEVs could not possibly exist, at least given the model of perception used in PCT. As the young man said to Father William “Pray, how did you manage to do it?”

BestÂ

Rick

Â

I suppose it does

happen, maybe with some kinds of figures and not with others?
Certainly with the one we analyxed and found that the switching
behaved as though a well-defined number of individual perceiving
units switched at random, and the reported switch depended both on
the history of what had been perceived and the numerical balance
between the two states of the individual perceiving units, nobody
ever reported being able to switch voluntarily
https://www.researchgate.net/publication/225324620_Stochastic_processes_in_reversing_figure_perception .

In other studies in different perceptual dimensions (hearing

repeated words, beep rhythm, repetitive motion, Necker Cube, among
others), we found that naive subjects without an academic background
did not necessarily see the forms that classical psychology
suggested they should. The Necker Cube, for example, might be seen
in more than ten different forms by any one subject. Some of them
were quite intrigued by the way it moved and changed, when the
physical presentation didn’t move at all. We found it very difficult
to produce an ambiguous presentation that would be seen in only two
ways by a naive subject. We did not use figures such as the
rabbit-duck and the old-young woman, because they seemed too complex
geometrically for our purposes.

One very interesting finding was that for figures such as the Necker

cube and the others mentioned above, if you plotted T, the number of
different transitions (Form i to Form j) ,against N, the number of
different forms reported so far, the fit was in most cases
extremely close to T = kN(N-1), sometimes so close that it was hard
to distinguish the dots representing data from the curve
representing the function. The curve is what you would expect if a
transition from Form i was equally likely to go to any previously
observed form, if it didn’t go to a new one.

We used this result in another experiment that demonstrated that

what people perceived (at least in this experiment) depended on what
they expected to perceive – something that is PCT-relevant.

In this later experiment, we asked subjects to listen to a sound

pattern repeated many times at a fixed interval. The sound patterns
were words that were clipped from a word or phrase of 8 syllables
(one I remember was “We ate some bananas today”). The clips were 4,
2, or 1 syllable long, always being of real words, but no matter
what the length of the clip, the repetition interval was the same.

We told all the subjects that the sounds they heard would be

changing very subtly, and gave them examples of clear changes (“We
aint so mad Anna today”). Some subjects were told that however the
sound changed, it would always be real words, while the other was
given an example including some nonsense gabble and told that some
of what they hear would be real words, and some wouldn’t.

Both groups of subjects again produced data that fitted the T =

kN(N-1) curve. But the “real words” group produced substantially
fewer forms and transitions than the “could be nonsense” group.

Question: *        Did the "real words" subjects

perceive some nonsense and simply not report it, or did they not
perceive any nonsense*?

Answer: If they had heard nonsense but didn't report it because the

experimenters had said they shouldn’t, they would have heard
transition sequences such as “real word i” → “nonsense k” →
“real word j”, which they would have reported as “realword i” →
“real word j”, and this would mean that the number of transitions
they reported would have been almost as many as the number reported
by the other group, which was not the case.

We concluded that in this ambiguous listening situation, you

perceive what you expect to perceive. By extrapolation (perhaps
unwarranted), I suggest that this is true of much of what we see,
hear, or (in my CSGnet expeerience particularly) read.

If you want to see a couple of the papers, they are here:

https://www.researchgate.net/publication/9572813_Verbal_transformations_and_an_effect_of_instructional_bias_on_perception
and
https://www.researchgate.net/publication/9478058_Transformations_of_perception_with_prolonged_observation

I didn't re-read them when writing the above, so you may find some

minor details differ.

Martin

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.09.1040)]

···

Erling Jorgensen (2017.07.07 1520 EDT)

RM: One thing that seems to be demonstrated by the fact that we can see the same physical reality in (at least) two different ways is that there is no such thing as an environmental variable (EV) that corresponds to a perception. In the young/old woman “illusion” what we know is the exact same environmental reality – the same lines and shadings on the paper or screen – can be perceived in two different ways: as the young woman or the old woman.Â

EJ:  This is what makes me think about two different reference specifications for the perceptual configuration of lines and shadings, one summoning the figure of an old woman, the other summoning the figure of a young woman (or alternatively, the rabbit or the duck on the screen). This seems consistent with your subjective experience, where you say –

RM: Yes, all I can go on is my subjective experience. And I can definitely switch from one perception to another voluntarily.Â

RM: The reason I don’t think this is the right explanation is because it assumes that the young and old woman are different states of the same perceptual variable – different values of the output of the same perceptual function. And the same perceptual function – at least as perceptual functions are conceived of in PCT _- can’t produce different outputs (the young versus the old woman perception) given the exact same input (the configuration of lines in the drawing).Â

Â

EJ: Switching between the two configurations seems to be a matter of switching the preferred reference for it. That means going up a level and, with apparent ease, indicating that some higher level wants the other configuration to be perceived.Â

RM:I agree that the switching is done by a higher level system but I think what is switched is the perceptual input to that higher level system. Apparently the higher level system can only produce its required perception using either the output of the young woman or old woman perceptual function, but not both at the same time. I think this switching could, indeed, be done by switching the references the higher level system sends to lower level systems. But it seems like what is being switched are that the references that are being sent to two different control systems with two different perceptual functions, one perceiving and controlling for young woman and the other perceiving and and controlling for old woman. These references would act in opposition so that when one specifies the young woman perception the other specifies no old woman perception and vice versa.Â

Â

EJ: Now, being able to see a given configuration in the first place seems to require reorganization. For instance, I initially only saw a) the young woman, b) the old woman, and c) the duck projected on the screen. The rabbit escaped me! It is only when you called attention (seemingly a reorganization phenomenon?) to ambiguity or reversibility also being on the screen that I then saw the rabbit.Â

RM: Yes, that’s a kind of reorganization. But a very simple one, in the sense that you didn’t need to learn to perceive in a new way; I 'm pretty sure you were already able to perceive rabbit caricatures (Bugs Bunny, for example). I think you didn’t notice the rabbit perception because it was being suppressed by the higher level system that could not perceive something that included the duck and rabbit  perception at the same time.Â

EJ: I think it was in Bill Powers’ article “A Cybernetic Approach to the Assessment of Children” that he proposed the order of developing a control system. If I recall accurately, the perceptual function has to come first. Until that is constructed, you just don’t see it (as Fred & I didn’t, with the reversible figure on the screen.) Once the perception becomes possible, then one can call from memory a reference specification for how one wants to see the perception again.Â

RM: Yes, but as I said, I don’t think this illusion required reorganizing (building) new perceptual functions. I think you already had the required “rabbit caricature” perceptual function built; it was just being suppressed. I think what my hint did was allow a reorganization in the sense that the higher level system suppressing the rabbit perception was able to unsuppress it long enough to see that it could be incorporated into the higher level perception by suppressing the duck. Â

RM:This is all just speculation of course. But I do think that whatever the correct explanation of this illusion is, it must involve separate perceptual functions for the two different ways of perceiving the same environmental situation – the lines and shadings in the cartoon. Since the same environment can result in two different perceptions, this illusion is a clear demonstration that there is no such thing as a complex environmental variable that corresponds to the perceptions we control. In PCT, perceptions (perceptual variables) are functions of environmental variables. Different functions of the same environmental variables produce different perceptual variables.Â

EJ: Having seen each figure once, it seems relatively easy for me to switch to its alternative. I still notice some tendency for each set to resolve into one OR the other, so that makes me wonder about Martin’s proposal for a flip-flop way to model the respective perceptual input functions.Â

RM: I didn’t catch Martin’s flip-flop model but my own guess about why the reversals happen does imply a flip-flip type of circuit organization, where activation of the reference for the perception controlled by one control system suppresses the reference for the perception controlled by the other. Â

EJ: But the initial reorganization seemed more effort-ful. I don’t think it was exactly random reorganization and selective retention, because it was guided by your suggestion of where to look. So it seems others can have some role in proposing possible perceptual-references to look for. And if I still had trouble, I suppose you might have said something like, 'What happens when you consider that the animal is looking the other way?' That seems to be what a teacher does, in general terms: propose ways to view things. They can’t guarantee that the student’s perceptions will arrive there, only point them in a certain direction.Â

RM: Yes, good teaching technique!Â

EJ:Â Nice find, for this doubly ambiguous figure!Â

RM: Thanks. Seemed like a rather relevant cartoon. Â

BestÂ

Rick

All the best,Â

Erling

---

Disclaimer: This message is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If the reader of this message is not the intended recipient, or the employer or agent responsible for delivering the message to the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately by telephone and delete the material from your computer. Thank you for your cooperation.

Â

–
Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.09.1810)]

[Martin Taylor 2017.07.09.17.08]

MT: Remarkable. I think I endorse most of what Rick says here (quoted in full below). A couple of minor comments are interleaved with the quoted text.

MT: But I don't see how he squares what he says here with his separate comment [From Rick Marken (2017.07.07.1205)] and [From Rick Marken (2017.07.09.0945)] that one environmental state cannot correspond to two or more different perceptual states.

RM: I said that I didn't think that the two perceptual states of the ambiguous figure could be two states of the same perceptual variable. Erling had proposed that the two perceptual states of the figure that we see (young and old woman) result from changing references. I took him to mean changing the reference for the same perceptual variable, which would imply a single perceptual function producing the different states of the figure. And I see no way a single perceptual function can produce different states of the same environmental (physical) variables. So I proposed that the two states of the figure are perceived by two different perceptual functions, one function producing the young woman perception and the other producing the old woman perception. Â

but my own guess about why the reversals happen does imply a flip-flip type of circuit organization, where activation of the reference for the perception controlled by one control system suppresses the reference for the perception controlled by the other.Â

MT: It's interesting that you would say that the reference for one inhibits the reference for the other, rather than that the perception of one inhibits the perception of the other. Is there a testable reason for that choice?

RM: The reason for the choice is that I tend to be biased against inputs knowing what should be done with themselves; seems to much like input guidance. I think the guidance -- in this case the guidance to not perceive the young and old woman at the same time -- should come from higher level control systems that would have a reason for not wanting this to happen. This idea should be testable but I can't think of how right now. At the moment I am less interested in why only one state of the ambiguous figure is seen at a time than what the fact that that happens says about the idea that the controlled perception, p, and its environmental correlate, q.i, are two separate entities. What I believe it says is that they are not separate entities. They are the same perceptual variable -- the controlled variable -- seen from different points of view, that of the control system (p) and that of the observer of the control system (q.i).Â
BestÂ
Rick
 >

···

EJ: But the initial reorganization seemed more effort-ful. I don't think it was exactly _random_ reorganization and selective retention, because it was guided by your suggestion of where to look. So it seems others can have some role in proposing possible perceptual-references to look for. And if I still had trouble, I suppose you might have said something like, 'What happens when you consider that the animal is looking the other way?' That seems to be what a teacher does, in general terms: propose ways to view things. They can't guarantee that the student's perceptions will arrive there, only point them in a certain direction.Â

RM: Yes, good teaching technique! >>>

EJ:Â Nice find, for this doubly ambiguous figure!Â

RM: Thanks. Seemed like a rather relevant cartoon. Â

Yes.

Martin

--
Richard S. MarkenÂ
"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.10.0950)]

···

Fred Nickols (2017.07.10.0723 ET)

Â

FN: Rick: I’m a little confused by your last paragraph below.

RM: This is it, I presume:

RM: The reason for the choice is that I tend to be biased against inputs knowing what should be done with themselves; seems to much like input guidance. I think the guidance – in this case the guidance to not perceive the young and old woman at the same time – should come from higher level control systems that would have a reason for not wanting this to happen. This idea should be testable but I can’t think of how right now. At the moment I am less interested in why only one state of the ambiguous figure is seen at a time than what the fact that that happens says about the idea that the controlled perception, p, and its environmental correlate, q.i, are two separate entities. What I believe it says is that they are not separate entities. They are the same perceptual variable – the controlled variable – seen from different points of view, that of the control system (p) and that of the observer of the control system (q.i).Â

Â

FN: Let’s say I’m the control system in question, looking at the cartoon. My perception of that cartoon, whether of a young woman or an old lady, is p. You are the observer, and your perception of the cartoon is q.i. Do I have that correct?

RM: Yes. But when these symbols are used in the PCT diagram, p is the controller’s perception of the controlled variable and q.i is the observer’s perception of the same variable. The point being that the controlled variable is a perceptual variable for both the controller and the observer of the controller. There is no complex environmental variable out there to which p corresponds; p (as well as q.i) is a variable aspect of the physical environmental variables that we never experience directly.Â

RM: The physical environmental variables of which the perceptions q.i and p are presumed to be a function are the theoretical entities of physics and chemistry that we never directly experience. Everything we do experience is presumed to be a function of these variables (or of the sensory effect of these variables). So these physical environmental variables are purely theoretical, as are the variables (and functions) in the control model. The only things that are not theoretical are the observer’s perceptions. So q.i, the observer’s perception  of the controlled variable, is a fact; p, the control system’s perception of the controlled variable, is part of a theoretical explanation of the fact that q.i is controlled. This is another reason why it’s incorrect to say that only p is controlled while q.i is a side effect of control of p. It privileges theory over fact, which is something, up with which I just can’t put.

Best

RickÂ

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.10.1130)]

···

Fred Nickols (2017.07.10.1316 ET)

Â

FN: Why is the observer’s perception of q.i. a fact and my perception of it is theoretical?

RM: Because the observer is not you. So your controlled perception is a theory for the observer; it’s a fact for you.

BestÂ

Rick

Â

Fred (Now I am really confused) Nickols

Â

From: Richard Marken [mailto:rsmarken@gmail.com]
Sent: Monday, July 10, 2017 12:54 PM
To: csgnet@lists.illinois.edu
Subject: Re: Perceptual Cartoon

Â

[From Rick Marken (2017.07.10.0950)]

Â

Fred Nickols (2017.07.10.0723 ET)

Â

FN: Rick: I’m a little confused by your last paragraph below.

Â

RM: This is it, I presume:

Â

RM: The reason for the choice is that I tend to be biased against inputs knowing what should be done with themselves; seems to much like input guidance. I think the guidance – in this case the guidance to not perceive the young and old woman at the same time – should come from higher level control systems that would have a reason for not wanting this to happen. This idea should be testable but I can’t think of how right now. At the moment I am less interested in why only one state of the ambiguous figure is seen at a time than what the fact that that happens says about the idea that the controlled perception, p, and its environmental correlate, q.i, are two separate entities. What I believe it says is that they are not separate entities. They are the same perceptual variable – the controlled variable – seen from different points of view, that of the control system (p) and that of the observer of the control system (q.i).Â

Â

FN: Let’s say I’m the control system in question, looking at the cartoon. My perception of that cartoon, whether of a young woman or an old lady, is p. You are the observer, and your perception of the cartoon is q.i. Do I have that correct?

Â

RM: Yes. But when these symbols are used in the PCT diagram, p is the controller’s perception of the controlled variable and q.i is the observer’s perception of the same variable. The point being that the controlled variable is a perceptual variable for both the controller and the observer of the controller. There is no complex environmental variable out there to which p corresponds; p (as well as q.i) is a variable aspect of the physical environmental variables that we never experience directly.Â

Â

RM: The physical environmental variables of which the perceptions q.i and p are presumed to be a function are the theoretical entities of physics and chemistry that we never directly experience. Everything we do experience is presumed to be a function of these variables (or of the sensory effect of these variables). So these physical environmental variables are purely theoretical, as are the variables (and functions) in the control model. The only things that are not theoretical are the observer’s perceptions. So q.i, the observer’s perception  of the controlled variable, is a fact; p, the control system’s perception of the controlled variable, is part of a theoretical explanation of the fact that q.i is controlled. This is another reason why it’s incorrect to say that only p is controlled while q.i is a side effect of control of p. It privileges theory over fact, which is something, up with which I just can’t put.

Â

Best

Â

RickÂ

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.11.1250)]

···

Bruce Nevin (2017.07.10. ET)]

RM: Excellent. I would just suggest putting the environmental variables explicitly into the environment, like so:

BN>Â Some of the confusion is that q.i is represented as though it is in the environment. To diagram the location of q.i requires the experimenter to be included in the diagram.

RM: I used v’s rather than x’s to represent environmental variables to be consistent with Bill’s notation in the Science paper.Â

BestÂ

Rick

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[bnhpct]

[From Bruce Nevin (2017.07.12.16:17 ET)]

 Rick Marken (2017.07.12.1245) –

So you and Martin agree that we know about the environment apart from our perceptions?Â

No, that’s unfair. Rather, I think that in both cases the diagram is a theoretical statement about control, about control-system structure of the organism, and about the environment. About the environment the diagram is a theoretical assertion that our perceptions are in fact the realities that we perceive (e.g some of the collectively controlled perceptions of physics, labeled here  v1, v2 … vn). We don’t think of it as a theoretical statement because it’s just common sense. As a theoretical statement that’s fine, as long as we are clear about what we are saying. Consistent with that, then, you should be able to accept Martin’s CEV as being in the environment in the same sense.

‘Perception’ has two aspects, the experience and the perceptual signal. Our experience is of things in the environment. Our perceptions as signals … that’s not so clear. It seems that they must be, because we habitually and probably necessarily equivocate between the two aspects of ‘perception’ to paper over the gap.

(Attachment TCV24.jpg is missing)

···

On Wed, Jul 12, 2017 at 3:43 PM, Richard Marken rsmarken@gmail.com wrote:

[From Rick Marken (2017.07.12.1245)]

[Bruce Nevin (2017.07.11.22:27 ET)]

Rick Marken (2017.07.11.1250)–

BN:I’m fine with that, so long as we understand that v.1, v.2, … v.n are someone’s perceptions that are being projected into the environment.

RM: I understand that the v’s represent physical variables in the environment. The letters themselves are perceptions that appear to be out in the environment. Is that what you mean by “perceptions being projected into the environment”? But in the diagram the v’s represent physical variables – the variables assumed by the models of physics and chemistry-- of which perceptions are ultimately a function. I think that’s all that need be understood about them.

Best

Rick

/Bruce

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

On Tue, Jul 11, 2017 at 3:51 PM, Richard Marken rsmarken@gmail.com wrote:

[From Rick Marken (2017.07.11.1250)]

Bruce Nevin (2017.07.10. ET)]

RM: Excellent. I would just suggest putting the environmental variables explicitly into the environment, like so:

BN>Â Some of the confusion is that q.i is represented as though it is in the environment. To diagram the location of q.i requires the experimenter to be included in the diagram.

RM: I used v’s rather than x’s to represent environmental variables to be consistent with Bill’s notation in the Science paper.Â

BestÂ

Rick

–
Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.07.12.1245)]

[Bruce Nevin (2017.07.11.22:27 ET)]

Rick Marken (2017.07.11.1250)–

BN:I’m fine with that, so long as we understand that v.1, v.2, … v.n are someone’s perceptions that are being projected into the environment.

RM: I understand that the v’s represent physical variables in the environment. The letters themselves are perceptions that appear to be out in the environment. Is that what you mean by “perceptions being projected into the environment”? But in the diagram the v’s represent physical variables – the variables assumed by the models of physics and chemistry-- of which perceptions are ultimately a function. I think that’s all that need be understood about them.

Best

Rick

(Attachment TCV23.jpg is missing)

···

On Tue, Jul 11, 2017 at 3:51 PM, Richard Marken rsmarken@gmail.com wrote:

[From Rick Marken (2017.07.11.1250)]

Bruce Nevin (2017.07.10. ET)]

RM: Excellent. I would just suggest putting the environmental variables explicitly into the environment, like so:

BN>Â Some of the confusion is that q.i is represented as though it is in the environment. To diagram the location of q.i requires the experimenter to be included in the diagram.

RM: I used v’s rather than x’s to represent environmental variables to be consistent with Bill’s notation in the Science paper.Â

BestÂ

Rick

–
Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery