Dormitive principles

[From Bill Powers (931003.0030 MDT)]

I am conducting a sort of low-level Gregory Bateson Memorial War
Against Dormitive Principles. This was Bateson's way of referring
to explaining a phenomenon by using its description, somewhat
transformed, as its explanation. Thus, Seconal makes you sleepy
because it contains a dormitive principle. By switching from the
English "sleep" to the French/Latin "dormir" you make it sound as
though you are naming a cause, whereas in fact you are simply
repeating the description in a sentence that has the form of an

This method of explanation is popular in medicine: you have red
itchy eyes because of conjunctivitis, and a red itchy nose
because of rhinitis, and are cross-eyed because of strabismus.
You break out in red spots and have a fever because you have
measles. In fact this is a popular mode of explanation in any
field where people keep pestering you for explanations and you
find it embarrassing or impolitic to keep saying "I don't know."

We have seen two examples of the dormitive principle recently on
CSGnet. One is in the discussion of phonemic contrasts, and the
other -- somewhat more subtle because more widespread -- in the
discussion of emotions.

Phonemic contrasts have been exemplified recently by two pairs of
vocalized words: pin and bin, and spin and sbin. In the first
pair, subjects say that there is a difference in the words that
are heard, while in the second pair this difference is not heard,
even when pains are taken to make sure the actual objective
sound-waves of pin and bin are preserved.

The explanation offered for this experienced phonemic difference
is that there is a phonemic contrast between bin and pin, while
there is none between spin and sbin. This is a dormitive
explanation, because "difference" and "contrast" mean essentially
the same thing. The explanation could just as easily be given
this way: there is a perceived contrast between pin and bin but
not between spin and sbin because in the former case there is a
phonemic difference, while in the second there is not. The second
way of putting it sounds, on the surface, just as explanatory as
the first.

The truth of the matter is that people perceive a contrast or
difference between pin and bin, but not between spin and sbin,
and nobody knows why, not even a linguist.



The second example, emotion, is a little harder to untangle
because some people set great store by emotions and don't like to
think that emotions might have a rather simple explanation.
Emotions, traditionally, are treated as a separate branch of
motivation, reaction, or experience, having a somewhat mysterious
kind of existence that is neither physical nor mental. Scientists
decry arguments that appeal to emotion rather than reason. Their
opponents often sneer at emotionless scientists for their
coldness or indifference to feelings. Both, when asked to explain
what they mean, fall back on dormitive principles.

Consider the emotion called anger. How do you know when you're
feeling anger? On _Star Trek: The Next Generation_, the android
Commander Data recently asked this question of Geordi, the blind
Chief Engineering Officer. To make this point, Data asked Geordi
to describe anger without using the word "angry." Geordi (and
presumably, the show's writers) were at a loss. You just -- you
know -- feel _angry_. If you don't know what anger is, how can
you understand a description of it? Geordi, admirably, refused to
fall back on a dormitive principle, and admitted that he couldn't
describe anger. Eventually, Commander Data understood that he had
in fact experienced anger for the first time (due to remote-
control meddling by his evil brother-android Lor) and therefore
_did_ know what Geordi meant, even though he couldn't define it
rationally. This, too, puzzled Data, by violating his concept (as
well as mine) of explanation as something different from
description .

Well, what does happen when you feel angry? You feel a surge of
sensations from your body, and an urge to do something energetic
to something. If you have no self-control you may well lash out
and do damage to something or somebody -- anger most often has an
object at which you're angry, and it's usually a person.

At one time in psychology there was one of those Scholastic
debates about emotion. In the terms just laid out, one side
argued that you feel a surge of sensations and an urge to strike
out because you're angry, while the other side argued that you're
angry because you feel the sensations and an urge to strike out.
Both sides were using the same dormitive principle with the terms

The term anger refers to an experience of a surge of bodily
feeling and an urge to so something extreme. Anger is just the
short way of saying " bodily feeling and an urge to do
something." "Anger" isn't an explanation: it's a word referring
to a phenomenon that needs an explanation. You don't feel the
sensations and the urge to act BECAUSE OF anger, or vice versa.
You feel the sensations and the urge to act, or alternatively,
you feel anger. The two ways of putting it say the same thing.
The word "anger" and the phrase "a surge of bodily feeling and an
urge to so something extreme" refer to the same experience.

How would we explain this experience in terms of the PCT model?
Clearly, "a surge of bodily feeling" is a perception, and an
"urge to do something extreme" implies a control system
containing a large error signal. Why, we may ask, would the
occurrance of a large error signal in a neural control system be
accompanied by a surge of bodily feeling? One answer that seems
reasonable is that the same output of the control system in
question that would set reference levels calling for extreme
action by the lower motor systems would also set reference levels
calling for an altered state of the biochemical systems that
support action. Thus we would expect blood sugar to rise,
respiration to increase, heart-rate to increase, and so forth --
the so-called "general adaptation syndrome." These sudden changes
in somatic state can obviously be sensed; they are experienced as
bodily feelings.

So when a reference signal is suddenly changed to a relatively
extreme value, or a large disturbance suddenly appears, the
result is an error-signal-driven urge to change the state of the
motor systems and the state of the biochemical systems by a large
amount. There is thus a surge of sensation from the body as the
biochemical systems are called upon to change to a significantly
different state.

Under normal circumstances and in a well-balanced system, the
heightened state of preparation of the body is immediately "used
up" by the accompanying motor action. There is a momentary sense
of elevated somatic state that is simply part of the sensed
action. The word "anger" would not be likely to be used to refer
to the result.

If, however, the person who experiences the large error has good
self-control, a conflict immediately ensues. One control system
receives a reference signal implying an immediate change of state
of the whole system, and at the same time a second control system
says "No, a civilized person like me does not punch a boor in the
nose, whatever the provocation." The "civilized" system cancels
the reference signals going to the motor systems, and the punch
does not take place.

However, the control system gearing up for the punch is still
there, and it is still telling the somatic systems to prepare for
violent action. This state of preparedness is now not dissipated
by the appropriate motor behavior and disappearance of the error
signal; it is maintained by the same error signal that would
throw the punch if lower systems were not receiving cancelling
reference signals from the "civilized" system. The reference
signal calling for extreme action is not matched by the
appropriate perception, so the urge to act continues and the
sensation from the body persists, too. NOW the person would say
"I am angry!"

Moreover, the person would say "I am angry AT HIM." The person
still wants to see and feel a fist mashing the other's nose, the
other person crying out in pain, falling, becoming abject and
apologetic and tearful and otherwise suffering all the
embellishments of a throughly satisfying retribution. All these
desires are the immediate source of the reference signal that
suddenly changed so as to call for an energetic punch. As long as
these desires are in effect, the "civilized" system will have to
keep cancelling the actual motor reference signals, and the anger
and hatred and whatever else we call it will continue. The
emotion will persist until the source of the reference signal is
turned off. One ceases to be angry when one ceases to want

This is a PCT explanation of anger that does not rely on a
dormitive principle. The same can be done for all the other
experiences we label with emotion-names. The feeling component is
the perception of a change in the biochemical state of the body,
or more generally, somatic state. The goal-component is the
reference signal that is calling for both motor action and the
somatic state appropriate to the action. If the goal is to get
the hell out of there, the same somatic changes take place as in
anger, but now the combination of goal and feeling is called
alarm, fear, fright, terror, panic, and so on. When the action is
prevented from succeeding in achieving the goal, the emotion is
felt the most strongly.

True conoisseurs of emotion have as large a vocabulary for
describing emotions as epicures have for describing tastes and
smells. We can speak of feeling annoyed, offended, irritated,
provoked, exasperated, angered, incensed, aroused, inflamed,
infuriated, and enraged. I've just arranged the terms under
"anger" from good old Roget more or less in order of increasing
error signal and increasing shift in somatic state, as I
understand them.

Notice how those adjectives imply the passive voice. It isn't
common to attribute emotions to one's own desires. Emotions --
particularly the somatic feeling part -- seem to arise as though
they're being done to us by something else, as if they're being
received from outside us. They _are_ being received from outside
our understanding; that's why we need models. But in this case
the model tells us we gambled on the wrong voice: we produce our
own emotions, which arise from what we want. All these terms
should be used in the active voice, which sounds really strange
when you do it. I'm exasperating at you?

Psychological explanation is riddled with dormitive principles.
PCT can eliminate them by offering, right or wrong, real
Best to all,

Bill P.