Re. Bimanual Control

[From Bjorn Simonsen (2004.07.09.13:05
EuST)]

[From Rick Marken
(2004.07.07.0845)]

I
am putting it up on the web in order to solicit suggestions,

requests
for clarifications and so on.

Bimanual control is interesting. May I talk
a little around?

It is customary to say that our left
cerebral hemisphere send output signals to our right extremities. I don’t know
what PCT/HPCT say about that. Have you talked about that earlier?

As I understand (and here I might be wrong)
the first level outputs come from the cerebral cortex. There are two cerebral
cortexes, one in each cerebral hemisphere. Signals from first level can pass to
the second level that we also find different places in the cerebral hemispheres.

I remember from B:CP Bill’s account
about Hess’s experiments. These experiments tell us that third level is in the
thalamus region. There is just one thalamus, therefore perceptual signals from
both cerebral hemispheres go to thalamus or further. All higher levels are
represented in one region, not as the motor cortex (two hemispheres).

Your model is a control system at the
sixth level (the logical one)??. The
output goes down to the third level and there are probably many references that
distributes to the two hemispheres and motor outputs goes to the right arm from
the left hemisphere and opposite. (You have a more simple model).

If they really control da/dt, I guess
the reference value to the first level systems would be the same value. But
most people have their best arm and I guess there are different gains and
slowing factors in the two systems. Because
of the feedback the two perceptual signals in the two systems become different.
Then we get a conflict??? Can this explain why people producing antiphase rotations would often end up
controlling for symmetry as they increased the speed of flag rotation?

I have tried some experiments myself. I
have placed two ordinary dinner plates on the table and moved my clenched fists
round and over the plates circumferences. It works very well when I do it
symmetric. If I do it 180 degrees asymmetric, I gradually continue with
symmetric movements.

When I move my arms symmetric I use the
same muscles at the same time, when I move my arms asymmetric I more or less
use “opposite” muscles at the same time. May also this be a reason why people producing antiphase rotations would often end up
controlling for symmetry as they increased the speed of flag rotation?

Over
to your Mechsner model.

On your web you have written, “While there was a 1:1 relationship between left hand
rotation and left flag rotation, there was only a .75:1 relationship between
right hand rotation and right flag rotation.”

I
prefer ““While there was a
1:1 relationship between left hand rotation and left flag rotation, there was
only a 1:1.33 relationship between right hand rotation and right flag
rotation.” Remember it is just what I prefer.

Your
figure 2 is indistinct on my Explorer. I had to copy it to a Paint program
where I can zoom it.

In
your figure 2 I see your model has two levels. That is nice, but you should
present a short account about PCT/HPCT (the first levels). You asked for our/my
views. You have a nice presentation |”The Hierarchical Behavior of Perception”.
Why not have a reference to that essay,

You
have a figure 3 that explain the diagram we get after running the simulation.
But I would have used some words describing the table. I would not used decimal
numerals telling each turn of the flag (z) and I would have more numbers on the y-axis. I would also
explained that your statistic is taken each 0.1 second (??) and telling that
the z-axis describes how many percent of the counting you get a certain angular-difference
value. I say this because I used much time to understand the z-axis. But that
is me.

This
was interesting. Maybe we can talk more about bimanual coordination after
Chicago. It tells much about the brain

Have
a nice conference in Chicago. If I had been there I would also visited The North
Western University. I know they do it well with Nano technology, with organic
molecules as transistors, molecules that can multiply themselves. They should
know about PCT.

bjorn

[From Rick Marken (2004.07.09.0930)]

Bjorn Simonsen (2004.07.09.13:05 EuST)

Bimanual control is interesting. May I talk a little around?
It is customary to say that our left cerebral hemisphere send output signals
to our right extremities. I don�t know what PCT/HPCT say about that. Have you
talked about that earlier?

No. I don't really think it's relevant. The nervous system is wired up so
that the left side of the brain is connected mainly to sensors and effectors
on the right side of the body, and vice versa for the right side of the
brain. But I don't think this has much to do with the bimanual control task
developed by Mechsner et al.

As I understand (and here I might be wrong) the first level outputs come from
the cerebral cortex.

First level efferent neurons probably come from the spinal cord and medulla.

I remember from B:CP Bill�s account about Hess�s experiments. These
experiments tell us that third level is in the thalamus region. There is just
one thalamus, therefore perceptual signals from both cerebral hemispheres go
to thalamus or further. All higher levels are represented in one region, not
as the motor cortex (two hemispheres).

I don't think that was the main conclusion based on the Hess experiments
(those involved deafferentiation, right?). The main conclusion I take from
the deafferentiation studies is that there is massive loss of control with
destruction of the perceptual paths, even when the motor paths are left
intact. But organisms can learn to compensate for this loss of control by
controlling higher level perceptions that still exist, though control is
quite sloppy without lower level control possible. I think Hess concluded
that deafferention shows that perception is not important to performance
because the deafferented organisms can still do things they could do before
deafferentiation. But PCT would point out that they do these things _very_
poorly and they can still do them only to the extent that many higher level
afferent paths are left intact after deafferantiation.

Your model is a control system at the sixth level (the logical one)??.

I never think about what the specific levels are when I build models like
this. I don't think it's necessary. Once we have many hierarchical models of
many different behaviors we should be able to see whether they all imply a
consistent hierarchical structure, or not.

If they really control da/dt, I guess the reference value to the first level
systems would be the same value. But most people have their best arm and I
guess there are different gains and slowing factors in the two systems.
Because of the feedback the two perceptual signals in the two systems become
different. Then we get a conflict??? Can this explain why people producing
antiphase rotations would often end up controlling for symmetry as they
increased the speed of flag rotation?

The model is set up so that rotation of the right and left hands affect the
two higher level controlled perceptions -- symmetry and speed --
differently, so there is no chance of conflict. I think the occasional
"defection to symmetry" in the antiphase case results from something other
than conflict. There really is no more conflict in the antiphase than in the
symmetry case. There is just something easier about controlling for
symmetry. I think the defection occurs because antiphase is simply harder to
_perceive_, when the flags are rotating rapidly, than is symmetry. I think a
higher level system just changes to controlling for symmetry rather than
antipahse when, for all intents and purposes, the antiphase perception
"disappears" as rotation speed increases. I think there are better
experimental ways to determine what's going on here. I've suggested some to
Franz and he was very receptive. Indeed, he said he had thought of some of
the same things himself and had already preparing some research along the
lines I suggested. Franz definitely sees behavior through PCT glasses, even
if he has not specifically ordered that prescription.

When I move my arms symmetric I use the same muscles at the same time, when I
move my arms asymmetric I more or less use �opposite� muscles at the same
time. May also this be a reason why people producing antiphase rotations would
often end up controlling for symmetry as they increased the speed of flag
rotation?

Actually, Franz's experiments are aimed at showing (successfully, I believe)
that this explanation of the difficulty of doing antiphase movements (which
is the prevailing explanation, by the way) is wrong. Note that in the demo
at the web site, the hands are often moving in antiphase rotation in order
to maintain the symmetric movement of the flags. I think this shows that the
difficulty of antiphase control lies in our perception, not in our ability
to produce antiphase movement.

Over to your Mechsner model.

On your web you have written, �While there was a 1:1 relationship between left
hand rotation and left flag rotation, there was only a .75:1 relationship
between right hand rotation and right flag rotation.�
I prefer ��While there was a 1:1 relationship between left hand rotation and
left flag rotation, there was only a 1:1.33 relationship between right hand
rotation and right flag rotation.� Remember it is just what I prefer.

Suggestion noted. I think it's a good one. I will consider changing this (it
may be the way Mechsner described it also).

Your figure 2 is indistinct on my Explorer. I had to copy it to a Paint
program where I can zoom it.

I'll try to re-do it as a GIF. Thanks.

In your figure 2 I see your model has two levels. That is nice, but you should
present a short account about PCT/HPCT (the first levels). You asked for
our/my views. You have a nice presentation |�The Hierarchical Behavior of
Perception�. Why not have a reference to that essay,

I agree!. Great suggestion, Thanks!!

You have a figure 3 that explain the diagram we get after running the
simulation. But I would have used some words describing the table. I would not
used decimal numerals telling each turn of the flag (z) and I would have more
numbers on the y-axis. I would also explained that your statistic is taken
each 0.1 second (??) and telling that the z-axis describes how many percent of
the counting you get a certain angular-difference value. I say this because I
used much time to understand the z-axis. But that is me.

I think you are absolutely right. The data displays definitely need more
explaining. I will work on this over the weekend!

This was interesting. Maybe we can talk more about bimanual coordination after
Chicago. It tells much about the brain

That would be great with me. I agree that it tells much about the brain (and
so would Mechsner, I'm sure, who, by the way, is a _biologist_, which is
cool, I think).

Have a nice conference in Chicago. If I had been there I would also visited
The North Western University.

Northwestern was my father-in-law's alma mater. I wish you could come to the
meeting! Maybe next year. Or perhaps the European meeting in 2005.

They should know about PCT.

I certainly agree. But more and more people are starting to find out about
it and use it (sometimes even using it appropriately!!) in their work. And
if some of it's enemies (most of whom are on CSGNet;-)) would get out there
and start sounding the alarm about how bad PCT is, PCT will really start
taking off.

Thanks for all your suggestions Bjorn. This was a very helpful post!

Skol (as an old alcoholic who spent much of that joyous time at a Norweigean
Lutheran college in Minnesota I particularly miss Norway's most delicious
and potent contribution to world peace, Aquavit)

Rick

···

--
Richard S. Marken
MindReadings.com
Home: 310 474 0313
Cell: 310 729 1400

[From Bill Powers (2004.07.09.1230 MDT)]

Rick Marken (2004.07.09.0930)--

> I remember from B:CP Bill�s account about Hess�s experiments. These
> experiments tell us that third level is in the thalamus region.

I don't think that was the main conclusion based on the Hess experiments
(those involved deafferentiation, right?).

No. See B:CP, p. 119 ff.

Best,

Bill P.

[From Rick Marken (2004.07.09.1330)]

Bill Powers (2004.07.09.1230 MDT)--

Rick Marken (2004.07.09.0930)--

I don't think that was the main conclusion based on the Hess experiments
(those involved deafferentiation, right?).

No. See B:CP, p. 119 ff.

Oops. That's what I get for not having a copy of B:CP at work. I'll buy a
couple copies of the paperback so I can keep one here and a couple copies of
the Chinese version as well (I work with a number of Chinese here who might
find it interesting).

I agree with what I said about the deafferentiation work. So what did Hess
do?

Best

Rick

···

--
Richard S. Marken
MindReadings.com
Home: 310 474 0313
Cell: 310 729 1400

[From Bill Powers (2004.07.09.1858 MDT)]

Rick Marken (2004.07.09.1330)--

>> I don't think that was the main conclusion based on the Hess experiments

>> (those involved deafferentiation, right?).
>
> No. See B:CP, p. 119 ff.

Oops. That's what I get for not having a copy of B:CP at work. I'll buy a
couple copies of the paperback so I can keep one here and a couple copies of
the Chinese version as well (I work with a number of Chinese here who might
find it interesting).

  So what did Hess.

No time. Read it when you get home. It's only about three pages at most.

Best,

Bill P.

[From Rick Marken (2004.07.10.1430)]

Rick Marken (2004.07.09.0930)--

Bjorn Simonsen (2004.07.09.13:05 EuST)

I remember from B:CP Bill�s account about Hess�s experiments. These
experiments tell us that third level is in the thalamus region. There
is just
one thalamus, therefore perceptual signals from both cerebral
hemispheres go
to thalamus or further. All higher levels are represented in one
region, not
as the motor cortex (two hemispheres).

I don't think that was the main conclusion based on the Hess
experiments
(those involved deafferentiation, right?).

I'm sorry Bjorn. My comments about deafferentiation had nothing to do
with the Hess experiments, which involved injecting electrical pulse
trains into neurons in the thalamus that apparently functioned as
carriers of reference signals for position. Injecting pulses at a
particular rate led to the animal assuming a particular body position.
Changing the pulse rate led to changes in position. Hess also
determined that the positions assumed by the animal were controlled,
ie. protected from disturbance. This work, which provides remarkable
support for the neurophysiological model implied by PCT, was done way
back in 1957, before, I suppose, all the new modern neurophysiological
work that shows how PCT has it all wrong.

Best regards

Rick

···

---
Richard S. Marken
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400

Richard S. Marken
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400