Emotions

[Martin_Taylor2]

[From Rick Marken (2004.01.02.1515)]

Martin Taylor (2004.01.02.1621)--

Marc Abrams (2004.01.02.1309)--

Your notion of the ECU
and its environment makes perfect sense. A disturbance to an ECU is a
disturbance, it doesn't make a difference where the disturbance
actually
comes from, its _external_ to the ECU. But I do not believe that this
is the
picture that is given in the PCT literature. I say this because in
showing
HPCT, in Mind Readings for example pg. 137 is a diagram of a three
level
hierarchy. In it the ECU's pass through the 'environment' only at the
first
level and are exposed to disturbances only at that point. They are
_not_
exposed to 'internal' disturbances on the way down the hierarchy from
level
to level.

True, again. But the idea there is simply to show that multi-level
hierarchies can work in the face of external disturbances. I don't
think it would be too hard for Rick to extend his spreadsheet model ,
as a proof of concept, to incorporate disturbances that enter into
the second or third level perceptions (not to be used as a model of
actual human processing). Rick??

Yes, it's easy to incorporate "internal" disturbances into my
spreadsheet hierarchy. But it's really not necessary to do this since
these internal disturbances already exist in the spreadsheet in the
form of the effects of systems at the same level on the perceptions
controlled by other system at that level.

Yes, I know (and knew) that, but for didactic purposes those effects
represent something different. They are (in theory) predictable given
the structure and weights in the hierarchy. I think it would be
useful to have random external disturbances introduced at different
levels. That would show that the hierarchic model is in fact robust
against wide-ranging disturbances (I mean wide-ranging through the
hierarchy), and I'm not aware that such a demonstration exists. Since
in a real organism, disturbances are almost guaranteed to be present
at any level, a demonstration that the hierarchic model handles them
with ease would be quite useful.

Martin

Martin

[Bill_Powers]

[From Bill Powers (2004.01.03.1127 MST)]

Bruce Gregory (2004.01.03.1051)–

As Rick reminded us, the
hierarchical model is capable of accommodating

memory, imagination, and emotion, both as controlled perceptions and
as

disturbances to controlled perceptions. Is it fair to say that the

absence of emphasis on these topics in HPCT is therefore a
reflection

of the lack of data?

It may be more a reflection of the difficulty of doing experiments with
things that are entirely internal to an organism. Memory is probably the
most accessible, since it’s possible to set up experiments that require
retaining some perception in memory. In Demo1, the demonstration of
controlling the pitch of a sound is in part a test of memory. A tone
sounds until the run starts, and then it begins to vary under
disturbance, and the subject must use the mouse to keep the tone at its
original pitch. There is nothing to indicate the original pitch, which
must be remembered and used as the reference pitch for the control
behavior. This could be the basis for a number of experiments with
different variables and different levels of perception. It would be very
easy to do studies of the way in which a remembered reference level of a
controlled quantity behaves – how much it drifts with time, how it’s
affected by other perceptions, how well it can be repeated after a few
minutes, hours, days, or years. Fine material for undergraduate or
graduate theses.
Imagination is more difficult to investigate, since there is no way for
an observer to verify that it’s taking place, much less record the state
of an imagined variable. One approach would be to test the hypothesis
that a perceived variable can’t simultaneously be imagined. If subjects
are asked to imagine a controlled variable being in some specific state
while they are controlling it to remain in a different state,
control should deteriorate during periods of imagining the variable. I’ve
described one experiment of this sort: try to look at the back of your
hand while at the same time imagining, as vividly as you can, the same
hand being held in the same place but palm up. I have great difficulty
with this – if I manage to get a clear imagined picture, the real
perception simply blanks out.

Unfortunately, since the experimenter can’t verify that imagination is
actually taking place as requested, the outcome depends on the subject’s
ability to imagine and report the experience accurately.

Emotion is still more difficult to experiment with. One thing that would
help would be for people to state exactly what they mean by this word. I
tried to create a definition, but nobody else seems to use it, so there’s
no way to know what’s being talked about when someone else mentions the
term. I think it would be very useful for participants here to volunteer
their own definitions, and give real-life examples with detailed
descriptions of specific perceptions of emotions. I know it’s difficult
to maintain an emotion while describing it, but remembered emotions would
do almost as well. If we had some concrete examples to talk about, maybe
more people would be enthusiastic about working up a consensus
definition. The question is, “How do you know when you’re
experiencing an emotion, and what exactly are you
experiencing?”

As long as we just talk about these things in the abstract, not much
progress is likely to be made.

Best,

Bill P.

[Bill_Powers]

[From Bill Powers (2004.01.03.1153 MST)]

Rick Marken (2004.01.02.1515)(or is this one from Martin Taylor?)--

I think it would be useful to have random external disturbances introduced
at different levels. That would show that the hierarchic model is in fact
robust
against wide-ranging disturbances (I mean wide-ranging through the
hierarchy), and I'm not aware that such a demonstration exists.

It wouldn't be easy for disturbances to act directly on perceptions at
levels above the first. About the only thing that would work would be a
chemical injection or an electrical signal from a probe.

However, there are plenty of opportunities for disturbances in the
environment to affect higher-order perceptions, through input channels that
are uncontrolled. In Rick's spreadsheet, all you have to do is eliminate
the control systems for some of the intermediate variables (retaining the
perceptual input functions). Consider the problem of a dog's chasing a cat.
The dog's perception of the cat begins with intensities, produces
sensations, configurations, transitions, and events (all uncontrolled in
any nontrivial sense), and finally becomes a relationship perception, the
cat's relationship to the dog, which is a controlled variable for the dog
(and the cat, too). The cat's running, dodging, and climbing are
disturbances of that relationship as the dog perceives it. The dog opposes
the disturbance by varying its own role in the relationship, the only
aspect of the relationship it can affect by its behavior. The disturbance
first becomes opposable at the relationship level because there are no
control systems in the dog that can alter the cat's appearance or behavior
at the lower levels (other than just not looking at it, which is what I
meant by "trivial").

Best,

Bill P.

[Richard_Marken2]

[From Rick Marken (2004.01.03.1200)]

Rick Marken (2004.01.02.1515)--
Yes, it's easy to incorporate "internal" disturbances into my
spreadsheet hierarchy. But it's really not necessary to do this since
these internal disturbances already exist in the spreadsheet in the
form of the effects of systems at the same level on the perceptions
controlled by other system at that level.

Yes, I know (and knew) that, but for didactic purposes those effects
represent something different. They are (in theory) predictable given
the structure and weights in the hierarchy.

Yes. But they are not predictable by the system controlling the
perception that is disturbed by changes in these other perceptions. The
main cause of unpredictable changes in these disturbing perceptions is
changes in higher level goals. This change can be seen in the
spreadsheet hierarchy by changing the level three goals (from 1 to -1
or vice versa) which are in cells D3-I3 and seeing what happens to the
net disturbance to the perception controlled by level 2 system 1 (what
I called p12) in cell D8. Recall that p12 is defined in the spreadsheet
as

p12 = a1*p11+a2*p21...a6*p61

Since the output of the system controlling p12 only influences the
value of p11, the rest of the perceptual components of p12 are a
disturbance to p12:

d12 = a1*p21+a2*p21...a6*p61

so p12 can be written in the form of the following formula:

p12 = a1*p11+d12

This formula shows that by taking the appropriate weighted sum of p21,
p31...p61 (cells E12-I12 in the spreadsheet) we can see how the
disturbance to p12 (cell D8) changes with changes in references to the
higher level systems. In the version of the spreadsheet I have (and
that I believe is on the net) the appropriate weights for p21,
p31...p61 (E12-I12) are 1,1,-1,-1 1, respectively. If you compute this
weighted sun in some available cell, you will be seeing the value of
the "internal disturbance", d12, to p12. This value changes slightly
over time (due to variations in lower level perceptions that result
from external disturbances) but it stays pretty close to some average
value (the weighted average of the references for the level 1
perceptions).

Large changes in d11 are created by changing higher level references.
For example, when the level 3 references for the logical perceptions
controlled controlled by level 3 systems 2, 3 , and 4 in cells E3-G3
are set to 1,-1, 1, d12 is about 52. When the same level 3 references
are 1,1,1, d12 is 108 and when they are 1,1,-1 d12 is 350. By the way,
I can't seem to get the level 3 systems to stabilize when these level 3
references are -1,-1,-1, indicating that my definition of the level
three (logical) perceptions has led to a conflict that occurs only when
the references for these perceptions are set at certain values. Fixing
this up in a real control system would be a job for the reorganizing
system.

Anyway, the point is that changes in the references for higher level
perceptions (level 3 in this case) change, in unpredictable ways (from
the point of view of the lower level system -- a level 2 system in this
case) the disturbances to the perceptions controlled by lower level
systems. Dealing with such disturbances -- whether they are caused by
variations in the environment ("external" disturbance) or by changes in
higher level references in a control hierarchy (internal disturbances)
-- is what control systems _do_. I don't believe that there is any
reason to deal with external and internal disturbances differently.

Disturbances are disturbances and a properly designed control system --
one that controls properly, keeping its perception matching its
reference -- acts to protect the controlled variable from such
disturbances. Internal disturbances are no more predictable that
external disturbances, at least this is true from the point of view of
the control system itself. Control systems are organized to bring
perceptions to references and maintain them at the reference value,
protected from disturbances, whatever the source of those disturbances.
  A control system that is not resisting disturbances to a controlled
perception is either not yet skilled at controlling that perception or
is in conflict with another control system that is trying to control
the same or a similar perception relative to a different reference.

I think it would be useful to have random external disturbances
introduced at different levels.

That is very easy to do using the spreadsheet. Just add one of the
randomly varying disturbance values in cells D24-I24 to one of the
level 2 perceptions. For example, the formula for the perception I've
been calling p12, which is cell D8, could be changed from:

=IF(D6="*",D7,MMULT(P1i,Pmlt1))

to

=IF(D6="*",D7,MMULT(P1i,Pmlt1)+F24)

The formula for p12 is then MMULT(P1i,Pmlt1)+F24 (the first part of the
IF just allows for p12 to be an imagined value, equal to the value of
the reference for that perception). The MMULT does a matrix
multiplication of the lower level perceptions (P1i, which is the vector
containing p11..p61) by the corresponding weights (Pmlt1, which the
vector containing a11..a61). The resulting perception now has an
externally generated disturbance added (F24). You will see that the
addition of this disturbance causes no problem for the system
controlling p12 (D8); the system keeps the perception under control
just as well as it did when the only disturbance was the "internal"
disturbance consisting of the all perceptions (except level 1
perception 1, p11) which are not influenced by the output of the system
controlling p12.

That would show that the hierarchic model is in fact robust
against wide-ranging disturbances (I mean wide-ranging through the
hierarchy), and I'm not aware that such a demonstration exists.

I agree. And I just described how to do the demonstration using the
spreadsheet. I agree that it is a useful demonstration of the
robustness of control in a control hierarchy.

Best regards

Rick

···

On Saturday, January 3, 2004, at 08:34 AM, Martin Taylor wrote:
---
Richard S. Marken
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400

[Bruce_Gregory4]

[From Bruce Gregory (2004.01.03.2200)]

Bill Powers (2004.01.02.0846 MST)

Do any of you see a coherent theory of emotion or behavior or
experience
arising out of what Marc has been proposing? I will apologize and
subside
if I have misconstrued what is going on. But I can't just let it go on
without voicing my protest.

I guess that depends on whether you think HPCT embodies a coherent
theory of emotion or behavior or experience. As Rick has pointed out,
Marc is saying nothing that cannot be embodied in the spreadsheet
model. In the spreadsheet model as it now exists, perceptual functions
and disturbances are "put in by hand" as physicists say. As I read
Marc, he is trying to explore ways that these functions could
themselves be modeled. Asking for details at this stage of the
investigation may be a bit premature.

Bruce Gregory

"Everything that needs to be said has already been said. But since no
one was listening, everything must be said again.'
                                                                                Andre Gide

"What is hateful to you, do not to your fellow men. That is the entire
Law; all the rest is commentary."

                                                                                The Talmud

[Bill_Powers]

[From Bill Powers (2004.01.03.2210 MST)]

The Mars Rover is safely down and transmitting a "strong signal." What a
nerve--racking 20 minutes that was, on NASA TV! We may get pictures at
11:00 PM Pacific Time: I shall struggle to stay awake.

Bruce Gregory (2004.01.03.2200)--

I guess that depends on whether you think HPCT embodies a coherent
theory of emotion or behavior or experience. As Rick has pointed out,
Marc is saying nothing that cannot be embodied in the spreadsheet
model. In the spreadsheet model as it now exists, perceptual functions
and disturbances are "put in by hand" as physicists say. As I read
Marc, he is trying to explore ways that these functions could
themselves be modeled. Asking for details at this stage of the
investigation may be a bit premature.

That's true. However, I am mainly concerned about the theory of emotion. I
ceased to believe theories I had heard when they seemed to imply that
emotions could respond to events without any need to detect, recognize, and
interpret them. To me that is simply an incoherent theory. This isn't about
details, it's about basic architecture. And it's about defining the
phenomenon called emotion so we start out knowing what we're trying to explain.

So I'd still like to hear what others mean by this term -- what sort of
experience, under what circumstances, etc.

Best,

Bill P.

[Marc_Abrams1]

from [Marc Abrams (2004.01.04.0120)]

[From Bill Powers (2004.01.03.2210 MST)]

And it's about defining the
phenomenon called emotion so we start out knowing what we're trying to

explain.

So I'd still like to hear what others mean by this term -- what sort of
experience, under what circumstances, etc.

Tongue firmly-in-cheek.

Interesting. Coherent, consistent? How can PCT 'explain' a phenomena that is
not yet defined and does not exist in the PCT world. What a wonderful
theory. Cosmologists would eat this up if they knew about it. PCT explains
the existence of everything, all we need to do is define it. I guess PCT is
like article 94 of the code of military justice. Article 94 covers all the
other 'infractions' that every other article might have missed or doesn't
cover. The translation is basically this; if you do something we don't like
we'll find a way to punish you, and I'm not making this up.

Bill, why bother defining something PCT already covers? I find it extremely
curious that every time you trash me and the ideas I have you suddenly
become very interested in the same topic I broached. Btw, how is you 500
node 'perceptual'model going? That was in response to my posting the idea
that the current hierarchy cannot build perceptions as constructed.

I know that these are simply coincidences, but you gotta admit they have a
curious timing to them.

Marc

[Bruce_Gregory4]

[From Bruce Gregory (2004.01.04.1105)]

[From Bill Powers (2004.01.03.2210 MST)]

That's true. However, I am mainly concerned about the theory of
emotion. I
ceased to believe theories I had heard when they seemed to imply that
emotions could respond to events without any need to detect,
recognize, and
interpret them. To me that is simply an incoherent theory. This isn't
about
details, it's about basic architecture. And it's about defining the
phenomenon called emotion so we start out knowing what we're trying to
explain.

I agree. My understanding is that Marc was talking about emotional
signals as being proportional to errors. To me this implied that events
must be detected, recognized, and interpreted before emotional signals
would would be sent, to the adrenal glands, for example. Presumably
this would be what happens in my Traffic Incident example.

Bruce Gregory

"Everything that needs to be said has already been said. But since no
one was listening, everything must be said again."
                                                                                Andre Gide

"What is hateful to you, do not to your fellow men. That is the entire
Law; all the rest is commentary."

                                                                                The Talmud

[Bill_Powers]

[From Bill Powers (2004.01.04.0932 MST)]

Marc Abrams (2004.01.04.0120)--

Interesting. Coherent, consistent? How can PCT 'explain' a phenomena that is
not yet defined and does not exist in the PCT world.

But that is exactly what a theory is for: to explain phenomena that are not
simply built into the theory. I am asking people to describe the phenomena
they mean by the term emotion: what they observe independently of any
theory. That's how we establish what we're trying to explain.

What a wonderful theory. Cosmologists would eat this up if they knew
about it. PCT explains the existence of everything, all we need to do is
define it.

No, it does not, and that is not the purpose of definitions. The question
is, what human experiences are we trying to explain here? We need to
describe and define those experiences in a way that is, to the best of our
abilities, completely independent of any theory we may propose to explain
them. It would be best to describe emotion without using the term emotion
at all, or any of its relatives naming specific emotions or feelings. Once
we can agree what we're trying to explain, we can begin proposing and
arguing about different explanations. This is a required step in any
scientific investigation.

I guess PCT is
like article 94 of the code of military justice. Article 94 covers all the
other 'infractions' that every other article might have missed or doesn't
cover. The translation is basically this; if you do something we don't like
we'll find a way to punish you, and I'm not making this up.

PCT is not like article 94. See how easy it is to prove things? You just
make positive statements of fact and threaten to explode in the face of
anyone who expresses doubts.

Bill, why bother defining something PCT already covers?

Because I want to get agreement that we're even talking about the same
thing when we use the word emotion. Defining the phenomenon is not the same
thing as explaining it. I have offered a PCT explanation for the phenomena
I think of as the meanings of the word emotion, but I'm not sure that
others assign the same meanings to that word. I'm willing to start over,
theoryless, with a simple but adequate description, and obtain a consensus
on the description. Then anyone can offer an explanation; I'll offer mine,
you can offer yours, and anyone else can chime in. But if you're explaining
apples and I'm explaining aardvarks, we will not get very far.

I find it extremely curious that every time you trash me and the ideas I
have you suddenly become very interested in the same topic I broached.

That's kind of weird, Marc. What should I do, talk about something else
instead? What would you say if I did that? And considering that I wrote
that chapter on emotion in about 1971 (when you were _how_ old?), it's kind
of strange that you should say you broached the subject first. You're
jumping into a conversation that started long before you showed up on the
scene.

Btw, how is you 500 node 'perceptual'model going? That was in response
to my posting the idea that the current hierarchy cannot build
perceptions as constructed.

It's coming along fine, and works. I demonstrated a 150-system (not "node")
version of it at the LA meeting in July of 2003, which was an improvement
over the collections of 20 or 30 systems I had tried to work with over the
previous 20 years (memory limitations in the older computers and Pascal
programs prevented going further). Right now it's on the shelf while I try
to figure out how to get it to reorganize properly. Also, I need to set up
a structured environment to see how that affects the way the multiple
systems reorganize. What are you accomplishing along those lines?

Was the LA meeting before or after your post? And was it you or I who first
commented that we do not even know how to model perceptual input functions
that can recognize configurations? I think I have commented on that, not
just recently but over the last 30 years.

I know that these are simply coincidences, but you gotta admit they have a
curious timing to them.

It's your comprehension of temporal sequence that is curious. I think you
start the timing when you first have an idea, and assume that if anyone
else speaks of that subject after that, they're just following along after
you. That's a pretty easy way to set yourself up as a pioneer, but sooner
or later someone is going to remind you that history did not begin with
you. You gotta admit that your suspicious are based on some pretty
egotistical assumptions.

Bill P.

[Bill_Powers]

[From Bill Powers (2004.01.04.1010 MST)]

Bruce Gregory (2004.01.04.1105)–

My understanding is that Marc was
talking about emotional

signals as being proportional to errors. To me this implied that
events

must be detected, recognized, and interpreted before emotional
signals

would would be sent, to the adrenal glands, for example. Presumably

this would be what happens in my Traffic Incident
example.

Yes, and that is also what my theory of emotion proposes (except the
signals are just ordinary error signals, not “emotional”
signals). However, Marc ended up saying, even in discussing your example,
that the pounding of the heart ocurred at the same time as the
cognitive experience of error, thus denying his earlier agreement that
the error came first and caused the physiological responses. He speaks of
emotions as arising from signals ascending the spinal cord, which
“notify” the organ systems at the same time they
“notify” the cognitive systems about something going wrong. You
can see in that proposal a vague suggestion of a hierarchy, but the
timing is wrong, as is the causal sequence. And the amount of knowledge
of the world that has to be built into the lowest levels of organization
is unacceptable to me – just as it was when I read the same ideas as
presented by old-time psychologists…

Best,

Bill P.

[Marc_Abrams1]

[From Marc Abrams (2004.01.04.1208)]

[From Bruce Gregory (2004.01.04.1105)]

I agree. My understanding is that Marc was talking about emotional
signals as being proportional to errors. To me this implied that events
must be detected, recognized, and interpreted before emotional signals
would would be sent, to the adrenal glands, for example. Presumably
this would be what happens in my Traffic Incident example.

My position, and it certainly could change in the face of contrary data, is
that _sometimes_ it does, and _somertimes_ it doesn't need to be
'interpreted'. 'Fear' _does_ and _is_ an interpreted emotion. You do have
some control over the intensity and duration. Pain does _not_ need to be
interpreted and you have very little control over the intensity and
duration. Moreover, we sometimes feel 'good'or 'bad' without having specific
cognitive thoughts in mind. I might ask, on occasion, rhetorically, where
does that feeling come from?. It is simply not a matter of either or here.
It is situation and context dependent. Error does _not_ cause _all_
emotions. It is my position that emotions/feelings are caused by
_everything_ we do. Emotions do not need to be earth shattering in order to
be intrusive.

An additional problem her might be in definition, which is why I was very
careful in pointing out that I was talking about emotions _AND_ feelings. To
me, they are the same. Some might argue that there is a difference between
emotions and physiological feelings. I say there isn't. They both come from
the same places and both are 'caused'by the same things. I think it's purely
a semantical difference.

Marc

[Bruce_Gregory4]

[From Bruce Gregory (2004.01.04.1245)]

Marc Abrams (2004.01.04.1208)

Error does _not_ cause _all_
emotions. It is my position that emotions/feelings are caused by
_everything_ we do. Emotions do not need to be earth shattering in
order to
be intrusive.

Well, that makes it pretty difficult to model, doesn't it? You may well
be right, but I don't see how to incorporate your insight into a
working model.

Bruce Gregory

"Everything that needs to be said has already been said. But since no
one was listening, everything must be said again."
                                                                                Andre Gide

"What is hateful to you, do not to your fellow men. That is the entire
Law; all the rest is commentary."

                                                                                The Talmud

[David_Goldstein1]

[From David Goldstein (2004.01.04.1236 EST)]
[From Bill Powers (2004.01.04.0932 MST)]

Bill and listmates,

I have some questions on the topic of emotions. I will only post a few
of these. If you all think this is a helpful way to proceed, maybe we
could have a good discussion on this important topic.

(1) In defining emotion, are we going to be taking the insider point of
view, or the outsider point of view or both?

David: The PCT view emphasizes the insider view, I think.

If a person asks himself/herself, am I experiencing an emotion, and the
person says no. But outsiders say yes, whose opinion do we go with?

People are sometimes not aware of their emotional state. Outsiders can
observe the nonverbal behavior of the person.

So that the difference of opinion might be related to the difference in
what is being observed.

(2) People undergo development. Emotions are not the same at all ages.
Are we going to be talking about the adult person?

David: I think it is probably a good idea to start with what we think
the adult state of development is. We can not ask a newborn infant
questions.

Piaget and other developmental psychologists have studied the
development of emotions in people. There is a literature.

(3) People in all cultures may not experience emotions the same. Are we
going to try to come up with ideas that apply to all people in all
cultures?

David: This seems like a good thing to do. We would want ideas that
apply to all people in all cultures.

(4) In defining emotion, are we going to be talking about the human
species only, or are we going to be talking about other animals?

David: I would guess that we are talking about people only. However,
people seem to think that other animals display behavior suggesting the
presence of emotions.

(5) In defining emotion, are we going to be talking about normal,
healthy people or are we going to be talking about people with emotional
problems or both?

David: I guess we would want a theory which could explain both.

Enough questions for now.

David
David M. Goldstein, Ph.D.

[Martin_Taylor2]

Re: Emotions
[Martin Taylor 2004.01.04.1311]

[From Bill Powers (2004.01.04.1010
MST)]
However, Marc ended up saying,…
that the pounding of the heart ocurred at the same time as the
cognitive experience of error, thus denying his earlier agreement that
the error came first and caused the physiological
responses.

In mitigation, it seems that the conscious perception of events
in the world is often (always?) delayed by many tens of milliseconds
(maybe even half a second) after the actions begin for their control.
If that’s so, is it not conceivable that the cognitive experience of
error is delayed long enough that the pounding of the heart really
does begin simultaneously with the experience?

The problem may not be with emotion and low-level stuational
analysis, but with the relation between conscious experience and the
perceptions that are unconsciously being controlled, at whatever high
level.

Just a thought, not a thesis.

Martin

[Marc_Abrams1]

from [Marc Abrams (2004.01.04.1336)]

[From Bruce Gregory (2004.01.04.1245)]

Well, that makes it pretty difficult to model, doesn't it?

Yes, but _not_ necessarily impossible. I have a couple of ideas on data
collection that just might prove to be viable.

You may well

be right, but I don't see how to incorporate your insight into a
working model.

Again, I have a couple of ideas but I'm _very_ reluctant to discuss either
of these issues on this forum. I certainly don't want to be treated like I
have been simply because of some ideas I presented. When I go in front of my
Doctoral committee I'll worry about those kinds of things. I don't want or
need the trouble.

Marc

[Marc_Abrams1]

from [Marc Abrams (2004.01.04.1345)]

[Martin Taylor 2004.01.04.1311]

The problem may not be with emotion and low-level stuational
analysis, but with the relation between conscious experience and the
perceptions that are unconsciously being controlled, at whatever high
level.

Just a thought, not a thesis.

And a good one, borne out by research. But Bill is right in a certain sense
as well. There _ARE_ emotions that are of 'cognitive' origin.
My position is that _whatever_ the initial impetus is; Cognitive or
physical, the same physiological processes take place to 'cause' the
_actual_ feelings we have.

Marc

[Marc_Abrams1]

from [Marc Abrams (2004.01.04.1706)]

[From David Goldstein (2004.01.04.1236 EST)]

If you all think this is a helpful way to proceed, maybe we
could have a good discussion on this important topic.

No, I don't think this is helpful

Marc

[Bill_Williams1]

[From Bill Williams 4 January 2004 5:20 PM CST]

Marc,

When David says,

Piaget and other developmental psychologists have studied the
development of emotions in people. There is a literature.

It seems to me that he is making a useful suggestion. Before possibly setting off to invent a wheel, wouldn't it be a good idea to consider what other people have come-up with when they've considered the question of emotions?

You might not be interested in doing a literature review, but I'm a bit vague about the issue of whether we run because we are frightened, or we are frightened because we run-- or whether there is some third or fourth already existing alternative that I'm not aware of.

Bill Williams

···

-----Original Message-----
From: Control Systems Group Network (CSGnet) on behalf of Marc Abrams
Sent: Sun 1/4/2004 4:07 PM
To: CSGNET@listserv.uiuc.edu
Subject: Re: Emotions

from [Marc Abrams (2004.01.04.1706)]

[From David Goldstein (2004.01.04.1236 EST)]

If you all think this is a helpful way to proceed, maybe we
could have a good discussion on this important topic.

No, I don't think this is helpful

Marc

[Bill_Powers]

[From Bill Powers (2004.01.04.1536 MST)]

Martin Taylor 2004.01.04.1311--

In mitigation, it seems that the conscious perception of events in the
world is often (always?) delayed by many tens of milliseconds (maybe even
half a second) after the actions begin for their control. If that's so, is
it not conceivable that the cognitive experience of error is delayed long
enough that the pounding of the heart really does begin simultaneously
with the experience?

Yes, and I have proposed that this would be expected if the error in
question were at a relatively low order (configurations to categories) so
the adjustment of somatic state occurs for reasons not apparent to
consciousness. The assumption is that consciousness normally (but not
always) resides in higher-level ("cognitive") systems such as programs,
principles, and system concepts. A disturbance applied to any variable
controlled at a lower level will result in adjustment of reference levels
sent to still lower systems, and also those sent to physiological control
systems. A person may consciously experience no error (if the lower-order
system successfully counteracts its error), but may become aware of a
momentary shift in bodily state which it interprets as an emotional
reaction -- a "shot of adrenaline". And then it may notice something going
on that might explain this -- a car speeding away down the street. If the
lower-order system fails to cancel the error, or is prevented from acting
for any reason, the change in body state will persist longer, the error
will get larger, and the disturbance will have effects at higher levels,
resulting in a more full-blown emotion.

My reason for not believing that the somatic state change can occur without
some error in the hierarchy to drive it is that this idea requires the
biochemical, spinal, and autonomic systems to contain knowledge about the
world beyond their immediate sensory milieu. It's possible that there can
be some inherited connections that operate in an S-R manner (Marc has
mentioned pain receptors that may directly give rise to biochemical
changes). However, even putting it that way denies an obvious fact, that
so-called automatic or unconditioned responses to pain are most probably
really the actions of control systems. The pain arises from effects of
impending injury, and the effect of the response to pain is organized to
reduce those effects or prevent further injury. You yank the hand away from
the fire, not toward it. So what traditionally has been described as a
response to a stimulus may well be just another example of the behavioral
illusion, in which disturbances of controlled variables result in actions
to oppose their effects, but the observer, failing to notice the controlled
variable, thinks it is a simple cause-effect connection. I don't claim that
this is uniformly the correct explanation, but at least those who propose
an S-R explanation are obligated to rule out the control-system explanation
in every specific case, now that it is known to be possible.

The problem may not be with emotion and low-level stuational analysis, but
with the relation between conscious experience and the perceptions that
are unconsciously being controlled, at whatever high level.

Yes, that is exactly what I have been proposing. When you aren't attending
to lower-level controlled variables, the associated systems act
automatically to oppose disturbances. According to my theory of emotion,
any significant or protracted action is accompanied by appropriate changes
in physiological states, the effects of which can be felt. The changes are
implemented by changing reference signals sent to the organs and glands, as
already discussed. Your words seem to suggest that even higher-level
control systems may be operating automatically, out of consciousness, and
that would be another source of apparently direct emotional reactions (as
seen by consciousness). Good idea.

I'm still hoping for a phenomenological description of emotions from
others. I have found that my own emotions have a feeling component (bodily
sensations), and a congitive component. The cognitive component, which is
what I want to do but am not succeeding in doing, always seems to explain
the feeling component. As I have said, this fits negative emotions very
nicely and in all cases I have experienced. Positive emotions aren't so
simple, since they seem to involve rates of change of error: an error which
has been going on for some time actually feels wonderful it suddenly starts
to get smaller. Another echo of my childhood in the form of a joke from my
mother who seemed to understand the 10-year-old mind very well. "Why are
you hitting yourself on the head with that hammer?" Answer: because it
feels so good when I stop. Funniest thing I ever heard in my life, then.
Also, it seems to me, a true observation. Lots of pleasure is the feeling
of errors getting smaller, and the concomitant relaxation of bodily states
toward neutral.

Best,

Bill P.

[Bill_Powers]

[From Bill Powers (2004.01.04.1618 MST)]

David Goldstein (2004.01.04.1236 EST)]

(1) In defining emotion, are we going to be taking the insider point of
view, or the outsider point of view or both?

David: The PCT view emphasizes the insider view, I think.

If a person asks himself/herself, am I experiencing an emotion, and the
person says no. But outsiders say yes, whose opinion do we go with?

I think the insider view, unless you're speaking of measuring physiological
correlates of emotions. But even then, you don't have a "correlate" unless
you have an independent way of knowing when an emotion exists and what it
is called. So it always comes down to a subjective report: the state I am
experiencing is or is not what I call an emotion. If the person says "I
have that feeling I call a rush of adrenaline," you have a subjective
report of something you can measure objectively, but it wouldn't be called
an emotion.

People are sometimes not aware of their emotional state. Outsiders can
observe the nonverbal behavior of the person.

But is behavior emotion? A prizefighter attacks another person viciously.
yet would probably not say he was angry at the time. In fact, allowing your
opponent to goad you into getting angry is a good way to lose a fight. I
don't think there is any way an outsider can scientifically judge whether
an emotion exists in any other person.

(2) People undergo development. Emotions are not the same at all ages.
Are we going to be talking about the adult person?

David: I think it is probably a good idea to start with what we think
the adult state of development is. We can not ask a newborn infant
questions.

Piaget and other developmental psychologists have studied the
development of emotions in people. There is a literature.

I've read some of it, and I don';t recall the definitions of emotion as
being very informative. But I agree that being able to communicate is
important, since the experience of emotion is subjective.

(3) People in all cultures may not experience emotions the same. Are we
going to try to come up with ideas that apply to all people in all
cultures?

David: This seems like a good thing to do. We would want ideas that
apply to all people in all cultures.

Yes, I agree with that. I think that we have a better chance to come up
with cross-cultural observations if we speak of specific feelings
(sensations) and the desire to carry out specific goals (attack, flee,
hide, get help). Relying on emotion-names alone, it seems to me, would
allow for far wider variations.

(4) In defining emotion, are we going to be talking about the human
species only, or are we going to be talking about other animals?

David: I would guess that we are talking about people only. However,
people seem to think that other animals display behavior suggesting the
presence of emotions.

I agree that we are concerned with human emotion. I don't think we can
build a theory if we have to rely on hints that only "suggest" things. Too
easy for the observers biases and beliefs to come into play.

(5) In defining emotion, are we going to be talking about normal,
healthy people or are we going to be talking about people with emotional
problems or both?

David: I guess we would want a theory which could explain both.

Yes, both. My current thought is that so-called "emotional" problems are
just problems, period. They are produced by trying to control things and
not succeeding. The more important the errors are to us, the harder we will
try to correct them, and therefor the more "emotion" we will experience
while trying. Emotion is the experience of trying hard to correct an error.
The more error, the more emotion.

In other words, I think it is unjustifiable to put emotion into a separate
category of experience or behavior. It's a normal part of controlling
perceptions.

Best,

Bill P.