[From Bill Powers (941209.0715 MST)]
Jeff Vancouver (941208) --
That is, the experience of a chill only makes sense if we understand
that the perceptual signal of the temperature system is compared again
the reference signal (which the immune system set to a higher than
normal level). Thus, the subjective experience is of the ERROR signal,
not the perceptual signal. I thought we did not experience
(consciously) error signals.
So did I, and I'm still not sure how to deal with situations like this.
For a lot of general reasons I have assumed that we consciously
experience only perceptual signals -- that is signals that are part of
the input side of control processes. But this really is an assumption,
not a fact. I don't know of any way to test it directly but to induce
neural signals in various pathways (short-circuiting the signals to
prevent their getting into different pathways) and see whether the owner
Under this assumption, we are forced to make more assumptions to explain
certain subjective phenomena. This is how the "imagination connection"
was born. If we can't experience anything but upgoing signals, how can
we ever know of an intention without carrying it out? The only way left
is to route a copy of a downgoing reference signal back into the upgoing
pathways, the only ones assumed to be associated with consciousness. As
it happens this has some nice side-effects, such as making sure that the
signals are put through the same interpretive functions that are used
when the reference signal is used in the normal way to tell lower
systems what perceptions to produce. So the added assumption we are
forced to make reveals a problem we had not noticed, and solves it at
the same time. That gives some added credibility to the first
assumption. Another unnoticed problem that is solved is how we can
imagine doing something without actually doing it -- how we can plan
actions without the plan immediately turning into actions.
Also, the assumption that we experience only upgoing signals leads us to
understand certain things in a new way that turns out to make sense.
Consider the sense of "effort." When we exert a lot of effort, we think
of the effort as an output: it's as though we are sensing what we are
doing to the world outside. But we have assumed that we can't experience
outgoing effects or the signals that cause them (except in imagination).
So the sense of effort must come from sensory receptors, and sure
enough, there are multitudes of receptors of the right kind. Sensory
signals arise from tendon receptors and stretch receptors in muscles and
skin pressure receptors and joint pressure and angle receptors, all of
which monitor the _consequences_ of outgoing neural signals and muscle
contractions. So when we pick up a "heavy" object, attributing the
heaviness to the object, we are really sensing the kinesthetic
consequences of attempting to pick it up. Now we can explain certain
illusions and aftereffects, and also why we experience unexpected
heaviness when we try to pick up something that we know to be light, but
which happens to be nailed to the floor. We can also explain why loss of
kinesthetic sensations leads to feeling paralyzed even though the
outgoing pathways are functional.
Sidenote: a trick learned in high school. A strong boy can lift a girl
by grasping her waist and lifting (which makes the trick fun). But if
the girl takes hold of the boy's forearms and pulls UP, she suddenly
feels a great deal heavier, and often it seems impossible to lift her.
Feeling a chill isn't understood quite so easily. It does seem that we
are experiencing an error signal. However, all may not be lost.
What if we are experiencing not the error signal itself, but something
that depends on the error signal? That should produce the same
subjective effect. So what depends on the error signal?
How about sweating, shivering, increased moscle tension,
vasoconstriction, and vasodilation (etc), the actions that are produced
by the error signal to correct it? These actions have sensory effects,
and under the primary assumption we can experience sensory effects. This
is like explaining the sense of heaviness or effort as perceptions of
the sensory consequences of the output signals.
We can also experience temperature, of course, because we have
temperature-sensitive receptors. However, this doesn't explain why
feeling the same temperature as before is experienced as feeling "too
cold." We do need some sensory effects that depend on the error signal.
I think it is plausible that the required effects arise from what the
temperature control systems are trying to do about the error signal --
from the actions being taken to raise the body temperature.
There's always a conflict about how far to carry plausible explanations
like this. If there'a better and simpler answer, then the sooner it is
found, the less embarrassing it will be. But once you've proposed a
general explanation, it has to be tried out thoroughly enough to see if
it can really handle the difficult cases -- and as long as there's no
evidence to the contrary, it might be right. What we really hope is that
eventually the basic explanation will lead us to expect something that
can be tested directly, to see if there is a crack in the bland surface
of pure plausibility -- or if we are left with no alternative but to
accept the explanation.
A certain amount of this sort of rationalization is, I think, essential
in trying out any new theory. You don't want to abandon it at the first
sign of trouble. But when the rationalizations begin to pile up, one
assumption leading to the need for more assumptions and those leading to
still more, there comes a point when insecurity starts to set in, and
there is a growing yearning for some sort of experimental verification.
Houses of cards have a tendency to sudden collapse. This is why I get to
a certain point in defending the assumptions of PCT, and become
unwilling to go any further without some experimental backing. I think
there are phenomena that are so difficult to explain -- that require so
many unverified assumptions -- that it's better not to try to explain
them, and just say "I don't know." I don't think that puts PCT at a
disadvantage; other theories have built their card-houses to much more
dizzying heights without experimental verification. But some theorists
don't seem to have the sense to know when they should feel insecure.