Latash, 2010: "Motor Control Theories and Their Applications"

Thanks Rick, yes, that's coming really close to how I was thinking about it. Putting aside whether it is closed loop for a moment, I think it is interesting that they choose length and rate of change of length as what we could call the controlled perceptual references and they don't have force as a sensed or controlled variable. This is what our joystick paradigm is doing at the moment and getting good results. Just like Bill's inverted pendulum model, I am predicting that controlling position at an upper level in a hierarchy to set the reference rate of change in position will lead to a more accurate model. This is because if the limb is already moving at a fast enough rate to the reference position but not got there yet, then no error will be detected and if it is too fast it will be slowed down, helping it not to overshoot its target reference position. I know the Latash people don't understand it this way, but the physiological data they have used can help us focus on what the controlled variables are likely to be and our PCT model is consistent with other findings and known physiological constraints. Given that PCT is such a broad theory I think we need domain-specific information to build the models. That said, most of the threshold model is not original and based on physiology that was understood by Bill prior to 1973... So it is a hard political game to play...
Warren

···

Sent from my iPhone

On 31 Dec 2013, at 20:31, Richard Marken <rsmarken@GMAIL.COM> wrote:

[From Rick Marken (2013.12.31.1230)]

Bruce Abbott (2013.12.30.1930 EST)--

RM: So the lambda model is what I thought it was in the first place; an
output generation mode, not a control model. It will fail miserably as soon
as you include disturbances to the apparently controlled variable.

BA: If you examine the larger portion of Figure 3 (showing a representation
of the skull and jaw at right, you will see an arrow labeled "muscle lengths
and rate of change in muscle lengths" running back to the muscles. These
changes alter the outputs of the muscle spindles, contributing to motor
neuron depolarization thresholds, so that a lengthening muscle generates a
feedback signal that opposes the lengthening. Any disturbance that began to
stretch the muscle beyond the "equilibrium point" established by the lambda
control inputs generates this opposing force.

RM: You're right. If I read the arrows as representing neural signal
paths and the boxes as comparators (points of synapse), then the
central command (lambda) is a reference signal and "length" and "rate
of change in length" are perceptual signals entering a comparator
function called "Length Dependent". The comparator is presumably what
carries out the calculation of the muscle activation signal per
equation 1 (or 3, which takes neural delays into account). Actually
the output of the comparator (A in equations 1 and 3) goes through a
series of subsequent functions (called "Time Dependent " and Velocity
Dependent") but eventually the output signal produces a force, which
has feedback effects on the "length" and "rate of change in length"
perceptions.

So if I'm reading this right this is precisely a PCT model of control
of a perception of a combination of muscle length and rate of change
in muscle length. This perception is being controlled relative to
lambda, the reference input. Their diagram does leave out some of the
nice functional labels that we would include in a PCT model,
particularly the "length" and "rate of change in length" perceptual
functions; it also leaves out disturbances to the controlled variable.
Disturbances are forces that affect the muscle length/rate of change
perception that are independent of the forces produced by the control
system. So those forces would enter from outside the "Force generating
mechanisms" box and join with the forces generated by the control
system at the circle containing the plus sign, adding to the forces
that have an affect on the controlled variable (muscle length/rate of
change). Maybe this is what Bill Powers saw as their model being
similar to PCT.

So I would have to agree that this model is not only similar to PCT;
it is exactly PCT. At least, that's true of the diagram. So it's not
an equilibrium model; it's a closed loop control model, just like PCT.
It's a muscle length control model, like the one Bill implemented in
the little man, except that it adds the derivative of muscle length
into the percpetion for some reason.

From my point of view I see nothing PCT has to learn from this model;
but I see a lot (in terms of clarity) that the people who developed
this model could learn from PCT.

I've just downloaded their more "detailed" presentation of the model
--Laboissiere , Ostry and Feldman ( in press ) -- because the math
that they show in the Appendix to the subject paper doesn't seem to
match the model shown in the diagram. The main problem is that the
math doesn't describe the closed loop that is implied by the diagram.
Because there is no feedback equation showing the dependence of
length/rate of change in length on the output of the system the
length/rate of change perception can be treated as though it were an
independent (rather than a controlled) variable. I suspect that this
is what is going on because the behavior that they show in figure 5
could be produced by a filtered version of the "command signal" and
the series of functions they show in the diagram in Figure 3 looks a
lot like a cascade of filters.

I've scanned the detailed model paper and I can't find any feedback
function stuff in their either.

I guess I'm at the point where I can't tell what the heck this paper
shows. The diagram suggests that they have a control model of muscle
length (with the addition of the derivative of muscle length into the
perception). Their math suggests that they have an output generation
model, where the output generated by the lambda command signal is
filtered by the processes that turn the difference between lambda and
length/rate of change in length into the forces that change length. If
it's a PCT model, then the main question would be whether they got the
controlled variable right; is it really length/rate of change in
length that is controlled or just length? This should be easily
testable. It it's an output generation model then it is completely
different than PCT and it should be easy to show that it fails where
PCT succeeds.

So I suggest that, rather than arguing about whether these equilibrium
theory folks have anything to teach (or learn from) us, we should just
implement their model as a computer program. Once we agree that we've
got their model right we can go on to see what we can learn from it. A
computer model would (I think) go a long way to helping me see how
their model works and what it does.

I can tell you one thing their model does for me; it makes me long for
the clarity and simplicity of PCT.

Best

Rick
--
Richard S. Marken PhD
www.mindreadings.com

The only thing that will redeem mankind is cooperation.
                                                  -- Bertrand Russell

[From Bruce Abbott (2014.01.01.0910 EST)]

Happy New Year everyone!

Warren, you've practically taken the words right out of my mouth (off of my
screen?) -- it was going to be the subject of my next post. While reading
the various papers on the equilibrium point model I couldn't help but wonder
whether much (if not all) of the phenomena that EP has been designed to
account for easily might be handled by a hierarchical model and/or a model
controlling several variable independently at a single level. In the first
category are Bill Powers' inverted pendulum model and Rick Marken's baseball
fielder model, both of which exhibit apparently complex behavior that would
appear to require complex cognitive processing and decision-making, but
which accomplish their goals without relying on any such processes. The
inverted pendulum, for example, responds to a change in the reference
horizontal position of the bob by first moving the cart in the OPPOSITE
direction -- away from the reference -- (which tips the pendulum rod toward
the target position), then reverses direction and accelerates the cart
enough to keep the bob from falling while maintaining the rod's tilt, thus
accelerating the bob horizontally toward the bob reference while reducing
the bob's vertical acceleration to zero. As the bob approaches its
reference position the cart accelerates further so that its velocity now
exceeds that of the bob, thus raising the bob while getting ahead of it. The
cart then slows. As the pendulum rod is now leaning slightly backward, this
decelerates the bob while continuing to raise it. The cart decelerates just
enough to being both itself and the bob to a stop just as the bob reaches
its reference position, with the cart directly beneath it. All this is
accomplished by a four-level control hierarchy, properly tuned, that acts
smoothly and continuously on bob acceleration, bob velocity, bob position,
and cart velocity -- there is not a single "if this, then that" decision
being made anywhere in the system. Rick's baseball fielder exhibits similar
apparent judgment and decision-making in a similar way.

Bill Powers' Crowd demo illustrates the approach in which control systems
acting at a single level accomplish life-like behavior, the "people" doing
things like avoiding obstacles and following each other while simultaneously
seeking a goal position on the opposite side of the screen. Simple control
systems in each "person" control for proximity to another person, proximity
to the goal, and avoidance of obstacles.

The equilibrium point folks have posited complex control signals in their
model, arising from a higher level that orchestrates a complex pattern of
equilibrium point changes in order to have their model move the "end point"
of a limb (e.g., finger) rapidly from a starting position to a target
position. These changes in EP lambda signals provide the right
accelerations and decelerations to the limb (and its joints) at the right
times while also adjusting muscle "stiffness" as required, in order to match
their model's behavior to their observations. (Stiffness is increased by
tensing antagonistic muscles against one another, which in the model is done
by increasing the lambda signals of both muscles together; joint movement
occurs when the two lambdas are changed in opposite directions. Stiffness
can be varied independently of joint angular position.) A properly designed
PCT model of this system well might take account of these changes without
requiring any cognitive "motor program" at a higher level.

Bruce

···

----------------------------------------------------------------

WM: Warren Mansell
Sent: Wednesday, January 01, 2014 5:56 AM

Rick Marken

WM: Thanks Rick, yes, that's coming really close to how I was thinking about
it. Putting aside whether it is closed loop for a moment, I think it is
interesting that they choose length and rate of change of length as what we
could call the controlled perceptual references and they don't have force as
a sensed or controlled variable. This is what our joystick paradigm is doing
at the moment and getting good results. Just like Bill's inverted pendulum
model, I am predicting that controlling position at an upper level in a
hierarchy to set the reference rate of change in position will lead to a
more accurate model. This is because if the limb is already moving at a fast
enough rate to the reference position but not got there yet, then no error
will be detected and if it is too fast it will be slowed down, helping it
not to overshoot its target reference position. I know the Latash people
don't understand it this way, but the physiological data they have used can
help us focus on what the controlled variables are likely to be and our PCT
model is consistent with other findings and known physiological constraints.
Given that PCT is such a broad theory I think we need domain-specific
information to build the models. That said, most of the threshold model is
not original and based on physiology that was understood by Bill prior to
1973... So it is a hard political game to play...
Warren

On 31 Dec 2013, at 20:31, Richard Marken <rsmarken@GMAIL.COM> wrote:

I guess I'm at the point where I can't tell what the heck this paper
shows. The diagram suggests that they have a control model of muscle
length (with the addition of the derivative of muscle length into the
perception). Their math suggests that they have an output generation
model, where the output generated by the lambda command signal is
filtered by the processes that turn the difference between lambda and
length/rate of change in length into the forces that change length. If
it's a PCT model, then the main question would be whether they got the
controlled variable right; is it really length/rate of change in
length that is controlled or just length? This should be easily
testable. It it's an output generation model then it is completely
different than PCT and it should be easy to show that it fails where
PCT succeeds.

So I suggest that, rather than arguing about whether these equilibrium
theory folks have anything to teach (or learn from) us, we should just
implement their model as a computer program. Once we agree that we've
got their model right we can go on to see what we can learn from it. A
computer model would (I think) go a long way to helping me see how
their model works and what it does.

I can tell you one thing their model does for me; it makes me long for
the clarity and simplicity of PCT.

[From Bruce Abbott (2014.01.01.1000 EST)]

Rick Marken (2013.12.31.1230) --

RM: I guess I'm at the point where I can't tell what the heck this paper
shows. The diagram suggests that they have a control model of muscle length
(with the addition of the derivative of muscle length into the perception).
Their math suggests that they have an output generation model, where the
output generated by the lambda command signal is filtered by the processes
that turn the difference between lambda and length/rate of change in length
into the forces that change length. If it's a PCT model, then the main
question would be whether they got the controlled variable right; is it
really length/rate of change in length that is controlled or just length?
This should be easily testable. It it's an output generation model then it
is completely different than PCT and it should be easy to show that it
fails where PCT succeeds.

RM: So I suggest that, rather than arguing about whether these equilibrium
theory folks have anything to teach (or learn from) us, we should just
implement their model as a computer program. Once we agree that we've got
their model right we can go on to see what we can learn from it. A computer
model would (I think) go a long way to helping me see how their model works
and what it does.

Excellent proposal Rick, assuming that we can figure out exactly what that
model is. And once we understand what observations the EP model is designed
to account for, we could try out hands at developing a PCT model.

RM: I can tell you one thing their model does for me; it makes me long for
the clarity and simplicity of PCT.

BA: Me too.

Bruce

[From Matti Kolu (2013.12.24.0920 CET)]

"Biological objects belong to the physical world and they are alive.
So, help with formulating questions may be expected to come from
physics and biology (physiology), not from control theory and
engineering developed to deal with objects in the inanimate world.
Physics of the inanimate nature, while being a highly developed
science, has troubles dealing with typical problems of motor control.
First, in contrast to movements in the inanimate world, movements of
biological objects are intentional and purposeful. These two notions
cannot be easily incorporated into physics."

-- Latash, Levin, Scholz, Sch�ner. Motor control theories and their
applications. Medicina (Kaunas) 2010; 46 (6): 382-392.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3017756/ (free full-text)

Matti

Yes, Bill and I had some interactions with this group on email. They seem completely resistant to the fact that control theories are functional models of biological systems, such as homeostasis, and unwilling to appreciate that perceptual control is purposeful. It is very sad that they don’t want to work with us instead of against us!

Warren

···

On Tuesday, December 24, 2013, Matti Kolu wrote:

[From Matti Kolu (2013.12.24.0920 CET)]

"Biological objects belong to the physical world and they are alive.

So, help with formulating questions may be expected to come from

physics and biology (physiology), not from control theory and

engineering developed to deal with objects in the inanimate world.

Physics of the inanimate nature, while being a highly developed

science, has troubles dealing with typical problems of motor control.

First, in contrast to movements in the inanimate world, movements of

biological objects are intentional and purposeful. These two notions

cannot be easily incorporated into physics."

– Latash, Levin, Scholz, Schöner. Motor control theories and their

applications. Medicina (Kaunas) 2010; 46 (6): 382-392.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3017756/ (free full-text)

Matti


Dr Warren Mansell
Reader in Psychology
Cognitive Behavioural Therapist & Chartered Clinical Psychologist
School of Psychological Sciences
Coupland I

University of Manchester
Oxford Road
Manchester M13 9PL
Email: warren.mansell@manchester.ac.uk

Tel: +44 (0) 161 275 8589

Website: http://www.psych-sci.manchester.ac.uk/staff/131406

See teamstrial.net for further information on our trial of CBT for Bipolar Disorders in NW England

The highly acclaimed therapy manual on A Transdiagnostic Approach to CBT using Method of Levels is available now.

Check www.pctweb.org for further information on Perceptual Control Theory

[From Rick Marken (2013.12.25.1120)]

···

On Tue, Dec 24, 2013 at 12:14 PM, Warren Mansell wmansell@gmail.com wrote:

Yes, Bill and I had some interactions with this group on email. They seem completely resistant to the fact that control theories are functional models of biological systems, such as homeostasis, and unwilling to appreciate that perceptual control is purposeful. It is very sad that they don’t want to work with us instead of against us!

RM: Not too surprising considering they are pushing “physics” (in the form of wht is basically an “attractor” or “equilibrium” model) as the explanation purposive behavior and dismissing control theory. This paper is a perfect example of what I was talking about in my criticism of conflating physical stability with control. These guys are not interested in control theory precisely because they believe that purposive behavior (what we know to be control) can be explained by Newton’s instead of Powers’ laws;-)

Merry Merry

Rick

Warren

On Tuesday, December 24, 2013, Matti Kolu wrote:

[From Matti Kolu (2013.12.24.0920 CET)]

"Biological objects belong to the physical world and they are alive.

So, help with formulating questions may be expected to come from

physics and biology (physiology), not from control theory and

engineering developed to deal with objects in the inanimate world.

Physics of the inanimate nature, while being a highly developed

science, has troubles dealing with typical problems of motor control.

First, in contrast to movements in the inanimate world, movements of

biological objects are intentional and purposeful. These two notions

cannot be easily incorporated into physics."

– Latash, Levin, Scholz, Schöner. Motor control theories and their

applications. Medicina (Kaunas) 2010; 46 (6): 382-392.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3017756/ (free full-text)

Matti


Dr Warren Mansell
Reader in Psychology
Cognitive Behavioural Therapist & Chartered Clinical Psychologist
School of Psychological Sciences

Coupland I

University of Manchester
Oxford Road
Manchester M13 9PL
Email: warren.mansell@manchester.ac.uk

Tel: +44 (0) 161 275 8589

Website: http://www.psych-sci.manchester.ac.uk/staff/131406

See teamstrial.net for further information on our trial of CBT for Bipolar Disorders in NW England

The highly acclaimed therapy manual on A Transdiagnostic Approach to CBT using Method of Levels is available now.

Check www.pctweb.org for further information on Perceptual Control Theory


Richard S. Marken PhD
www.mindreadings.com
The only thing that will redeem mankind is cooperation.

                                               -- Bertrand Russell

[From Bruce Abbott (2013.12.26.10:20 EST)]

Rick Marken (2013.12.25.1120) –

···

On Tue, Dec 24, 2013 at 12:14 PM, Warren Mansell wmansell@gmail.com wrote:

Yes, Bill and I had some interactions with this group on email. They seem completely resistant to the fact that control theories are functional models of biological systems, such as homeostasis, and unwilling to appreciate that perceptual control is purposeful. It is very sad that they don’t want to work with us instead of against us!

RM: Not too surprising considering they are pushing “physics” (in the form of wht is basically an “attractor” or “equilibrium” model) as the explanation purposive behavior and dismissing control theory. This paper is a perfect example of what I was talking about in my criticism of conflating physical stability with control. These guys are not interested in control theory precisely because they believe that purposive behavior (what we know to be control) can be explained by Newton’s instead of Powers’ laws;-)

The above exchange was in reference to the Latash et al. (2010) paper reviewing two “motor control” hypotheses, including one called the equilibrium point hypothesis.

According to a Wiki article reviewing different models of “motor” control http://en.wikipedia.org/wiki/Degrees_of_freedom_problem ),

In the Equilibrium Point hypothesis, all movements are generated by the nervous system through a gradual transition of equilibrium points along a desired trajectory. “Equilibrium point” in this sense is taken to mean a state where a field has zero force, meaning opposing muscles are in a state of balance with each other, like two rubber bands pulling the joint to a stable position. Equilibrium point control is also called “threshold control” because signals sent from the CNS to the periphery are thought to modulate the threshold length of each muscle. In this theory, motor neurons send commands to muscles, which changes the force–length relation within a muscle, resulting in a shift of the system’s equilibrium point. The nervous system would not need to directly estimate limb dynamics, but rather muscles and spinal reflexes would provide all the necessary information about the system’s state.

I don’t know enough about the equilibrium point hypothesis to weigh in on its adequacy, but what Latash et al. (2010) describe isn’t a passive equilibrium system:

Fundamental to the EP theory is the notion that threshold position control underlies intentional motor actions. To perform such actions, electrochemical influences descending from the brain in the presence of proprioceptive feedback to motoneurons are transformed into changes in the threshold muscle lengths or joint angles at which these motoneurons begin to be recruited, thus setting the spatial activation range in reference to the body geometry. This allows control levels of the CNS to specify where, in spatial coordinates, muscles are activated without being concerned about exact details on when and how they are activated.

The EP is described as something that the CNS can set to varying values as a means of changing joint angles or relative muscle tensions; in fact, Latash et al. refer to it as a “referent” (reference?). And note the inclusion of proprioceptive feedback to motoneurons in their model. It theirs is an equilibrium model, how did THAT get in there?

Compare this to your description:

RM: Not too surprising considering they are pushing “physics” (in the form of what is basically an “attractor” or “equilibrium” model) as the explanation purposive behavior and dismissing control theory. This paper is a perfect example of what I was talking about in my criticism of conflating physical stability with control. These guys are not interested in control theory precisely because they believe that purposive behavior (what we know to be control) can be explained by Newton’s instead of Powers’ laws;-)

Nothing that you stated here is true. They are not pushing an attractor or equilibrium model (whatever it is, it is not that) as the explanation of purposive behavior, they are not dismissing control theory, and therefore they are not dismissing it because of a belief that behavior “can be explained by Newton’s instead of Powers’ laws.”)

Bruce

[From Erling Jorgensen (2013.12.26 1210 EST)]

Bruce Abbott (2013.12.26.10:20 EST)

The above exchange was in reference to the Latash et al. (2010) paper
reviewing two "motor control" hypotheses, including one called the
equilibrium point hypothesis.

...

In the Equilibrium Point hypothesis, all movements are generated by the
nervous system through a gradual transition of
<http://en.wikipedia.org/wiki/Equilibrium_point&gt; equilibrium points along a
desired trajectory. "Equilibrium point" in this sense is taken to mean a
state where a field has zero force, meaning opposing muscles are in a state
of balance with each other, like two rubber bands pulling the joint to a
stable position.

I was looking at some of this "equilibrium point" approach to motor control
a few months back, I think when Matti(?) raised a question about whether
joint angles per se get controlled.

The distinct impression I had was that the "equilibrium points along a
desired trajectory" provided a way to understand reference signals being
provided for joint positioning. I'll have to look more at the articles you
cite, to see if that impression is borne out.

Equilibrium point control is also called "threshold
control" because signals sent from the CNS to the periphery are thought to
modulate the threshold length of each muscle.

I remember getting a little confused by specifics such as what you cite
here, as well as how the muscles & joints operate & some of their
dynamics. But it did register that the sequence of equilibrium points
could be a specification for reference signals. So I am in agreement with
what you seem to be wondering about this EP theory. Thanks for bringing
the matter back up & providing the links.

All the best,
Erling

[Martin Taylor 2013.12.26.12.50]

Rick's comment doesn't seem to gibe very well with the bit Matti

quoted: “Physics of the inanimate nature, while being a highly
developed science, has troubles dealing with typical problems of
motor control. First, in contrast to movements in the inanimate
world, movements of biological objects are intentional and
purposeful. These two notions cannot be easily incorporated into
physics.”
Or with this, from the Latash et.al. paper: “In the most advanced
formulation of the EP theory, activity of each muscle ,
without any programming, depending on the difference between the
actual configuration of the body and its threshold (referent)
configuration, as well as on the rate of changes of this difference
().”
As is all too common, Rick has commented on something he imagines
his imagined adversary has said, rather than on what a legitimate
scientist has actually said. The wording “conflating physical
stability with control” is yet another illustration of something
that happened only in his own mind.
A legitimate complaint can, however, be raised against the paper,
given that PCT is very much an incorporation of purpose and intent
into physics. Maybe that fact is one of the issues that make it hard
for PCT to be accepted into the mainstream.
Martin

···

[From Rick Marken (2013.12.25.1120)]

        On Tue, Dec 24, 2013 at 12:14 PM,

Warren Mansell wmansell@gmail.com wrote:

          Yes,

Bill and I had some interactions with this group on email.
They seem completely resistant to the fact that control
theories are functional models of biological systems, such
as homeostasis, and unwilling to appreciate that
perceptual control is purposeful. It is very sad that they
don’t want to work with us instead of against us!

          RM: Not too surprising considering they are pushing

“physics” (in the form of wht is basically an “attractor”
or “equilibrium” model) as the explanation purposive
behavior and dismissing control theory. This paper is a
perfect example of what I was talking about in my
criticism of conflating physical stability with control.
These guys are not interested in control theory precisely
because they believe that purposive behavior (what we know
to be control) can be explained by Newton’s instead of
Powers’ laws;-)

emergesFeldman, in press

                    On Tuesday, December 24, 2013, Matti Kolu

wrote:

                      [From Matti Kolu

(2013.12.24.0920 CET)]

                      "Biological objects belong to the physical

world and they are alive.

                      So, help with formulating questions may be

expected to come from

                      physics and biology (physiology), not from

control theory and

                      engineering developed to deal with objects in

the inanimate world.

                      Physics of the inanimate nature, while being a

highly developed

                      science, has troubles dealing with typical

problems of motor control.

                      First, in contrast to movements in the

inanimate world, movements of

                      biological objects are intentional and

purposeful. These two notions

                      cannot be easily incorporated into physics."



                      -- Latash, Levin, Scholz, Schöner. Motor

control theories and their

                      applications. Medicina (Kaunas) 2010; 46 (6):

382-392.

                      [http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3017756/](http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3017756/)
                      (free full-text)



                      Matti

[From Rick Marken (2013.12.26.1400)]

Bruce Abbott (2013.12.26.10:20 EST)

Rick Marken (2013.12.25.1120) –

WM: Yes, Bill and I had some interactions with this group on email. They seem completely resistant to the fact that control theories are functional models of biological systems, such as homeostasis, and unwilling to appreciate that perceptual control is purposeful. It is very sad that they don’t want to work with us instead of against us!

RM: Not too surprising considering they are pushing “physics” (in the form of wht is basically an “attractor” or “equilibrium” model) as the explanation purposive behavior and dismissing control theory.

BA: The EP is described as something that the CNS can set to varying values as a means of changing joint angles or relative muscle tensions; in fact, Latash et al. refer to it as a “referent” (reference?). And note the inclusion of proprioceptive feedback to motoneurons in their model. It theirs is an equilibrium model, how did THAT get in there?

Compare this to your description:

RM: Not too surprising considering they are pushing “physics” (in the form of what is basically an “attractor” or “equilibrium” model) as the explanation purposive behavior and dismissing control theory.

BA: Nothing that you stated here is true. They are not pushing an attractor or equilibrium model (whatever it is, it is not that) as the explanation of purposive behavior, they are not dismissing control theory, and therefore they are not dismissing it because of a belief that behavior “can be explained by Newton’s instead of Powers’ laws.”)

RM: So what are they pushing as an explanation of purposive behavior? The fact is I’m not sure I know what the heck they are talking about in that paper. I tried my best to make sense of it based on their verbal descriptions but, truth be told, I don’t have a very clear idea of what their model is.

You seem to understand it, Bruce; how about a functional diagram of their model. For example, I have no idea what this means : “To perform such actions, electrochemical influences descending from the brain in the presence of proprioceptive feedback to motoneurons are transformed into changes in the threshold muscle lengths or joint angles at which these motoneurons begin to be recruited, thus setting the spatial activation range in reference to the body geometry.” I have no idea how to translate this into a functional model. What, for example, does “…in the presence of proprioceptive feedback” mean? Are the referents specifications for particular values of proprioceptive feedback? If so, what proprioceptive feedback are they talking about; muscle length, muscle tension? What’s the controlled proprioceptive variable? So many questions, so many hands waving in my face;-)

To me this article is just a bunch of blather – a Rorshach test in which people can see traces of PCT if they want. When I look at it I see attractor theory; the referent (R) seems to set the attractor point of limb position by setting the appropriate tensions of the muscles (it’s not clear how it knows that the muscles tensions are appropriately set; maybe that’s what the proprioceptive signals are for). Sort of like tipping the bowl this way and that so that the ball ends up in different positions on the bowl (gravity being the attractor). I’m sure someone else could look at it and see some other theory.

I guess my feeling at this point is that whatever theory of purposive behavior is being proposed by Latish et al it is so incoherent that it’s not clear whether it is similar to or a competitor to PCT. To me it suggests attractor theory; no control involved. Maybe that’s just because I’m too stupid to understand it. But when I compare Latish et al’s description of their theory to Bill Powers description of PCT it’s just night and day; not because one theory is better than the other but because one theory (PCT) is so clearly and completely described while the other is so completely opaque.

For example, Powers was very careful to define and describe the phenomenon that his theory explains: control, which is also known as purposive behavior. He maps out all the variables involved in control and makes it concrete in terms of a simple control task: compensatory tracking. He then explains the model that accounts for the phenomenon of control (control theory) and shows how the model maps to the controller/environment system and details how the model could be implemented in a nervous system (given what we know of how the NS works) as well as in a computer simulation. And he showed exactly how the model accounts (nearly perfectly) for observed control behavior.

There is none of this in the Latish et al paper. They talk about purposive behavior but never explain what it is. They never say exactly what behavior they are trying to account for. They give only a verbal description of the theory; I can find no hint of how one might turn it into a working model. And they don’t show how the model accounts for the experimental results they present (which I find somewhat incomprehensible as well; I really don’t have a good idea of what the heck the subjects were doing in the flexor torque study). That’s why I think of it all as hand-waving; there is just no coherent there there, for me anyway.

But maybe my lack of understanding of Latish et al is a result of my own stupidity (is there such a thing as “grandchild simple”, something like “blood simple”?). So it would be great if Bruce Abbott or Martin Taylor, both of whom are orders of magnitude smarter than me, would help me see just what model is being proposed by Latish et al. I would really appreciate it if they could do it in nice, simple terms – using diagrams and pictures; simple equations would be OK too. That’s what I loved about Bill: even a dumb guy like me could understand him (Yes, Boris, I got my PhD from a grab bag at Pep Boys Auto Parts;-).

Thanks.

Best

Rick

···


Richard S. Marken PhD
www.mindreadings.com

The only thing that will redeem mankind is cooperation.

                                               -- Bertrand Russell

[From Rick Marken (2013.12.26.1505)]

···

Martin Taylor (2013.12.26.12.50)–

MT: As is all too common, Rick has commented on something he imagines

his imagined adversary has said, rather than on what a legitimate
scientist has actually said.

RM: Leaving aside the question of who is the “legitimate scientist”, I agree that imagination was involved in my response to the Latish et al paper. But what I imagined was not what they said but what they meant by what they said. That is, I was trying to imagine what their model was – and how it worked – based on the words they used to describe it. As I said in my earlier post, their description of their model was not even close to the clarity of Powers’ description of the PCT model. So I ask you (along with Bruce) to please tell me what their model is – ow it works – in a functional diagram, hopefully, so that I can simulate it on a computer and we can design experimental tests to discriminate between their model and PCT, just as though we were legitimate scientists too;-)

Best

Rick

Richard S. Marken PhD
www.mindreadings.com
The only thing that will redeem mankind is cooperation.
– Bertrand Russell

[From Bruce Abbott (2013.12.26.2045 EST)]

Rick Marken (2013.12.26.1400)

Bruce Abbott (2013.12.26.10:20 EST)

Rick Marken (2013.12.25.1120) –

BA: The EP is described as something that the CNS can set to varying values as a means of changing joint angles or relative muscle tensions; in fact, Latash et al. refer to it as a “referent” (reference?). And note the inclusion of proprioceptive feedback to motoneurons in their model. It theirs is an equilibrium model, how did THAT get in there?

Compare this to your description:

RM: Not too surprising considering they are pushing “physics” (in the form of what is basically an “attractor” or “equilibrium” model) as the explanation purposive behavior and dismissing control theory.

BA: Nothing that you stated here is true. They are not pushing an attractor or equilibrium model (whatever it is, it is not that) as the explanation of purposive behavior, they are not dismissing control theory, and therefore they are not dismissing it because of a belief that behavior “can be explained by Newton’s instead of Powers’ laws.”)

RM: So what are they pushing as an explanation of purposive behavior? The fact is I’m not sure I know what the heck they are talking about in that paper. I tried my best to make sense of it based on their verbal descriptions but, truth be told, I don’t have a very clear idea of what their model is.

RM: You seem to understand it, Bruce; how about a functional diagram of their model. For example, I have no idea what this means : “To perform such actions, electrochemical influences descending from the brain in the presence of proprioceptive feedback to motoneurons are transformed into changes in the threshold muscle lengths or joint angles at which these motoneurons begin to be recruited, thus setting the spatial activation range in reference to the body geometry.” I have no idea how to translate this into a functional model. What, for example, does “…in the presence of proprioceptive feedback” mean? Are the referents specifications for particular values of proprioceptive feedback? If so, what proprioceptive feedback are they talking about; muscle length, muscle tension? What’s the controlled proprioceptive variable? So many questions, so many hands waving in my face;-)

You omitted the following from my post:

BA: I don’t know enough about the equilibrium point hypothesis to weigh in on its adequacy, but what Latash et al. (2010) describe isn’t a passive equilibrium system:

So, how do you get from there to the idea that I understand the equilibrium point hypothesis? All I said is that I understand enough to tell that it isn’t the passive equilibrium system model you asserted that it is.

Like you, I find the language Latash et al. use to describe their hypothesis rather difficult to decipher. Perhaps by reviewing the literature involved (including previous descriptions of the hypothesis and the empirical basis for it) I might get a clearer picture of it. Until then, I simply have nothing further to say about it.

The problem for PCT is that no one involved with PCT has proposed a competing model that will either (a) account for the observations on which the equilibrium hypothesis was built, or (b) show those observations to be either inaccurate or perhaps an irrelevant side-effect of the system’s operation. It’s that kind of model, together with supporting physiological evidence, that is needed if PCT is to gain greater acceptance in the scientific community.

Bruce

[From Rick Marken (2013.12.26.2035)]

Bruce Abbott (2013.12.26.2045 EST) –

RM: So what are they pushing as an explanation of purposive behavior?

BA: You omitted the following from my post:

BA:I don’t know enough about the equilibrium point hypothesis to weigh in on its adequacy, but what Latash et al. (2010) describe isn’t a passive equilibrium system:

BA: So, how do you get from there to the idea that I understand the equilibrium point hypothesis?

RM: Well you know enough to say that it is an “active” system, by which I presume you mean that it is a system that actively compensates for disturbance like a control system, rather than one that passively responds to a disturbance after it is removed, like an attractor system (like the ball in the bowl). The Latash et al. model looks to me just like a “passive” attractor system.

BA: The problem for PCT is that no one involved with PCT has proposed a competing model that will either (a) account for the observations on which the equilibrium hypothesis was built, or (b) show those observations to be either inaccurate or perhaps an irrelevant side-effect of the system’s operation.

RM: Actually, someone has. Me. The competing model and the demonstrations that observations on which equilibrium models are based are irrelevant side effects of control are described in my Psychological Science “Degrees of freedom in behavior” paper, which is reprinted in Mind Readings, p. 185. I think Bill has done this, too, in his “Little Man” model, with which I believe you are quite familiar;-) The Little Man does everything the equilibrium model says it does (move a limb purposefully) and it does it backwards and in high heels … er, I mean in the presence of continuously varying disturbance, something that no equilibrium model can even dream of doing.

BA: It’s that kind of model, together with supporting physiological evidence, that is needed if PCT is to gain greater acceptance in the scientific community.

RM: Well, my Psych Science paper didn’t seem to do it. Maybe there is some other more compelling observations on which the equilibrium hypothesis is built. If you would let me know what they are I’ll work on building a model that can handle them.

But I’m willing to bet my ill-gotten PhD that nothing will convince believers in the equilibrium hypothesis (and similar “dynamic attractor” models) that control theory is the solution to the problem of understanding purposeful behavior. Because I know that they are absolutely convinced that purposive behavior can be accounted for in lineal cause-effect terms.

Best

Rick

···


Richard S. Marken PhD
www.mindreadings.com

The only thing that will redeem mankind is cooperation.

                                               -- Bertrand Russell

[From Bruce Abbott (2013.12.27.0850 EST)]

Rick Marken (2013.12.26.2035)]

Bruce Abbott (2013.12.26.2045 EST) –

RM: So what are they pushing as an explanation of purposive behavior?

BA: You omitted the following from my post:

BA:I don’t know enough about the equilibrium point hypothesis to weigh in on its adequacy, but what Latash et al. (2010) describe isn’t a passive equilibrium system:

BA: So, how do you get from there to the idea that I understand the equilibrium point hypothesis?

RM: Well you know enough to say that it is an “active” system, by which I presume you mean that it is a system that actively compensates for disturbance like a control system, rather than one that passively responds to a disturbance after it is removed, like an attractor system (like the ball in the bowl). The Latash et al. model looks to me just like a “passive” attractor system.

It’s an active system because it actively alters λ, a system parameter that affects the equilibrium point in their model, and this changes the balance of muscle tensions of opposing muscles across the joint, leading to movement or a change in exerted force. (Or so it seems from my admittedly limited understanding of their model.) An equilibrium model wouldn’t produce any movements unless you gave it an impulse disturbance, after which the limb in question would just settle back to its pre-disturbance position.

BA: The problem for PCT is that no one involved with PCT has proposed a competing model that will either (a) account for the observations on which the equilibrium hypothesis was built, or (b) show those observations to be either inaccurate or perhaps an irrelevant side-effect of the system’s operation.

RM: Actually, someone has. Me. The competing model and the demonstrations that observations on which equilibrium models are based are irrelevant side effects of control are described in my Psychological Science “Degrees of freedom in behavior” paper, which is reprinted in Mind Readings, p. 185. I think Bill has done this, too, in his “Little Man” model, with which I believe you are quite familiar;-) The Little Man does everything the equilibrium model says it does (move a limb purposefully) and it does it backwards and in high heels … er, I mean in the presence of continuously varying disturbance, something that no equilibrium model can even dream of doing.

Hmmm. Bill’s “Little Man” model doesn’t model any actual physiology that must be involved in limb movement. The outputs of the control systems in the model somehow cause changes in joint angle without specifying how the outputs of these systems accomplish that action in the body. Little Man was not intended to include that level of detail; it was designed to show how a particular set of control systems might be organized to move a finger-tip on the end of a human arm to a target location in 3D space. Bill assumed that the visual system produces signals indicating the position of the target, but as we do not yet know enough about how the visual system does this, he substituted a bit of trigonometry to provide similar information in the Little Man model. Similarly, the Little Man model does not include the details of the first-order control systems that contract the muscles – no motor neurons, muscle spindles, gamma afferents, and so on. It is just assumed that the first-order systems will do their jobs and produce perceptions (e.g., sensed joint angles) that match their references.

I don’t believe that your paper deals with this level of physiological detail either. So neither your paper nor Bill’s Little Man demo offer a competing explanation for the observations that Latash et al.’s model deals with.

What Latash and colleagues have been attempting is a detailed model of the first-order systems involved in controlling joint angle. Bill provided such a model in B:CP based on what was known at the time about the anatomy and physiology involving control of a single muscle (such as a biceps), but “motor control” researchers apparently have encountered observations that this relatively straight-forward model does not account for. However, PCT doesn’t stand or fall on the way a particular model is organized; all that is necessary is a demonstration that the system works by controlling relevant perceptions. The specifics of the models can be altered as required by reliable observations so as to account for the data.

BA: It’s that kind of model, together with supporting physiological evidence, that is needed if PCT is to gain greater acceptance in the scientific community.

RM: Well, my Psych Science paper didn’t seem to do it. Maybe there is some other more compelling observations on which the equilibrium hypothesis is built. If you would let me know what they are I’ll work on building a model that can handle them.

Hey, it’s not my job to “let you know what they are.” I think it would be great if you would develop such a model; in fact if you succeeded it would be a great leap forward for PCT. But you will have to immerse yourself in what is by now an extensive literature in order to become familiar enough with the relevant observations to know what the model will have to incorporate within its structure, and what data it will have to account for.

But I’m willing to bet my ill-gotten PhD that nothing will convince believers in the equilibrium hypothesis (and similar “dynamic attractor” models) that control theory is the solution to the problem of understanding purposeful behavior. Because I know that they are absolutely convinced that purposive behavior can be accounted for in lineal cause-effect terms.

Well, you may be convinced of that, but I’m not.

What is “the equilibrium hypothesis” and how is it a “dynamic attractor” model? (Dynamic attractors are just descriptions of system behavior, they are not models.)

Bruce

[From Bruce Abbott (2013.12.26.1225 EST)]

Thanks, Erling, I appreciate your interest in this topic.

Bruce

Erling Jorgensen (2013.12.26 1210 EST)]

Bruce Abbott (2013.12.26.10:20 EST)

The above exchange was in reference to the Latash et al. (2010) paper
reviewing two "motor control" hypotheses, including one called the
equilibrium point hypothesis.

...

In the Equilibrium Point hypothesis, all movements are generated by the
nervous system through a gradual transition of
<http://en.wikipedia.org/wiki/Equilibrium_point&gt; equilibrium points
along a desired trajectory. "Equilibrium point" in this sense is taken
to mean a state where a field has zero force, meaning opposing muscles
are in a state of balance with each other, like two rubber bands
pulling the joint to a stable position.

I was looking at some of this "equilibrium point" approach to motor control
a few months back, I think when Matti(?) raised a question about whether
joint angles per se get controlled.

The distinct impression I had was that the "equilibrium points along a
desired trajectory" provided a way to understand reference signals being
provided for joint positioning. I'll have to look more at the articles you
cite, to see if that impression is borne out.

Equilibrium point control is also called "threshold control" because
signals sent from the CNS to the periphery are thought to modulate the
threshold length of each muscle.

I remember getting a little confused by specifics such as what you cite
here, as well as how the muscles & joints operate & some of their dynamics.
But it did register that the sequence of equilibrium points could be a
specification for reference signals. So I am in agreement with what you
seem to be wondering about this EP theory. Thanks for bringing the matter
back up & providing the links.

All the best,
Erling

···

-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2014.0.4259 / Virus Database: 3658/6952 - Release Date: 12/26/13

[From Rick Marken (2013.12.27.0940)]

Bruce Abbott (2013.12.27.0850 EST)--

BA: It's an active system because it actively alters οΏ½, a system parameter that affects the equilibrium point in their model

RM: I see that as just a motor output generation model with οΏ½ being the motor output that, as you said, sets the equilibrium point. It is not a reference for an input. This is not what I meant by "active", though, since this system is no different from any other "control of output" motor control model, all of which (unlike PCT) view behavior as a process of output generation. In this case, the output is the setting of the tensions in a group of muscle, which defines an equilibrium point, in the sense that if a transient force is applied to the limbs attached to the muscles the limbs will spring back to the equilibrium setting. That is not an active process from my point of view; it's just passive output generation: S-R.

BA: An equilibrium model wouldn't produce any movements unless you gave it an impulse disturbance, after which the limb in question would just settle back to its pre-disturbance position.

RM: Actually, it would produce movements resulting from the changes in
the setting of οΏ½, the equilibrium point. Variations in οΏ½ will produce
very nice movements in a disturbance free environment. I found a nice
paper showing life like movements produced by variations in οΏ½ in a
human avatar. οΏ½ is kind of like a faux reference setting; if there are
no disturbances in the environment then variations in οΏ½ will look like
"controlled" movements. But put in disturbances and you will see that
controlled output models of purposeful behavior don't work.

BA: I don't believe that your paper deals with this level of physiological detail either. So neither your paper nor Bill's Little Man demo offer a competing explanation for the observations that Latash et al.'s model deals with.

RM: I see only behavioral data in the Latash et al model. And to the
extent that I understand their experiment, I think PCT can provide a
very strong competing explanation of these observations. Indeed, I
didn't even see how Latash et al. explain their findings using a
model.

BA: What Latash and colleagues have been attempting is a detailed model of the first-order systems involved in controlling joint angle.

RM: If so, I certainly didn't pick that up from the paper. I didn't
see any physiological model, certainly not a model as detailed as the
physiological model described in B:CP.

BA: Bill provided such a model in B:CP based on what was known at the time about the anatomy and physiology involving control of a single muscle (such as a biceps), but "motor control" researchers apparently have encountered observations that this relatively straight-forward model does not account for.

RM: I think motor control researchers have "encountered observations"
that make them feel comfortable about their S-R approach to
understanding purposive behavior. I think one's model of behavior has
a big influence on what one "observes" about nervous system behavior.

RM: Well, my Psych Science paper didn't seem to do it. Maybe there is some other more compelling observations on which the equilibrium hypothesis is built. If you would let me know what they are I'll work on building a model that can handle them.

BA: Hey, it's not my job to "let you know what they are."

RM: OK, so it's just your job to say that there are observations that
PCT can't handle but I won't tell you what they are? Great job.

RM: But I'm willing to bet my ill-gotten PhD that _nothing_ will convince believers in the equilibrium hypothesis (and similar "dynamic attractor" models) that control theory is the solution to the problem of understanding purposeful behavior. Because I know that they are absolutely convinced that purposive behavior can be accounted for in lineal cause-effect terms.

BA: Well, you may be convinced of that, but I'm not.

RM: I know;-)

BA: What is "the equilibrium hypothesis" and how is it a "dynamic attractor" model? (Dynamic attractors are just descriptions of system behavior, they are not models.)

RM: I think that's a good distinction. Attractor models are
descriptive or what I would call "curve fitting" models. "Equilibrium
hypothesis" models are often working versions of "attractor" models;
the equilibrium point (defined by the relative tensions in a set of
muscles, for example) is the functional equivalent of an attractor.

Best

Rick

···

--
Richard S. Marken PhD
www.mindreadings.com

The only thing that will redeem mankind is cooperation.
                                                   -- Bertrand Russell

Hi, I am in line with Bruce here. In our discussions with them, Bill and I were more frustrated that they didn't acknowledge that PCT-like idea were already in their model, than thinking they were S-R. Bill was will to concede and learn from them concerning the equilibrium idea because he recognised that muscles are on paired groups across joints, but only if they could see the relevance of PCT to their overall functioning models, which they didn't, refusing to accept any connection between control theory and purposeful biological systems. I still think their physiological ideas could be helpful to us but within a HPCT model of the nervous system...
Warren

···

Sent from my iPhone

On 27 Dec 2013, at 17:42, Richard Marken <rsmarken@GMAIL.COM> wrote:

[From Rick Marken (2013.12.27.0940)]

Bruce Abbott (2013.12.27.0850 EST)--

BA: It's an active system because it actively alters λ, a system parameter that affects the equilibrium point in their model

RM: I see that as just a motor output generation model with λ being the motor output that, as you said, sets the equilibrium point. It is not a reference for an input. This is not what I meant by "active", though, since this system is no different from any other "control of output" motor control model, all of which (unlike PCT) view behavior as a process of output generation. In this case, the output is the setting of the tensions in a group of muscle, which defines an equilibrium point, in the sense that if a transient force is applied to the limbs attached to the muscles the limbs will spring back to the equilibrium setting. That is not an active process from my point of view; it's just passive output generation: S-R.

BA: An equilibrium model wouldn't produce any movements unless you gave it an impulse disturbance, after which the limb in question would just settle back to its pre-disturbance position.

RM: Actually, it would produce movements resulting from the changes in
the setting of λ, the equilibrium point. Variations in λ will produce
very nice movements in a disturbance free environment. I found a nice
paper showing life like movements produced by variations in λ in a
human avatar. λ is kind of like a faux reference setting; if there are
no disturbances in the environment then variations in λ will look like
"controlled" movements. But put in disturbances and you will see that
controlled output models of purposeful behavior don't work.

BA: I don't believe that your paper deals with this level of physiological detail either. So neither your paper nor Bill's Little Man demo offer a competing explanation for the observations that Latash et al.'s model deals with.

RM: I see only behavioral data in the Latash et al model. And to the
extent that I understand their experiment, I think PCT can provide a
very strong competing explanation of these observations. Indeed, I
didn't even see how Latash et al. explain their findings using a
model.

BA: What Latash and colleagues have been attempting is a detailed model of the first-order systems involved in controlling joint angle.

RM: If so, I certainly didn't pick that up from the paper. I didn't
see any physiological model, certainly not a model as detailed as the
physiological model described in B:CP.

BA: Bill provided such a model in B:CP based on what was known at the time about the anatomy and physiology involving control of a single muscle (such as a biceps), but "motor control" researchers apparently have encountered observations that this relatively straight-forward model does not account for.

RM: I think motor control researchers have "encountered observations"
that make them feel comfortable about their S-R approach to
understanding purposive behavior. I think one's model of behavior has
a big influence on what one "observes" about nervous system behavior.

RM: Well, my Psych Science paper didn't seem to do it. Maybe there is some other more compelling observations on which the equilibrium hypothesis is built. If you would let me know what they are I'll work on building a model that can handle them.

BA: Hey, it's not my job to "let you know what they are."

RM: OK, so it's just your job to say that there are observations that
PCT can't handle but I won't tell you what they are? Great job.

RM: But I'm willing to bet my ill-gotten PhD that _nothing_ will convince believers in the equilibrium hypothesis (and similar "dynamic attractor" models) that control theory is the solution to the problem of understanding purposeful behavior. Because I know that they are absolutely convinced that purposive behavior can be accounted for in lineal cause-effect terms.

BA: Well, you may be convinced of that, but I'm not.

RM: I know;-)

BA: What is "the equilibrium hypothesis" and how is it a "dynamic attractor" model? (Dynamic attractors are just descriptions of system behavior, they are not models.)

RM: I think that's a good distinction. Attractor models are
descriptive or what I would call "curve fitting" models. "Equilibrium
hypothesis" models are often working versions of "attractor" models;
the equilibrium point (defined by the relative tensions in a set of
muscles, for example) is the functional equivalent of an attractor.

Best

Rick
--
Richard S. Marken PhD
www.mindreadings.com

The only thing that will redeem mankind is cooperation.
                                                  -- Bertrand Russell

[From Rick Marken (2013.12.28.1200)]

WM: Hi, I am in line with Bruce here. In our discussions with them, Bill and I were more frustrated that they didn't acknowledge that PCT-like idea were already in their model, than thinking they were S-R. Bill was will to concede and learn from them concerning the equilibrium idea because he recognised that muscles are on paired groups across joints, but only if they could see the relevance of PCT to their overall functioning models, which they didn't, refusing to accept any connection between control theory and purposeful biological systems. I still think their physiological ideas could be helpful to us but within a HPCT model of the nervous system...

RM: I think Bill often bent over too far backwards in the hope of
getting people to accept PCT. Bill knew (as I do) that you can't force
people to accept your ideas, even when you present them along with
facts (see my previous post). One of Bill's favorite sayings was "A
man convinced against his will is of the same opinion still". So Bill
often tried to sneak PCT in while saying how similar it was to things
people already believed in, hoping that such an approach would lead to
less resistance. And it did.

There was little resistance to Bill (until he started getting specific
at which point people often became livid). But there was also little
understanding of PCT. I actually can't think of _anyone_ who came to
accept (or understand) PCT as a result of Bill's strategy. In my
experience the only people who really accept and understand PCT are
those who 1) come to PCT without an existing, strongly held agenda and
2) saw the merits of PCT on their own and were willing to learn it
themselves (and/or with Bill's help).

I happen to take a different approach to dealing with opponents of PCT
than Bill (which is why he often got mad at me on the net). I'm less
accepting of what I consider to be wrong ideas so I'm not expecting to
be as well loved as Bill. But I've got enough friends to keep me happy
and I prefer the true satisfaction of understanding PCT over the false
high of what Mary Powers called "understandingness". Like Bill, I
don't mind being wrong; but I also don't like using false flattery to
get people to agree with me, and I do think Bill did that a bit too
much; it's the only criticism I have of Bill but it pales into
insignificance when one considers his incredible intellectual
accomplishments and kind and gentle nature.

Best regards

Rick

···

On Sat, Dec 28, 2013 at 7:50 AM, Warren Mansell <wmansell@gmail.com> wrote:

Warren

Sent from my iPhone

On 27 Dec 2013, at 17:42, Richard Marken <rsmarken@GMAIL.COM> wrote:

[From Rick Marken (2013.12.27.0940)]

Bruce Abbott (2013.12.27.0850 EST)--

BA: It's an active system because it actively alters οΏ½, a system parameter that affects the equilibrium point in their model

RM: I see that as just a motor output generation model with οΏ½ being the motor output that, as you said, sets the equilibrium point. It is not a reference for an input. This is not what I meant by "active", though, since this system is no different from any other "control of output" motor control model, all of which (unlike PCT) view behavior as a process of output generation. In this case, the output is the setting of the tensions in a group of muscle, which defines an equilibrium point, in the sense that if a transient force is applied to the limbs attached to the muscles the limbs will spring back to the equilibrium setting. That is not an active process from my point of view; it's just passive output generation: S-R.

BA: An equilibrium model wouldn't produce any movements unless you gave it an impulse disturbance, after which the limb in question would just settle back to its pre-disturbance position.

RM: Actually, it would produce movements resulting from the changes in
the setting of οΏ½, the equilibrium point. Variations in οΏ½ will produce
very nice movements in a disturbance free environment. I found a nice
paper showing life like movements produced by variations in οΏ½ in a
human avatar. οΏ½ is kind of like a faux reference setting; if there are
no disturbances in the environment then variations in οΏ½ will look like
"controlled" movements. But put in disturbances and you will see that
controlled output models of purposeful behavior don't work.

BA: I don't believe that your paper deals with this level of physiological detail either. So neither your paper nor Bill's Little Man demo offer a competing explanation for the observations that Latash et al.'s model deals with.

RM: I see only behavioral data in the Latash et al model. And to the
extent that I understand their experiment, I think PCT can provide a
very strong competing explanation of these observations. Indeed, I
didn't even see how Latash et al. explain their findings using a
model.

BA: What Latash and colleagues have been attempting is a detailed model of the first-order systems involved in controlling joint angle.

RM: If so, I certainly didn't pick that up from the paper. I didn't
see any physiological model, certainly not a model as detailed as the
physiological model described in B:CP.

BA: Bill provided such a model in B:CP based on what was known at the time about the anatomy and physiology involving control of a single muscle (such as a biceps), but "motor control" researchers apparently have encountered observations that this relatively straight-forward model does not account for.

RM: I think motor control researchers have "encountered observations"
that make them feel comfortable about their S-R approach to
understanding purposive behavior. I think one's model of behavior has
a big influence on what one "observes" about nervous system behavior.

RM: Well, my Psych Science paper didn't seem to do it. Maybe there is some other more compelling observations on which the equilibrium hypothesis is built. If you would let me know what they are I'll work on building a model that can handle them.

BA: Hey, it's not my job to "let you know what they are."

RM: OK, so it's just your job to say that there are observations that
PCT can't handle but I won't tell you what they are? Great job.

RM: But I'm willing to bet my ill-gotten PhD that _nothing_ will convince believers in the equilibrium hypothesis (and similar "dynamic attractor" models) that control theory is the solution to the problem of understanding purposeful behavior. Because I know that they are absolutely convinced that purposive behavior can be accounted for in lineal cause-effect terms.

BA: Well, you may be convinced of that, but I'm not.

RM: I know;-)

BA: What is "the equilibrium hypothesis" and how is it a "dynamic attractor" model? (Dynamic attractors are just descriptions of system behavior, they are not models.)

RM: I think that's a good distinction. Attractor models are
descriptive or what I would call "curve fitting" models. "Equilibrium
hypothesis" models are often working versions of "attractor" models;
the equilibrium point (defined by the relative tensions in a set of
muscles, for example) is the functional equivalent of an attractor.

Best

Rick
--
Richard S. Marken PhD
www.mindreadings.com

The only thing that will redeem mankind is cooperation.
                                                  -- Bertrand Russell

--
Richard S. Marken PhD
www.mindreadings.com

The only thing that will redeem mankind is cooperation.
                                                   -- Bertrand Russell

[From Bruce Abbott (2013.28.1745 EST)]

Warren Mansell:

WM: Hi, I am in line with Bruce here. In our discussions with them, Bill and I were more frustrated that they didn't acknowledge that PCT-like idea were already in their model, than thinking they were S-R. Bill was will to concede and learn from them concerning the equilibrium idea because he recognised that muscles are on paired groups across joints, but only if they could see the relevance of PCT to their overall functioning models, which they didn't, refusing to accept any connection between control theory and purposeful biological systems. I still think their physiological ideas could be helpful to us but within a HPCT model of the nervous system...
Warren

I've been doing some on-line reading of the literature on the equilibrium point (EP) model and can see why Latash et al. would not be aware of any connection of their model with PCT. In fact, I've found it difficult to understand their model, given the descriptions I've read of it. Thus far I have not seen a discussion of how the "commands" of the system are derived; it could very well be an open-loop system at that level, or maybe not. It's like presenting a PCT diagram of a first-order control system without indicating what determines the reference signal values -- they could be determined by the output of an open-loop or closed-loop system.

As for the first-order system that contracts the muscles, the lambda version of the EP model refers to a control signal ("command"), lambda, determines the equilibrium point of opposing muscles. The equilibrium point represents a condition in which the force being exerted by opposing muscles balances across the joint so that there is no net torque about the joint. It is important to understand that this "equilibrium point" takes account of the load (force) being exerted by a disturbance; thus the opposing muscles do not necessarily (or even usually) generate equal and opposite torques under static conditions.

The equilibrium point is shifted by changing lambda, a neural input to the first-order system that affects the neural threshold of the motor neurons. Lambda shifts these thresholds in opposite directions for the motor neurons innervating opposing muscles. These thresholds are also affected by input to the neurons from the muscle spindles.

Imagine that the joint in question is the elbow joint and that you are holding your forearm level as a result of establishing a certain value for lambda. The biceps contracts enough to resist the torque exerted by the force of gravity on your forearm, and the triceps is relaxed. I now place a weight in the palm of your hand. This increases the torque on the elbow and your forearm begins to drop. This movement stretches the biceps, resulting in a change in the output of the stretch receptors in the biceps muscle. This is signal is conveyed to motor neurons connected to the biceps, where it alters the activation thresholds of those neurons. (This signal and lambda add together at the motor neuron.) Consequently these motor neurons begin to fire at a higher rate and more motor neurons become active, increasing the contractile force of the biceps muscle. A new equilibrium point is reached in which the biceps is only slightly stretched because of the negative feedback relation between the stretch of the muscle and the contractile force the muscle generates. (The increased stretch generates an increase in opposing force, compensating (mostly) for the change in the disturbance level (extra weight in the palm).

In this situation the equilibrium point is a point along a curve. Where you are on the curve depends on load, etc. Changing lambda moves the entire curve left or right, creating a different set of equilibrium point values.

I don't guarantee that the above description is entirely accurate, because I may not have understood the model correctly.

Bruce

···

On 27 Dec 2013, at 17:42, Richard Marken <rsmarken@GMAIL.COM> wrote:

[From Rick Marken (2013.12.27.0940)]

Bruce Abbott (2013.12.27.0850 EST)--

BA: It's an active system because it actively alters λ, a system
parameter that affects the equilibrium point in their model

RM: I see that as just a motor output generation model with λ being the motor output that, as you said, sets the equilibrium point. It is not a reference for an input. This is not what I meant by "active", though, since this system is no different from any other "control of output" motor control model, all of which (unlike PCT) view behavior as a process of output generation. In this case, the output is the setting of the tensions in a group of muscle, which defines an equilibrium point, in the sense that if a transient force is applied to the limbs attached to the muscles the limbs will spring back to the equilibrium setting. That is not an active process from my point of view; it's just passive output generation: S-R.

BA: An equilibrium model wouldn't produce any movements unless you gave it an impulse disturbance, after which the limb in question would just settle back to its pre-disturbance position.

RM: Actually, it would produce movements resulting from the changes in
the setting of λ, the equilibrium point. Variations in λ will produce
very nice movements in a disturbance free environment. I found a nice
paper showing life like movements produced by variations in λ in a
human avatar. λ is kind of like a faux reference setting; if there are
no disturbances in the environment then variations in λ will look like
"controlled" movements. But put in disturbances and you will see that
controlled output models of purposeful behavior don't work.

BA: I don't believe that your paper deals with this level of physiological detail either. So neither your paper nor Bill's Little Man demo offer a competing explanation for the observations that Latash et al.'s model deals with.

RM: I see only behavioral data in the Latash et al model. And to the
extent that I understand their experiment, I think PCT can provide a
very strong competing explanation of these observations. Indeed, I
didn't even see how Latash et al. explain their findings using a
model.

BA: What Latash and colleagues have been attempting is a detailed model of the first-order systems involved in controlling joint angle.

RM: If so, I certainly didn't pick that up from the paper. I didn't
see any physiological model, certainly not a model as detailed as the
physiological model described in B:CP.

BA: Bill provided such a model in B:CP based on what was known at the time about the anatomy and physiology involving control of a single muscle (such as a biceps), but "motor control" researchers apparently have encountered observations that this relatively straight-forward model does not account for.

RM: I think motor control researchers have "encountered observations"
that make them feel comfortable about their S-R approach to
understanding purposive behavior. I think one's model of behavior has
a big influence on what one "observes" about nervous system behavior.

RM: Well, my Psych Science paper didn't seem to do it. Maybe there is some other more compelling observations on which the equilibrium hypothesis is built. If you would let me know what they are I'll work on building a model that can handle them.

BA: Hey, it's not my job to "let you know what they are."

RM: OK, so it's just your job to say that there are observations that
PCT can't handle but I won't tell you what they are? Great job.

RM: But I'm willing to bet my ill-gotten PhD that _nothing_ will convince believers in the equilibrium hypothesis (and similar "dynamic attractor" models) that control theory is the solution to the problem of understanding purposeful behavior. Because I know that they are absolutely convinced that purposive behavior can be accounted for in lineal cause-effect terms.

BA: Well, you may be convinced of that, but I'm not.

RM: I know;-)

BA: What is "the equilibrium hypothesis" and how is it a "dynamic
attractor" model? (Dynamic attractors are just descriptions of system
behavior, they are not models.)

RM: I think that's a good distinction. Attractor models are
descriptive or what I would call "curve fitting" models. "Equilibrium
hypothesis" models are often working versions of "attractor" models;
the equilibrium point (defined by the relative tensions in a set of
muscles, for example) is the functional equivalent of an attractor.

Best

Rick
--
Richard S. Marken PhD
www.mindreadings.com

The only thing that will redeem mankind is cooperation.
                                                  -- Bertrand Russell

-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2014.0.4259 / Virus Database: 3658/6956 - Release Date: 12/28/13

[From Rick Marken (2013.12.29.1600)]

Bruce Abbott (2013.28.1745 EST)–

BA: I’ve been doing some on-line reading of the literature on the equilibrium point (EP) model and can see why Latash et al. would not be aware of any connection of their model with PCT.

RM: It’s not that they are “not aware” of a connection; they explicitly reject a connection because they reject control theory.

BA: In fact, I’ve found it difficult to understand their model, given the descriptions I’ve read of it. Thus far I have not seen a discussion of how the “commands” of the system are derived; it could very well be an open-loop system at that level, or maybe not.

RM: I’m pretty sure it’s an open loop model. Their “commands” are not like reference signals because they not specifications for input. Rather, they are commands for output – they cause the tensions in the muscles (outputs) that determine the equilibrium point of the limbs to which the muscles are attached.

BA: It’s like presenting a PCT diagram of a first-order control system without indicating what determines the reference signal values – they could be determined by the output of an open-loop or closed-loop system.

RM: I think it’s like presenting a PCT diagram of a first order control system without the first order perception and comparator. In other words, it’s nothing like a PCT diagram; it’s a diagram of a motor output system and I can only assume that they persist in going with this approach because they want to explain control in lineal causal terms. The relationship of Latish et at to PCT is one of complete opposition.

BA: As for the first-order system that contracts the muscles, the lambda version of the EP model refers to a control signal (“command”), lambda, determines the equilibrium point of opposing muscles.

RM: Yes, and it determines the equilibrium point by sending signals to the muscles. But the theorists don’t say how the system knows which signals to send to the muscles since there is no feedback regarding whether the equilibrium point has been achieved by the tensions commanded; or even whether teh tensions commanded have been achieved (muscles don’t always respond to neural commands in exactly the same way; they fatigue, for example). I think their model is just a bunch of hand-waving, S-R BS. The fact that you and Martin (and apparently others) see this model as “similar” to PCT or that it explains something that PCT doesn’t is just astounding to me (and not a little depressing).

BA: The equilibrium point represents a condition in which the force being exerted by opposing muscles balances across the joint so that there is no net torque about the joint. It is important to understand that this “equilibrium point” takes account of the load (force) being exerted by a disturbance; thus the opposing muscles do not necessarily (or even usually) generate equal and opposite torques under static conditions.

RM: Oh, really? And how in the world does it “take account” of the load being exerted by a disturbance? Because Latash et al say so? PCT takes it into account by controlling a perception of the “equilibrium point” (defined by the tensions of the muscles on the limbs) and varies the “commands” sent to the muscles, as necessary, to maintain that perception in the face of varying loads on the limbs. The equations of closed-loop control explain how it does this.

BA: The equilibrium point is shifted by changing lambda, a neural input to the first-order system that affects the neural threshold of the motor neurons. Lambda shifts these thresholds in opposite directions for the motor neurons innervating opposing muscles. These thresholds are also affected by input to the neurons from the muscle spindles.

RM: If these inputs are compared to references for what they should be then you’ve got the equivalent to a first order intensity control system and equilibrium theory is precisely identical to control theory. I’d have to see a diagram of their model to see if this is the case. Since they rather rudely rejected Bill’s overtures to discuss the control theory approach to limb control I doubt very much that there is feedback control involved in their equilibrium theory.

BA: Imagine that the joint in question is the elbow joint and that you are holding your forearm level as a result of establishing a certain value for lambda. The biceps contracts enough to resist the torque exerted by the force of gravity on your forearm, and the triceps is relaxed. I now place a weight in the palm of your hand. This increases the torque on the elbow and your forearm begins to drop. This movement stretches the biceps, resulting in a change in the output of the stretch receptors in the biceps muscle. This is signal is conveyed to motor neurons connected to the biceps, where it alters the activation thresholds of those neurons.

RM: You are describing an active control process. The stretch receptor output is the controlled perceptual signal. This signal is “conveyed to motor neurons” via a comparator resulting in an error signal that is the motor neuron signal to the biceps, resulting in contraction of the biceps that brings the stretch signal back to the reference. This is control theory, not equilibrium theory.

BA: (This signal [the stretch signal] and lambda add together at the motor neuron.)

RM: That would produce positive feedback; the perceptual signal should be subtracted from the reference in order to get negative feedback. But if that’s really the equilibrium model then it’s a control model that doesn’t control, which would distinguish it from PCT, but in a different way than if it were a lineal causal model (which it is).

BA: Consequently these motor neurons begin to fire at a higher rate and more motor neurons become active, increasing the contractile force of the biceps muscle. A new equilibrium point is reached in which the biceps is only slightly stretched because of the negative feedback relation between the stretch of the muscle and the contractile force the muscle generates. (The increased stretch generates an increase in opposing force, compensating (mostly) for the change in the disturbance level (extra weight in the palm).

RM: Yes, you have described the PCT model of control of limb position. I believe the equilibrium model is quite different, at least, as I have read about it on the net). According to what I have read, equilibrium theory treats the muscles as a set of rubber bands attached to the limbs. The lambda “command” signal specifies an equilibrium point in terms of motor commands that specify the appropriate tensions in these rubber bands. If a disturbance is applied (such a a ball dropping on the hand) the tension in the rubber bands (and, hence, the location of the equilibrium point) changes but when the disturbance is removed the bands (and equilibrium point) return to their original tension configuration (equilibrium point). Unlike the control system approach, the equilibrium model cannot maintain the equilibrium point in the face of continuous disturbance.

BA: In this situation the equilibrium point is a point along a curve. Where you are on the curve depends on load, etc. Changing lambda moves the entire curve left or right, creating a different set of equilibrium point values.

RM: Correct; lambda changes the equilibrium point open loop. The equilibrium point is “defended” against transient disturbances in the same way that the ball in the bowl is defended from transient disturbances; once the disturbance is removed, the elastic forces of the muscles return the limb to the equilibrium point just as gravity returns the ball to the bottom of the bowl.

BA: I don’t guarantee that the above description is entirely accurate, because I may not have understood the model correctly.

RM: I don’t know either. It’s hard to find a nice, simple description of the model, in diagram form, mapped to a limb, for example. But since you and Martin seem to think that PCT has to be able to explain the data that you say only the equilibrium theory model can handle, how about explaining what Figure 1 in the Latash et al paper is all about.

Best

Rick

···

Bruce

On 27 Dec 2013, at 17:42, Richard Marken rsmarken@GMAIL.COM wrote:

[From Rick Marken (2013.12.27.0940)]

Bruce Abbott (2013.12.27.0850 EST)–

BA: It’s an active system because it actively alters ë, a system
parameter that affects the equilibrium point in their model

RM: I see that as just a motor output generation model with ë being the motor output that, as you said, sets the equilibrium point. It is not a reference for an input. This is not what I meant by “active”, though, since this system is no different from any other “control of output” motor control model, all of which (unlike PCT) view behavior as a process of output generation. In this case, the output is the setting of the tensions in a group of muscle, which defines an equilibrium point, in the sense that if a transient force is applied to the limbs attached to the muscles the limbs will spring back to the equilibrium setting. That is not an active process from my point of view; it’s just passive output generation: S-R.

BA: An equilibrium model wouldn’t produce any movements unless you gave it an impulse disturbance, after which the limb in question would just settle back to its pre-disturbance position.

RM: Actually, it would produce movements resulting from the changes in
the setting of ë, the equilibrium point. Variations in ë will produce
very nice movements in a disturbance free environment. I found a nice

paper showing life like movements produced by variations in ë in a
human avatar. ë is kind of like a faux reference setting; if there are
no disturbances in the environment then variations in ë will look like

“controlled” movements. But put in disturbances and you will see that
controlled output models of purposeful behavior don’t work.

BA: I don’t believe that your paper deals with this level of physiological detail either. So neither your paper nor Bill’s Little Man demo offer a competing explanation for the observations that Latash et al.'s model deals with.

RM: I see only behavioral data in the Latash et al model. And to the
extent that I understand their experiment, I think PCT can provide a
very strong competing explanation of these observations. Indeed, I

didn’t even see how Latash et al. explain their findings using a
model.

BA: What Latash and colleagues have been attempting is a detailed model of the first-order systems involved in controlling joint angle.

RM: If so, I certainly didn’t pick that up from the paper. I didn’t
see any physiological model, certainly not a model as detailed as the
physiological model described in B:CP.

BA: Bill provided such a model in B:CP based on what was known at the time about the anatomy and physiology involving control of a single muscle (such as a biceps), but “motor control” researchers apparently have encountered observations that this relatively straight-forward model does not account for.

RM: I think motor control researchers have “encountered observations”
that make them feel comfortable about their S-R approach to
understanding purposive behavior. I think one’s model of behavior has

a big influence on what one “observes” about nervous system behavior.

RM: Well, my Psych Science paper didn’t seem to do it. Maybe there is some other more compelling observations on which the equilibrium hypothesis is built. If you would let me know what they are I’ll work on building a model that can handle them.

BA: Hey, it’s not my job to “let you know what they are.”

RM: OK, so it’s just your job to say that there are observations that
PCT can’t handle but I won’t tell you what they are? Great job.

RM: But I’m willing to bet my ill-gotten PhD that nothing will convince believers in the equilibrium hypothesis (and similar “dynamic attractor” models) that control theory is the solution to the problem of understanding purposeful behavior. Because I know that they are absolutely convinced that purposive behavior can be accounted for in lineal cause-effect terms.

BA: Well, you may be convinced of that, but I’m not.

RM: I know;-)

BA: What is “the equilibrium hypothesis” and how is it a "dynamic

attractor" model? (Dynamic attractors are just descriptions of system
behavior, they are not models.)

RM: I think that’s a good distinction. Attractor models are

descriptive or what I would call “curve fitting” models. “Equilibrium
hypothesis” models are often working versions of “attractor” models;
the equilibrium point (defined by the relative tensions in a set of

muscles, for example) is the functional equivalent of an attractor.

Best

Rick

Richard S. Marken PhD
www.mindreadings.com

The only thing that will redeem mankind is cooperation.
– Bertrand Russell


No virus found in this message.
Checked by AVG - www.avg.com

Version: 2014.0.4259 / Virus Database: 3658/6956 - Release Date: 12/28/13


Richard S. Marken PhD
www.mindreadings.com

The only thing that will redeem mankind is cooperation.

                                               -- Bertrand Russell