Tai Chi

[From Bill Powers (940503.0740 MDT)]

Avery Andrews (940503.1508)--

RE: Tai Chi

You make very slow and deliberate movements, with the long term
goal of getting good coordination of upper and lower body
movements. I think that this might be a matter of developing
good level 2 perceptions (sensations). The low speeds of
movement and often rather awkward-feeling positions provide
what ought to be ideal conditions for reorganization to produce
new postural sensations and control-systems for them.

I can see some suggestions of level 3 (configuration) and level 4
(transition) in your descriptions. I think of level 2 as consisting
of sensations like effort, the kind of sensation you get when you
pull or push on a heavy object, or when you "make a muscle" by
tensing opposing muscles. A configuration, on the other hand, would
be a set of limb and body poses controlled by varying muscle
efforts, with a fixed reference-configuration corresponding to a
particular static pose or arrangement of body parts. A fixed
reference level for transition, then, would be a fixed rate of
change of configuration, with the magnitude of the reference signal
corresponding to the speed of change.

This seems to be supported by your description:

Another principle is that you never go ballistic-any movement
should be able to be halted or reversed at any time (`walking
like a cat', as opposed to 'walking like a dog'). So control
of motions seems to be firmly based on a prior control of

Yes, positions as I think of level 3: fixed configurations. The
principles relating the hierarchical levels require that

1. For the higher perception to be controlled the lower must be
variable, and

2. Lower perceptions, if changed in the right combinations, can be
changed without changing the higher perception (this implies, of
course, that there are unused degrees of freedom at the lower

If you hold an arm out in a particular configuration, such as with
the elbow bent, you can vary the effort sensations without changing
the configuration, by running the muscles in parallel up and down
their tension curves (raising and lowering the opposing tensions
simultaneously). But to alter the configuration, you must alter the
relative contraction of opposing muscles: that is, the muscle
sensations are a little different in the new pose from the
sensations in the old one.

If you change the reference signal for configuration, the actual
configuration will track the change. As you say, stopping the change
in reference signal stops the change in configuration; reversing the
change in reference signal reverses the change in configuration.
With configuration control, the configuration is set by the current
value of the reference signal, and remains constant as long as the
reference signal is constant.

For the slow movements of Tai Chi, I would guess that the rate of
change of configuration (transition) is being sensed and controlled.
A constant transition reference signal requires a constant rate of
change of configuration to exist, to produce a constant perception
of velocity. The velocity is set by the magnitude of the transition
reference signal. For slow movements, the reference signal is set to
a constant low value; for faster movements, to a higher constant
value. The output function of a transition system is probably a set
of integrators, so that a small error signal causes a small rate of
change of output, and so on. That changing output becomes the
changing configuration reference signal. Reducing the transition
reference signal to zero leaves the reference signal for
configurations constant (the output integrators hold their output
values when the error drops to zero).

It seems possible that the reason for the slow movements is to
extend the dynamic range of transition control as near to zero as

This has nothing to do with the rate feedback at the spinal level,
which is present mainly for damping and stabilization, and is not
separately sensed.

Since configurations involve joint angles, they are relative to the
body rather than a fixed external coordinate system. This means that
configurations relative to the environment can change without
changing transitions: the whole body can move while maintaining many
configurations unchanging. What can change while what else remains
constant depends on the available degrees of freedom.

From the little I have seen of Tai Chi, there are also at least

"event" perceptions under control. That is, there seem to be
stereotyped movements which involve smoothly varying velocities and
positions through a set pattern of changes.

And finally, the event reference signals might be the outputs of
relationship control systems. While the positions, transitions, and
events are controlled for each limb, the limbs appear to be
controlled in a way that establishes particular controlled
relationships among them. One arm slowly extends while the other
slowly retracts, for example, maintaining a particular relationship,
such as symmetry, between them.

You have a nice portable laboratory and instructors who are very
experienced with its use. How do your experiences with Tai Chi jibe
with my guesses about the levels of control?



Bill P.

[Lash LaRue, 940503.1400 EDT]

(Bill Powers, 940503.0740 MDT & Avery Andrews 940503.1508)

Bill & Avery's comments about Tai Chi give me a way of posing a
question that I am puzzled about.

Bill & Avery were focusing their attention on the question: which
level in the hierarchy of levels is being controlled?, which leads to
the question of how the reorganization of control is generated. I
like these questions, but I also want to ask another one.

As I understand it, Tai Chi trains one in new ways of using the body,
and I think that its practicioners contrast the new ways with the old
"habits." The contrast with "habit" is what I would like the group
to clarify for me.

Of course, that word, "habit," has no precise analytical meaning, but
when I have listened to experts in physical training use the word,
they seem to have a concept in mind that is narrower that ordinary
usage and that is very close to what some people call a "reflex."
The teacher seem to be saying (but maybe I misinterpret) that our
habits of body use contain as a part (not the whole) some learned
reflexes that can be unlearned. What does this mean?

Consider a clear case (?) of a reflex, the "startle reflex." A loud
noise that startles sets off a complex set of bodily motions that are
"automatic." In this complex, the upright posture of the body
collapses, muscles shorten and tighten, the head is pulled down into
the shoulders, the arms are pinned in toward the body, etc.

First question: is this a case where the use of stimulus-response
analysis is proper? Or is this also a case for a PCT analysis.

Without regard to that question, let me move back to Tai Chi and
other body work in which teachers try to retrain us to use our bodies
in different ways. As I understand it, the teachers contend that our
ordinary (and bad) use of our body is all too similiar to the startle
reflex. In other words, the collapse, the shortening, the
tightening, the pulling of the head down into the shoulders, is
likely to accompany routine action. (To a lesser degree, of course.)
And the point of the retraining is to train one into non-reflexive

Indeed, I had one teacher who explained the matter as follows: our
habitual actions are all too likely to be reflexes triggered by a
stimulus; the goal of the retraining is to bring our actions under
conscious control; the means of the retraining is to inhibit the old
habitual response and to explore new ways of moving that are now
under our control.

Second question: is this a case of "moving up a level"?

I do not know if I have asked these questions clearly enough.
Perhaps some critique of the way I have asked the questions might be
as useful to me (and perhaps others) as would be any answer.

Best wishes,

Lewis Henry LaRue
Washington and Lee University
School of Law
Lexington, VA 24450

e-mail address: LHL@FS.LAW.WLU.EDU
telephone: 703-463-8513

[Avery.Andrews 940506.0928]
(Lash LaRue, 940503.1400 EDT)
(Bill Powers (940503.0740 MDT))

My Tai Chi instructors don't say much about habits, old or new, but
it seems to me that their view is that the idea is to replace old systems
with new, more effective and efficient ones (developing the new and letting
the old recede into the background on its own). Part of this is developing
'body awareness', but I doubt that putting actual movement under conscious
control to a greater extent is really what they have in mind. Indeed,
for the martial arts applications they are quite explicit that things
happen much too fast for any kind of conscious decision-making. Presumably
these applications involve control of relationships between yourself
and the opponent, but this stuff doesn't get taught until after several years
of study (fitting nicely with the logic of the PCT hierarchy).

So my interpretation of their view is that the idea is to replace old
'habits' by new ones, and I conjecture that the method works by triggering
reorganization, the trigger being the feelings of awkwardness, poor
balance, etc. occasioned by the postures (setting things up so that the
only way to eliminate the errors is to fix the non-optimal control systems
(c.f. Bill's discussion of reducing error caused by the inadequacies of
one control system by messing around with others).


Text item: Text_1

     I discussed the recent traffic on the net about habit, Tai Chi, and
     the martial arts with one of the senior sensei at the Shotokan dojo
     where I study. He and I agree that habit is not the objective of Tai
     Chi or the martial arts.

     One practices basics and katas in order to develop the specific
     strength and muscle memory necessary to perform a particular movement,
     because centuries of study have shown that doing that movement that
     way is less likely to result in self-damage or imbalance than is doing
     some other way. The objective is NOT, however, to develop a habitual
     reaction to any given situation. That is precisely what karate, kung
     fu, akido, et al try to keep you from doing. The desired skill is to
     move, without thinking, to deal with the environment - whether a
     threat is present at that moment or not. The objective is not habit,
     but rather awareness. Repetition of basic movements and stylized
     practices (katas) are intended to make one aware of ones own movement,
     not to ingrain a habitual way of moving.

     The reason one feels so awkward when first learning, for example, the
     Tai Chi step, is that the motion, although absolutly natural, is
     unfamiliar. That is because most people, other than martial arts
     nuts, professional dancers, and physical therapists, spend much time
     THINKING about human movement. Is walking, jete', or washi giri
     habitual movement? I don't think so. I think they are the absence of
     habit - the constant reassessment of position and desired location.
     Aren't these elements of PCT?

[LaRue, 940507.1145 EDT]

     I discussed the recent traffic on the net about habit, Tai Chi, and
     the martial arts with one of the senior sensei at the Shotokan dojo
     where I study. He and I agree that habit is not the objective of Tai
     Chi or the martial arts.

Did I misword my message? I thought that I said that "habits" of
movement were the problem, and that the various body awareness
techniques were a solution (dissolution) (a disolving in a solvent)
of the problem of habitual misuse of the body.

My question was: what does "habit" mean in this context? In orthodox
psychologly, habit is described with stimulus-response explanations;
so I was asking for an alternative account.

     [the techniques] are the absence of
     habit - the constant reassessment of position and desired location.
     Aren't these elements of PCT?

Nicely put.

Lewis Henry LaRue
Washington and Lee University
School of Law
Lexington, VA 24450

e-mail address: LHL@FS.LAW.WLU.EDU
telephone: 703-463-8513