[From Bill Powers (991021.0710 MDT)]
Mark Lazare 991020 1600 hrs)-
HUNGER is not a CV it is an error signal -- LIKE ALL ERROR -- there is a
discrepancy between what you what and what you are getting. You either are
not getting enough of what you want (FOOD, and are weak and dizzy with a
headache) or you have had too much, (heart burn, and Bloated with a stomach
ache) and feel like crap. If I was having Thanksgiving diner and I ate
everything on my plate and went back for more, I would feel like I do every
Thanksgiving--- Full as a Tick and laid out on the couch holding my stomach
saying "I shouldn't have ate that much" "My stomach hurts"
A perception is not necessarily an error signal just because it represents
something that we don't like. If I come back to my car and see a big dent
in its side, that is NOT an error signal; it's just a perception. The error
signal arises when I compare the actual appearance of the side of the car
with the reference appearance, the appearance I wish it had or that I
remember it having the last time I saw it and would prefer that it had now.
But only the actual appearance is experienced. The comparison is automatic
and unconscious.
If I came back to my old beater car and found that someone had smoothed out
all the dents and given it a paint job in my favorite color, that would
also generate an error signal if my preference were for a car that looked
like my old beater.
The experiences we call hunger are simply perceptions, information. We do
have reference levels for these experiences; they are usually set to zero
or some fairly low value, but not always (consider recent remarks about
dieting or hunger strikes). Comparing the actual value of these signals
with their reference values generates an error signal if the perceptual
signal is greater than its reference level (a one-way control system). I
claim that we do _not_ experience the error signal, although we may
experience the activities that the error signal generates in our lower
level systems, if those activities create or change perceptual signals.
Pain is a similar perception. It is just information. An error signal is
generated when this information is compared with a reference level for that
information. In most people most of the time, the reference level for
pain-signals is zero, so any amount of pain-perception results in
generation of an error signal. The error signal is what causes the changes
in action that normally remove us from the cause of the pain-signal, or it
from us. But G. Gordon Liddy explained how a person can hold the palm of
his hand over a candle flame until the flesh bubbles: you have to not care.
That is, you have to set the reference signal for that perceptual pain
signal higher than the actual level of perceptual pain signal generated by
this action. This turns off the "pain reflex."
This is how we can handle situations that look as if we are experiencing
error signals, without saying that we experience ANYTHING other than
perceptual signals. The basic underlying assumption in the model is that we
experience ONLY perceptual signals; that is, signals arising from
perceptual input functions that are receiving the outputs of multiple lower
perceptual input functions (or sensory nerves, the PIFs of first-order
systems).
Note that we don't experience reference signals, either, according to the
main hypothesis. That hypothesis is the entire reason for postulating the
imagination connection. If we could experience reference signals directly,
why would that connection be needed? If you understand that I am accepting
the proposition that we experience ONLY perceptual signals, you will
understand why the model is the way it is.
If we don't experience reference signals, how is it that we can know (that
is, experience) what we want or intend? If we perceive nothing but
perceptual signals, then somehow the reference signals have to end up in
the same channels where perceptual signals normally occur. And that, of
course, is a beautiful solution even if it has some loose ends. To send a
reference signal back into the perceptual channels is exactly the same as
what would happen if the lower system received that reference signal and
matched its perceptual signal to it, with a copy of the perceptual signal
going upward into the perceptual apparatus of the higher system -- which is
how the basic HPCT model has always worked. Imagining the reference signal
is just like controlling it perfectly and instantly (except that nothing
actually happens at lower levels or in the outside world). This is why we
can control things in imagination that we can't control in the real world.
Should we continue to insist that we experience ONLY perceptual signals?
Only as long as there is an easy and sensible -- and preferably elegant --
way to explain the cases where it _seems_ that we are experiencing error
signals. And if you think about it, the basic hypothesis, and the revisions
of common sense needed to make it seem acceptable, also save us from having
to explain many other anomalies. For example, suppose you perceive the
amount of food you see on the table as an error: too little food. If
someone kept putting food on the table, eventually you would be
experiencing another error: too much food. But wait a minute. If a certain
amount of food is perceived as a too-little error, then how could
increasing the amount of food also be perceived as an error? We get away
from such problems by saying that the perceived amount of food is just a
perception, whether it represents a tiny amount or a huge amount. It is
judged as an error only when compared (invisibly) with a reference signal
by a specific control process.
My stomach can hurt from the lack of food or the excess of food. Either is
an Error, either in judgment or fortune. In both cases you "feel" and
"sense" the ERROR. If this happens often enough, you might get an insight
and Reorganize.
This is what I'm trying NOT to say. You perceive fullness of your stomach.
But do you also perceive the reference fullness, so you can perceive that
one is greater than the other? I say not. You know that the amount of
fullness is too much only when you feel yourself, for some unknown reason,
starting to reject the input of any more food. You're beginning to
experience sensations that seem bad to you, but they seem bad because you
find yourself trying to avoid them or make them go away. If I asked you
_why_ that's a bad sensation, you couldn't tell me. It's just bad. But
while you were eating, it was good. You can't tell me what made it good,
either.
The control-system model is intended to explain what we _do_ experience by
filling in the details of what we _don't_ experience about our own
operation. If you look at the world around you, you won't see any reference
signals or error signals. You'll just see what your senses are telling you
about what is (apparently) out there. In that world, you can observe
intensities, sensations, configurations, transitions, events,
relationships, categories, sequences, logical processes, principles, and
system concepts, all existing in the objective world that looks as if it's
outside you. It's really made of neural signals in your head, according to
the theory. But even if the theory's right, you can't see any reference
signals or error signals "out there."
In this modeling game, the objective is to get by with as few ad-hoc
assumptions as possible. Any time you think you can account for something
more simply, and just as completely, the old explanation has to go. And any
time you propose something new as part of the model, you have to live with
it not just in the situation that prompted you to propose it, but from then
on forever, in every situation.
So far I haven't seen anything to make us drop the proposal that our
experiences are associated ONLY with perceptual signals. Instead, other
interpretations that make error signals and reference signals part of
experience lead to absurdities and suggest phenomena that we just don't
observe. Or at least that _I_ don't observe. Anyway, whether you agree with
my assumption or not, you can understand my arguments about things like
hunger, thirst, pain, and such if you realize that I do not think we
experience anything but perceptual signals as defined in the HPCT model.
Best,
Bill P.