Memory in perceptual input (was Website updated)

[Bill_Powers1]

[From Bill Powers (2005.09.02.0749 MDT)]

Rick Marken (2005.09.01.2115) --

While it is true that B:CP shows imagined perceptions at one level branching up to be the inputs to perceptual functions at higher levels, I think this was done to keep the imagination control model of thinking consistent (structurally) with the perceptual control model of behavior. I don't believe it was based on any evidence that when we imagine we change the way we perceive. It may very well be that imagining does change the way we perceive but it will take more than anecdotal stories to convince me that this is the case.

The point of the BCP model is that at levels higher than the level where the imagination connection exists, there is no way to tell whether the higher perception is based on real perceptions, imagined perceptions, or a mix of the two kinds from the lower level. If you're conscious at a given level, you experience real perceptions at that level. But if you go down a level and examine what those perfectly real perceptions are coming from, you may well find that some of the elements which normally make up the higher perception are missing, and are being supplied in imagination. I don't think this is at all unusual. The model in BCP was based largely on my examination of my own experiences, but I don't think I am remarkably different from other people, or that they are that different from me.

Read this stacked-up phrase in the triangle:

···

*
         * *
       * Paris *
     * in the *
   * the Spring *
* * * * * * * *

Most people, on first reading this, perceive it as "Paris in the Spring," not, as it actually says, "Paris in the the spring." What you read at a higher level may not accurately represent what perceptions you remember to be there at a lower level. Yet when you focus on the individual words instead of the familiar string, the extra "the" is obvious. A missing word can even harder detect. (and two missing words can even harder to detect, as in the previous sentence -- not to mention one in this sentence). As you say, an input function designed to respond to a particular word or phrase will respond (though a little less) if some of the word or phrase is missing, but if you then say that the word you just read was spelled correctly, it's your imagination that's inserting the missing letter or word. And sometimes we can imagine that we have explained something, when we have actually omitted an entire sentence -- the explanation.

We can read imagined perceptions into any real perception. In the extreme, this is one of the symptoms called paranoia, where we see hostility in a stranger's eyes, or admiration, or envy, or jealousy. We imagine that we see something there when in fact the stranger isn't even aware of looking at us, and is thinking "I hope that's my bus coming around the corner" (behind us).
-----------------------------------------------
In response to your previous post, no, there is no depth information at all in the conventional Necker Cube. Overlap is not depth information: this is actually a flat figure. Furthermore, exactly the same flat figure would result if it was a shadow of a wire-frame cube held in either of the two orientations. Any perception of depth at all is an illusion, an imagined attribute of what you are actually perceiving. Of course you are actually perceiving a three-dimensional cube. That is, the perception is just as real as the perception of a real three-dimensional cube. However, this real perception is not being generated in the usual way, by assembling lower-level perceptions in which there is binocular disparity, convergence of spatially parallel lines, diminution of texture for more distant parts of the figure, and so on. What most people do (from what I had read) is try to generate an imaginary sense of closeness for one of the corners, and when they succeed, that is enough to make the whole figure look three-dimensional. That is, the higher perceptual input function is fooled into generating that sense of depth for you to experience. It's the same sense of depth you would get from looking at a real wire-frame cube, but obtained in a different way.

I think I've about run out of examples. Convince a man against his will, he's of the same opinion still.

Best,

Bill P.

[Dick_Robertson]

[From Dick Robertson,2005.09.021030CDT]

Bill Powers wrote:

[From Bill Powers (2005.09.02.0749 MDT)]

As I recall your original discussion of "the imagination arrow" in B:CP you argued that the missing perceptual signals could be supplied by memory, where _memory_ meant stored recordings of previous real perceptions. I wonder if someone can come up with an example of a "false memory" in which there is no previously stored real perception, but a quasi perception has been built from stored recordings of similar but not identical perceptions.

Best,

Dick R.

···

Rick Marken (2005.09.01.2115) --

While it is true that B:CP shows imagined perceptions at one level branching up to be the inputs to perceptual functions at higher levels, I think this was done to keep the imagination control model of thinking consistent (structurally) with the perceptual control model of behavior. I don't believe it was based on any evidence that when we imagine we change the way we perceive. It may very well be that imagining does change the way we perceive but it will take more than anecdotal stories to convince me that this is the case.

The point of the BCP model is that at levels higher than the level where the imagination connection exists, there is no way to tell whether the higher perception is based on real perceptions, imagined perceptions, or a mix of the two kinds from the lower level. If you're conscious at a given level, you experience real perceptions at that level. But if you go down a level and examine what those perfectly real perceptions are coming from, you may well find that some of the elements which normally make up the higher perception are missing, and are being supplied in imagination. I don't think this is at all unusual. The model in BCP was based largely on my examination of my own experiences, but I don't think I am remarkably different from other people, or that they are that different from me.

Read this stacked-up phrase in the triangle:

          *
        * *
      * Paris *
    * in the *
  * the Spring *
* * * * * * * *

Most people, on first reading this, perceive it as "Paris in the Spring," not, as it actually says, "Paris in the the spring." What you read at a higher level may not accurately represent what perceptions you remember to be there at a lower level. Yet when you focus on the individual words instead of the familiar string, the extra "the" is obvious. A missing word can even harder detect. (and two missing words can even harder to detect, as in the previous sentence -- not to mention one in this sentence). As you say, an input function designed to respond to a particular word or phrase will respond (though a little less) if some of the word or phrase is missing, but if you then say that the word you just read was spelled correctly, it's your imagination that's inserting the missing letter or word. And sometimes we can imagine that we have explained something, when we have actually omitted an entire sentence -- the explanation.

We can read imagined perceptions into any real perception. In the extreme, this is one of the symptoms called paranoia, where we see hostility in a stranger's eyes, or admiration, or envy, or jealousy. We imagine that we see something there when in fact the stranger isn't even aware of looking at us, and is thinking "I hope that's my bus coming around the corner" (behind us).
-----------------------------------------------
In response to your previous post, no, there is no depth information at all in the conventional Necker Cube. Overlap is not depth information: this is actually a flat figure. Furthermore, exactly the same flat figure would result if it was a shadow of a wire-frame cube held in either of the two orientations. Any perception of depth at all is an illusion, an imagined attribute of what you are actually perceiving. Of course you are actually perceiving a three-dimensional cube. That is, the perception is just as real as the perception of a real three-dimensional cube. However, this real perception is not being generated in the usual way, by assembling lower-level perceptions in which there is binocular disparity, convergence of spatially parallel lines, diminution of texture for more distant parts of the figure, and so on. What most people do (from what I had read) is try to generate an imaginary sense of closeness for one of the corners, and when they succeed, that is enough to make the whole figure look three-dimensional. That is, the higher perceptual input function is fooled into generating that sense of depth for you to experience. It's the same sense of depth you would get from looking at a real wire-frame cube, but obtained in a different way.

I think I've about run out of examples. Convince a man against his will, he's of the same opinion still.

Best,

Bill P.

[Richard_Marken2]

[From Rick Marken (2005.09.02.0900)]

Bill Powers (2005.09.02.0749 MDT)--

Read this stacked-up phrase in the triangle:

         *
       * *
     * Paris *
   * in the *
* the Spring *
* * * * * * * *

Most people, on first reading this, perceive it as "Paris in the
Spring," not, as it actually says, "Paris in the the spring."

How did the lower level system know that it was supposed to imagine no
repeated "the"? You seem to be attributing some remarkable capabilities to
the imagination system; it is able, in real time, to make up for missing (or
to remove superfluous) perceptions that only the higher level system could
know were missing (or superfluous). So your model of the situation above is
that the higher level system perceives the two "the"s and tells the
appropriate lower level system to imagine the absence of one of the "the"s?
If not, what do you actually think is going on here?

As you say, an
input function designed to respond to a particular word or phrase
will respond (though a little less) if some of the word or phrase is
missing,

or repeated

but if you then say that the word you just read was spelled
correctly, it's your imagination that's inserting the missing letter
or word.

I don't see that. If what was on the page was "R ck" and you perceive "Rick"
as the output of the "Rick" detector perceptual function (sans imagination
input) then why do you need imagination to insert the "i"; it's already part
of the perception.

We can read imagined perceptions into any real perception. In the
extreme, this is one of the symptoms called paranoia, where we see
hostility in a stranger's eyes, or admiration, or envy, or jealousy.

I think all this can be explained by top down references coming from memory;
the paranoid is trying to see hostility, etc., and succeeding by producing
perceptions that match this reference.

In response to your previous post, no, there is no depth information
at all in the conventional Necker Cube. Overlap is not depth
information: this is actually a flat figure.

Overlap is counted as a monocular "cue" to depth along with linear
perspective, aerial perspective and so on.

Any perception of depth at
all is an illusion, an imagined attribute of what you are actually
perceiving.

I think you are using "imagined" here vernacularly. The image on the retina
is two dimensional as well so depth could be considered an illusion. But,
then, so could any other perception, like the perception of purple (if you
think of color as being a representation of light wavelength then there is
no wavelength that corresponds to purple). It seems to me that depth, like
the color purple, is a perception. At least, it doesn't sem to me to be
imagined.

However, this real perception [of the 3-D cube] is not
being generated in the usual way, by assembling lower-level
perceptions in which there is binocular disparity, convergence of
spatially parallel lines, diminution of texture for more distant
parts of the figure, and so on.

You can still see a real 3-D wire frame cube with one eye closed or at a
distance, where you only have convergence, texture gradient and the other
monocular "cues". I've seen the Necker reversal in a real wire frame cube
viewed binocularly at a distance of about 3 feet. So it works even when the
binocular information is there.

I think I've about run out of examples. Convince a man against his
will, he's of the same opinion still.

I'm afraid so. I'm not convinced at all.

You say in your post:

The point of the BCP model is that at levels higher than the level
where the imagination connection exists, there is no way to tell
whether the higher perception is based on real perceptions, imagined
perceptions, or a mix of the two kinds from the lower level.

I can't believe that this was the point. The imagination connection is a
switch, thrown voluntarily by an unspecified entity (call it consciousness).
This aspect of the model certainly corresponds to my experience; when I
imagine something,like the sour smell of the milk, it is done voluntarily.
If these voluntary imaginings always influenced higher level perceptions
(unbeknownst to the higher systems) then it seems to me that behaving
(controlling) skillfully would be quite a chore. I do think that we
sometimes err and start controlling an imagined perception; this is
certainly what happens when we wake ourselves by struggling to run from an
imagined foe in a dream. I think this kind of imagination "slip" can happen
when we are waking too; it's like the imagination connection stays in place
while the output is accidentally activated. But the idea that voluntary
imagining (what you do when you are awake) can influence the perceptions
controlled at a higher level -- unbeknownst to those systems -- not only is
something I have still not seen demonstrated but also seems like it would be
very poor system design.

But, again, an experimental demonstration would certainly convince me.

Best regards

Rick

···

--
Richard S. Marken
MindReadings.com
Home: 310 474 0313
Cell: 310 729 1400

--------------------

This email message is for the sole use of the intended recipient(s) and
may contain privileged information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply email and destroy all copies
of the original message.

[Bill_Powers1]

[From Bill Powers (2005.09.02.1235 MDT)]

Rick Marken (2005.09.02.0900) --

I think you are using "imagined" here vernacularly. The image on the retina
is two dimensional as well so depth could be considered an illusion. But,
then, so could any other perception, like the perception of purple (if you
think of color as being a representation of light wavelength then there is
no wavelength that corresponds to purple). It seems to me that depth, like
the color purple, is a perception. At least, it doesn't sem to me to be
imagined.

It's not. It's real, even when you're seeing a perfectly flat figure as having three dimensions. It's just as real as a perception of a 3-D object in the environment.But the experience of real depth can be obtained from imagined inputs instead of real ones. The depth perception isn't imagined, only some of the inputs to the perceptual input function that reports on depth.

You seem to be assuming that imagining is strictly voluntary: if you don't consciously choose to imagine, it doesn't happen. I would disagree with that. I think we can be fooled by things we involuntarily or automatically imagine, not at the level where we are conscious, but at the lower levels from which the elements of higher perceptions come.

Here is a possible approach to investigating this (just some vague ideas). Suppose we ask a person to identify certain aspects of a changing situation. If the person really wants to do this, and if there are no penalties for false positives, I think that ambiguous situations would result in the person's imagining missing information that would yield the wanted perception, and not other missing information that would yield a different perception. Basically, what we do is ask the person to demand certain perceptions from lower-level systems, some of which they can't provide because it's not available from the environment. I'm guessing that when a lower-level system fails to produce the perceptions its reference signal demands, it will go into the imagination mode to correct its own error. Or maybe that would involve putting one or more systems below it into the imagination mode when they fail to produce the perception. That looks like a hypothesis about "what it is that throws the imagination switch." It's as if the higher system is saying to the lower system, "Here's a specification for the perception I need from you. If you can't give it to me by controlling via your environment, then make it up and give it to me anyway." I can't decide which level would be responsible for throwing the switch: the one that demands the perception, or the one that fails to provide it. If it's the one that fails to provide it, then with awareness identified with the higher system, you would not be aware that some of the data from which your perception is drawn is imaginary. That's pretty much how I've been thinking during this discussion.

Any other ideas?

Best,

Bill P.

[Bill_Powers1]

[From Bill Powers (2005.09.02.1257 MDT)]

Dick Robertson,2005.09.02.1030 CDT]

As I recall your original discussion of "the imagination arrow" in B:CP you argued that the missing perceptual signals could be supplied by memory, where _memory_ meant stored recordings of previous real perceptions. I wonder if someone can come up with an example of a "false memory" in which there is no previously stored real perception, but a quasi perception has been built from stored recordings of similar but not identical perceptions.

Imagine your wife with a unicorn's horn sticking out of her forehead. When you have it constructed, put it aside for a while, then later try to bring the image you constructed back to attention. I think you can do it. Of course it would be even better if you made up the image completely instead of following my suggestion. Does this fit your criteria?

Best,

Bill P.

[Richard_Marken2]

[From Rick Marken (2005.09.02.1415)]

Bill Powers (2005.09.02.1235 MDT)--

Here is a possible approach to investigating this (just some vague
ideas). Suppose we ask a person to identify certain aspects of a
changing situation. If the person really wants to do this, and if
there are no penalties for false positives, I think that ambiguous
situations would result in the person's imagining missing information
that would yield the wanted perception, and not other missing
information that would yield a different perception...

It's a start. I'll noodle on it a bit and see if I can come up with an
experiment that might convince me. But I think it's going to be difficult.
How do you tell if a person actually saw the missing "i" in "R ck" from the
person who saw "Rick" without having actually imagined the "i". I think a
better experiment would be the one I suggested, where you have people
voluntarily imagine a low level perception and see if what is imagined has
an observable influence on control of a higher level perception.

with awareness identified with the higher system, you would not be
aware that some of the data from which your perception is drawn is
imaginary. That's pretty much how I've been thinking during this
discussion.

Yes. This helps. I think this model might fit into my model of error. The
higher level system wants to see a particular result, but if it's taking too
long to get that result, the higher level system somehow forces the lower
order system -- the one*(s) that can't give the higher level system what it
wants -- into imagination mode. Maybe an error study is the way to see if
this kind of "fill in" imagination actually occurs.

Best

Rick

···

--
Richard S. Marken
MindReadings.com
Home: 310 474 0313
Cell: 310 729 1400

--------------------

This email message is for the sole use of the intended recipient(s) and
may contain privileged information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply email and destroy all copies
of the original message.

[Dick_Robertson]

[From Dick Robertson, 2005.09.02.1840CDT]

Bill Powers wrote:

[From Bill Powers (2005.09.02.1257 MDT)]

Dick Robertson,2005.09.02.1030 CDT]

As I recall your original discussion of "the imagination arrow" in B:CP you argued that the missing perceptual signals could be supplied by memory, where _memory_ meant stored recordings of previous real perceptions. I wonder if someone can come up with an example of a "false memory" in which there is no previously stored real perception, but a quasi perception has been built from stored recordings of similar but not identical perceptions.

Imagine your wife with a unicorn's horn sticking out of her forehead. When you have it constructed, put it aside for a while, then later try to bring the image you constructed back to attention. I think you can do it. Of course it would be even better if you made up the image completely instead of following my suggestion. Does this fit your criteria?
  
Hm, yes it seems to fit my criteria. I'll schmooze on it for a while.

Best,

Dick R

···

.

[Martin_Taylor2]

[Martin Taylor 2005.09.02.23.10]

[From Rick Marken (2005.09.01.2115)]

I gotta say that I'm really puzzled by this sudden interest in the idea that imagination contributes significantly to perception.

Could it possibly have anything to do with [From Rick Marken (2005.08.19.1140)] commenting on Dag Forssell's "Once around the Loop"?:

I think a more serious problem exists in your description of how memory
affects perception. You start this right off in the second sentence of the
paragraph when you say "...the input function should again be thought of as
a neural network that receives the various [sensory] signals and constructs
interpretations of them using both the current input and signals retrieved
from memory". I don't believe that the PCT model of perception says anything about perceptions being constructed from sensory _and_ memory signals. In all the working PCT models I know of, perceptual signals are constructed only as a function of sensory and/or low order perceptual signals. Yet most of your discussion of the input function assumes that perception is based, at least in part, on memory signals. I think this gives a somewhat misleading idea of the role of perception (and memory) in the PCT model.
Perception is a function of sensory input. Memory is made up of stored
perceptual signals that can be compared to incoming perceptions (in which
case they function as reference signals) or replayed as thoughts
(imaginings).

I'd be tempted to start a parallel Thread based on the last quoted sentence of that 2005.08.19.1140 message, based on my conception that a reference signal is the desired value for some perceptual signal. The fact that a perception can be compared to a memory doesn't make that memory into a reference signal. The new perception might be great, as compared to the terrible memory, couldn't it? The fact that a comparator has a perceptual signal as input doesn't make its other input a reference signal!

But I won't start that Thread -- too busy, even after Sunday!

Martin

[Bill_Powers1]

[From Bill Powers (2005.09.03.0808 MDT)]

Rick Marken (2005.09.02.1415) --

How do you tell if a person actually saw the missing "i" in "R ck" from the
person who saw "Rick" without having actually imagined the "i".

I agree that that would be a poor test; the answer isn't clear.

I think a better experiment would be the one I suggested, where you have people voluntarily imagine a low level perception and see if what is imagined has an observable influence on control of a higher level perception.

Can you voluntarily fool yourself? I'm not sure, though I have seemed to do it once or twice (not that I was remembering the ether smell voluntarily -- I was reminded of it by mention of a hospital).

Since imagining and remembering and knowing are all very closely related phenomena, what you want done should be fairly easy -- in fact it's probably been done in those "priming" experiments. Priming can work only through memory, unless it gets confused with sensory adaptation.

Do you remember that classical example of THE CAT? I think this was done by Boring. In both words, the middle letter is exactly the same: you could call it either an H with the sides leaning in toward each other at the top, or an A with the sides pried apart and separated a little at the top. Of course we would explain this as two perceptual functions responding to less-than-perfect configurations, with no conflict because there are no words spelled TAE or CHT. But suppose we embed the ambiguous letter in a sentence.

The story is that this scribbled note was found in the belongings of a young yuppie who had been looking for an apartment for himself and his girlfriend...

Or the story is that the note was found in the belongings of a young man who had been going to graduate school while his wife worked to support them both.

So what does the note say? (The H of course is the one that could be either an A or an H). Here it is:

HONEY, I GOT THE PHD!.

(The H, of course, is the one that could be either an H or an A).

The prediction would be that the person who is imagining a graduate student will see an H while the one imagining a guy looking for an apartment will see exactly the same configuration as an A, even if you have the person read off the individual letters of the message.

The principle is that the imagined/remembered/known information influences even how we percieve categories -- if they're ambiguous.

Best,

Bill P.

[Richard_Marken2]

[From Rick Marken (2005.09.03.0845)]

Martin Taylor (2005.09.02.23.10) --

Rick Marken (2005.09.01.2115)--

I gotta say that I'm really puzzled by this sudden interest in the idea that imagination contributes significantly to perception.

Could it possibly have anything to do with [From Rick Marken (2005.08.19.1140)] commenting on Dag Forssell's "Once around the Loop"?:

I think a more serious problem exists in your description of how memory
affects perception. ..

Yes. Of course, it had everything to do with that. But I was so much older then; I'm younger than that now. I've learned that perceptual inputs to perceptual functions _can_ be imagined inputs -- or so says the model. More importantly, I've also learned that the fact that there is no evidence at all that this occurs -- that is, that perceptions are regularly (or ever) the result of "fill-in" by imagination -- and that none of the research and modeling that has been done on control of perception has required that imagined inputs be incorporated into the model, is of no absolutely no relevance. The received wisdom, supported by some intriguing just-so stories but no actual research, is that perception is "filled in" by imagination.

This whole discussion has helped me understand why almost no one bothers doing any research on PCT; apparently you discover more by writing about PCT than by going into the lab and studying it. More the fool I for having wasted my time on this research stuff.

Best regards

Rick

···

---
Richard S. Marken
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400

[Bill_Powers1]

[From Bill Powers (2005.09.04.1045 MDT)]

Rick Marken (2005.09.03.0845) –

I’ve learned that perceptual
inputs to perceptual functions can be imagined inputs – or so says the
model. More importantly, I’ve also learned that the fact that there is no
evidence at all that this occurs – that is, that perceptions are
regularly (or ever) the result of “fill-in” by imagination –
and that none of the research and modeling that has been done on control
of perception has required that imagined inputs be incorporated into the
model, is of no absolutely no relevance. The received wisdom,
supported by some intriguing just-so stories but no actual research, is
that perception is “filled in” by imagination.

This whole discussion has helped me understand why almost no one bothers
doing any research on PCT; apparently you discover more by writing about
PCT than by going into the lab and studying it. More the fool I for
having wasted my time on this research stuff.

When you get over being indignant about learning something, consider that
your bases for rejecting supposed instances have all been other models
you have proposed for what is going on, which you can no more support
with evidence than can those who propose that inputs to PIFs can come
from imagination rather than lower systems. I’m getting an impression,
perhaps groundless, that you just don’t like the idea that some elements
of human perceptions, mine and perhaps more specifically yours, might
sometimes be internally manufactured. I, on the other hand, have found
that to be true of myself often enough not to find anything remarkable in
that proposition. I have seen people as liking me simply because I
imagined that they did, and also have found my imagination leading to
false perceptions of dislike. I have imagined that the focus knob on my
telescope was broken because I was actually turning the declination knob
in the dark and imagined that it was the focus knob. I’m always checking
up on things I imagine to be true because I know that sometimes I just
don’t want to go to the trouble of dealing with them if they’re not. The
idea that anyone’s perceptions are always crisp, clear, true renditions
of That Which Is, with no taint of imagination to distort the picture,
just seems laughable to me – either that, or I’m in much worse shape
than anyone else is and am just waking up to the fact, and should be
despairing instead of laughing.
BCP has always shown imagined perceptions from one system passing upward
to the inputs of higher systems. I didn’t get in there and change it.
Martin Taylor made similar complaints about the role of memory in
generating reference signals, but he lost his copy of BCP long ago so
can’t really be blamed (except for being too penny-pinching to buy
another). We all imagine things that are different from what was actually
experienced, and base higher-order perceptions on what we imagine to have
been true rather than on what was true. We wax indignant over things that
never happened, and complain about things that nobody else remembers. If
memory doesn’t alter present-time perception, we had better be looking
for something else, because something certainly does
that.

Best,

Bill P.

[Martin_Taylor2]

[Martin taylor 2005.09.04.17.34]

[From Bill Powers (2005.09.04.1045 MDT)]
Martin Taylor made similar complaints about the role of memory in generating reference signals, but he lost his copy of BCP long ago so can't really be blamed (except for being too penny-pinching to buy another).

Actually, I did, at the Crieff Hills meeting. But it's still in its wrapping, so blame me for losing my Swiss Army knife, instead

Martin

[Martin_Taylor2]

Re: Memory in perceptual input (was Website
updated)
[Martin Taylor 2005.09.04.17.36]

[From Bill Powers (2005.09.04.1045
MDT)]
If memory doesn’t alter
present-time perception, we had better be looking for something else,
because something certainly does that.

I’d go further, and say that if memory didn’t alter present-time
perception, we would be in a pretty precarious situation trying to
control our perceptions! It would be extraordinarily inefficient of
Nature to have allowed complex organisms to evolve in such a way that
past experience did not affect their perceptions of the present.
Evolved systems usually seem to take advantage of whatever information
is available to create their perceptions. Eiminating everything that
happened prior to the current sensory input would be a really major
exception to that geralization. Possible, but in my view highly
unlikely.

I know, it’s only a predisposition on my part to say that for all
their rococo excesses, evolved systems usually work nearly as well as
they could in their ecological niches, but it’s more often nearly true
than it is nearly false. So, if this topic had never come up in the
form it did, the natural presumption would be that memory affects
current perception, and the burden of proof would be on those who
argue that it does not.

Martin

[Bill_Powers1]

[From Bill Powers (2005.09.04.1554 MDT)]

Martin Taylor 2005.09.04.17.36 --

I'd go further, and say that if memory didn't alter present-time perception, we would be in a pretty precarious situation trying to control our perceptions! It would be extraordinarily inefficient of Nature to have allowed complex organisms to evolve in such a way that past experience did not affect their perceptions of the present.

This is an ambiguous statement, so I half agree with it. The half I don't agree with implies that specific past experiences influence the present forms of perceptual input functions.

I think perceptual input functions attain their present forms in two major stages of development. One is the evolutionary stage in which the ability to form PIFs of certain types -- say, sequence detectors -- comes into being in a species. The second is the reorganizational phase of an individual lifetime, in which interactions with the environment determine which of all possible forms the PIFs of the given type will come to have.

The influence I envision for imagination is of neither kind just mentioned. Imagination simply supplies the requested value of a perceptual input to a higher system as if the normal control process had been successful. If you're carrying out a sequence and are not sure one element has occurred, you imagine it and go on with the sequence. I don't mean you do this in a conscious intentional way, but that it happens. You go on as if the missing element had occurred, instead of just waiting forever for it. This is the best course of action often enough that we quite often do it.

There are lots of situations where in order to proceed with actions, it is necessary to assume that the right perception has occurred whether it has or not.

Best,

Bill P.

[Richard_Marken2]

[From Rick Marken (2005.09.05.1750)]

Bill Powers (2005.09.04.1045 MDT)--

When you get over being indignant about learning something, consider that your bases for rejecting supposed instances have all been other models you have proposed for what is going on, which you can no more support with evidence than can those who propose that inputs to PIFs can come from imagination rather than lower systems.

I am also basing it on empirical observation. Particularly, the observation that the flavor of milk was no different when I imagined it sour versus when I imagined it fresh. According to your model, imagined perceptions enter high level perceptual functions as inputs. Smell is a perception that we know is combined with taste to form a higher level perception of flavor. So the smell imagination should have been combined with the taste (from the tongue) to give a different flavor in the two cases of imagination. It didn't. So that is evidence against the idea that imagined perceptions at one level are inputs to perceptual functions at a second level.

I'm getting an impression, perhaps groundless, that you just don't like the idea that some elements of human perceptions, mine and perhaps more specifically yours, might sometimes be internally manufactured.

Yes, that would be a groundless impression since I have always assumed that perceptions are internally manufactured (by perceptual functions) constructions based on sensory input.

I, on the other hand, have found that to be true of myself often enough not to find anything remarkable in that proposition. I have seen people as liking me simply because I imagined that they did

Why not just say that you _perceived_ them as liking you? Saying you _imagined_ that they liked you implies that there was a correct way to perceive the situation. Have you now become a naive realist? Would we perceive everything correctly if it weren't for our pesky imagination?

and also have found my imagination leading to false perceptions of dislike.

But you can only call the perception "false" because subsequent information led you to conclude that what you had perceived was dislike. That doesn't mean you were imagining the dislike when you perceived it. In the Gelb illusion you see a white square illuminated in a dark room; when the lights come on you see the square as black. This is because the perception of brightness is relative. Once you find out that the square is black in a lighted room you could say that seeing it as white was a false perception. But it was not. It was what the square looked like when it was illuminated in a dark room. And it will always look that way -- white -- even if you know that it appears black in a lighted room. Imagination is not required to explain this "false" perception; the explanation is in the workings of the perceptual function itself. I believe that there is no reason to think that this is not true of all the cases of "false" perception that you are suddenly so keen to explain in terms of imagined inputs to perceptual functions.

BCP has always shown imagined perceptions from one system passing upward to the inputs of higher systems. I didn't get in there and change it.

Perhaps it's time to rethink it.

Best

Rick

···

---
Richard S. Marken
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400