B:CP impressions - 1st cut

CSGnet Archive CSG1994
1

[Paul George 940720 17:00]

Having read through the first 8 chapters, I begin to see the source of the
misunderstanding of my posts, as most discussion seem to focus on the first
three levels (orders) of the CNS.PCT does agree with my mental model, there
were just some scope and terminology issues. All in all a great piece of work.

I guess it's lack of acceptance is due to perceptual filtering, or more likely
that 'other schools' never actually _read_ (with understanding) the book. I
found Bill Power's post on the e. coli paper review commitee revealing as to
the nature of 'peer review'. However you should be careful that you don't
follow the same pattern in judging other's work. It is important to evaluate a
piece of work using the _author's_ frame of reference, not your own.
Translation is your job when you are not using terminology in a 'standard' way
or when using 'non-standard' concepts.

May I suggest that the 'FAQ'/intro add a summary of the nature of neural
currents, the types of neural circuits, and the distinction between first order
and higher order systems. Since B:CP is relatively hard to find (lacking money
to order it from the Powers), this would avoid a lot of confusion. {I had to
get my copy via inter-library loan from a Cincinnati public library (I live in
Cleveland). It apparently wasn't available from the universities in the area,
of which there are many)

The essence of PCT is the functioning of organisms (a normal focus of
psychology as opposed to sociology). It can be summarized thusly:
All action and sensation is produced via the interaction of neurons and
muscles. Further, the central nervous system is composed of neurons. It
therefor follows that any perception or behavior, regardless of complexity,
must be producible via the interaction of neurons through neural currents (in
the absence of some other mechanism - 'a ghost in the machine'). We can
demonstrate that negative feedback control is the mechanism used for first
order interaction with the environment, and at the level of spinal reflex. We
can use models to show it can work for higher order behavior. Thus, by default
we presume it is the mechanism used at all levels.

When I talk about a system or control system I am usually looking at more than
one entity. A biological analogy to a process control system would be a
supervisor monitoring 2 operators, who are each operating a set of tools
involved in a manufacturing process. This is similar to HPCT, except that
neural currents are not the _only_ mechanism for interaction of the 'nodes'. I
don't think that this really changes anything, except by adding 1st level input
and output function errors and adding the possibility of non pulse coded
signals.

Please bear with me on the following as I am using a 20 year old work. There
may have been elaborations in the mean time, but I presume there would be a
'2nd edition' if there were major changes.

An observation: error signals don't theoretically have to only have positive
values from an information standpoint. You can cheat by 'biasing' the signal.
If the signal may vary from 0-30 ppm, I can set the comparitor and output
function so that 15 = no error (a logical 0). I could then set an 'upper
threshold' at 25 and a lower threshold at 5. The output function could use this
information to 'select' the proper action or magnitude of action. This could
allow a stepped sawtooth output function rather than a continuous one. I'm not
saying that this _ever_ happens in biology, or that it can't be implemented
with a network of simple nodes, just that it would work. This might be useful
for designing automata using PCT.

Question: why is it illegal to pass an error signal (directly or through a
'repeater' output function) to another node as a 'sensation'? While the analogy
may not bear close examination, pain might be something of this type. The
output is just passed up to tell a higher level controller that something is
wrong, and how badly. It can be used to indicate that 'control' is not working
and another strategy must be applied. It is up to the higher level to set
reference levels elsewhere in the nnetwork so that the error signal is
mitigated. It's not very useful at the lowest couple of orders, but might be
useful at the 'cognitive' levels of control. This also might correspond to
'alerting'. It is just another input signal, but generated by a comparator
rather than a lower order input function. Again, I don't assert that it
actually exists in nervous systems, or that it is necessary. OTOH in
engineering we sometimes find that a more complex structure is more efficient
than the equivalent constructed from simple structures. If nothing else you
have propagation and processing lags. Biology has a limitation in that it
developed by 'growing like topsy'. New levels and nodes were added to existing
ones that worked.

As your thinking has evolved over 20 years, do you have any problem with having
'neural logic' as a part of input, output, or comparator functions? Is there
any theoretical problem with a single node (or subsystem) being at different
levels of the hierarchy with respect to other nodes or subsystems {Higher
apparently means 'sets another's reference level'} at the same time? My mental
model can envision a network rather than a true hierarchy

On another minor matter, I am not sure that nodes interact only through neural
currents in biological systems. Some output functions result in the release of
hormones, neurochemicals or substances like adrenaline (blanking on the name).
These are certainly sensed by other nodes, and not necessarily hierarchically
(i.e 4th level affecting 4th or other level). I guess this could be viewed as
an 'envronmental' interaction of physical laws, but seems to me more like a 2nd
or higher order interaction. I'm not sure it really matters if a signal is
transmitted in terms of neural current frequency or in terms of a chemical
concentration which must be translated via an input or reference signal
transducer function. Again, it may be just where you draw the system boundary.

Re Tom Bourbon [940719.1202]

I don't think creating a taxomy of types of signals or types of control nodes
is silly in and of itself, any more than distinguishing between orders of
control systems or sections of the brain. Yea it is all 'just control' or 'just
neural currents', but the distinctions are sometimes useful. We wouldn't get
very far in medicine or physiology if we fixated on the idea that 'cells are
just cells' and ignored their differentiations and groupings (e.g. organs). We
may be able (and have) to define standard structures or patterns used for
various 'types' of perception and control. The question is which groupings or
distinctions make sense.

Enough digital diarrhea for today
Paul

2

[Paul George 940721 11:00]

[Bill Leach 940720.23:22 EST(EDT)]

Me:

It can be summarized thusly: All action and sensation is produced via
the interaction of neurons ...
            ...
Thus, by default we presume it is the mechanism used at all levels.

I think "them's fighten' words" :slight_smile: PCT is about the idea that Behaviour
is the control of perception.

I _knew_ I was going to get zapped for that :-}.
To me there are 2 significant factors in PCT, while y'all seem to focus on one.
The first is that you can demonstrate that behavior is a hierarchical control
process and that a neural mechanism can produce it. The second is that
perception is being controlled rather than the environment or actions upon the
environment. I took the latter en passant, as should anyone with any
familiarity with epistimology.

However note that even control engineers do not view that which is being
corrected as that which is being controlled. From _their point of view_ the
controller is controling the process, i.e. trying to control the environment.
Of course the only way we know what is going on in the environment or with our
actions (which you normally consider part of it), is through our senses which
are abstracted up into perceptions through a hierarchy (or series, or network)
of input functions. Engineers would grant that the control loop and hierarchy
functions as described in PCT.

I grant that this control distinction is a major sticking point between you and
other psychologists, but I am not sure that the first point is not the most
significant in terms of being able to analyze human and social behavior. The
second point seems to be more one of terminology or point of view than of real
overriding significance. But then I'm not a psychologist or behavioral
researcher :-).

When I talk about a system or control system I am usually looking at
more than one entity.

PCT does not have a problem in dealing with multiple control systems.

I meant the same thing as you did in your post on process control. We tend to
draw the boundary of 'a system' in a different place that PCT'rs usually do.
The theory of course has no problem with distribution of control nodes or
systems, other than the mechanism of communication would possibly vary from the
physiological model.

I should like an example of this. This sounds a bit like the arguement
that it is impossible to write a program in assembler that is as
efficient and effective as one written using an optimizing high level
compiler.

I intended it as more the other way around. It is more efficient to use high
level graphic design systems than to hack code, which is much more efficient
than bit twiddling. (this is after all my field of expertise). I am reminded of
a 'programmer' on one project who always hard coded incrementing loop counters,
because he had never heard of a fortran 'do loop' (and never thought to look in
a manual). I know there are examples in electronics of more sophisticated
circuits replacing sets of simpler ones, but can't think of a specific on the
spur of the moment.

What do you mean by this? Higher level engineered control structures
don't have propagation and processing lags?

No, but intra-node communication can be faster than inter-node, due to physical
distance if nothing else. An IC based CPU is a might faster (and more reliable)
than the equivalent made of tubes or transistors. As a team grows in size the
increased communication and coordination overhead makes the increase in
capacity diminish on the order of np^2 (a.k.a an NP complete problem).
Centralizing computation or logic minimizes communication at the expense of
complexity. All I am saying is that in a given situation a 'logic chain' and
single control loop may be more efficient than a hierarchy of control loops.

3

<[Bill Leach 940721.20:16 EST(EDT)]

[Paul George 940721 11:00]

I _knew_ I was going to get zapped for that :-}.

Then why d'ja do it? Huh, Huh? :slight_smile:

To me there are 2 significant factors in PCT, while y'all seem to focus
on one. The first is that you can demonstrate that behavior is a
hierarchical control process and that a neural mechanism can produce it.
The second is that perception is being controlled rather than the
environment or actions upon the environment. I took the latter en
passant, as should anyone with any familiarity with epistimology.

There are really three that you have listed. The necessity for
demonstrating that the neural structure for control exists is really
"someone elses bag." OTOH, if what you mean by "and that a neural
mechanism can produce it." is that "we" can show that biological beings
exhibit control system behaviour then that is another matter altogether.

The real question there is "Can anyone show any other method of operation
for living entities?" PCT can an does demonstrate that it can model to a
high degree of accuracy the operation of living systems in limited
sphears. What is important here with the phrase "limited" is that the
evidence points to the limited ability to model rather than limits upon a
particular model match (human in this case) behaviour.

At risk of being slapped into line by Bill P., I am going to propose some
overall "driving concepts" for PCT:

1. The macro view of the activities of mankind over as long a history as
    detailed information is available clearly demonstrate "closed loop
    negative feedback control system operation".

    This single observation, I gather, was one of the overwhelming
    influences upon Bill Powers concerning the possible "nature of man."
    You can probably add the scientist/engineers' disdain for the "mushy"
    opinion based existing behavioural "sciences".

2. Actual attempts at modeling aspects of human behaviour were found to
    correllate well with real human behaviour in the conduct of simple
    experiments. (and if I remember the story correctly, an error in
    prediction made by those discussing PCT was discovered but BOTH the
    human and the model behaved in the same fashion -- thus
    interpretation of the significance of control theory was what was in
    error, not the theory itself).

3. There seems to be physical evidence of the existance of actual
    control systems in living systems (including, of course, people).
    The discovery of actual control system loops is "encouraging" but is
    not central to the theory. If the biologists had not found such
    structures, the rest of the behavioural "sciences" would still have
    the problem of explaining closed loop control system behaviour some
    other way and none has been demonstrated to work.

I grant that this control distinction is a major sticking point between
you and other psychologists, but I am not sure that the first point is
not the most significant in terms of being able to analyze human and
social behavior. The second point seems to be more one of terminology or
point of view than of real overriding significance. But then I'm not a
psychologist or behavioral researcher :-).

If you reword that first line to "... is THE major ...", I believe that
you will have phrased it correctly.

As I tried to indicate above, the physical evidence of structure is not
at issue and would not be even if it appeared to be conflicting. If the
biological evidence indicated something other than closed loop control
systems, their would still be the problem of explaining how something
that is not a closed loop control system behaves as though it is one.

PCT does not have a problem in dealing with multiple control systems.

I meant the same thing as you did in your post on process control. We
tend to draw the boundary of 'a system' in a different place that PCT'rs
usually do. The theory of course has no problem with distribution of
control nodes or systems, other than the mechanism of communication would
possibly vary from the physiological model.

Indeed, if I have learned anything from Tom Bourbon at all, it is that
"social aspects" of human behaviour are explainable by analyzing the
dymanics of control systems operating in a common environment.

I should like an example of this. This sounds a bit like the arguement
that it is impossible to write a program in assembler that is as
efficient and effective as one written using an optimizing high level
compiler.

I intended it as more the other way around. It is more efficient to use
high level graphic design systems than to hack code, which is much more
efficient than bit twiddling. (this is after all my field of expertise).
I am reminded of a 'programmer' on one project who always hard coded
incrementing loop counters, because he had never heard of a fortran 'do
loop' (and never thought to look in a manual). I know there are examples
in electronics of more sophisticated circuits replacing sets of simpler
ones, but can't think of a specific on the spur of the moment.

It is (generally) a more efficient use of resources including time, to
program in a high level language. However, in anything but a trivial
program it is always possible that assembly code could be written that
would be more effecient as a program than that generated by a compiler.
That programs have reached levels of complexity that would make it not
only impractical to make the attempt but would be physically impossible
for even a medium sized programming team to do so.

The statement is true because of the very nature of programming and logic
processing. Indeed, high level languages are nothing more than programs
written to convert tokens and symbols to binary code but such conversions
are somewhat general. You might want to take a look at some of the
comparisons that have been made between code size and function as a
function of the method of generation.

In terms of efficiency, computers have moved away from efficiency at a
remarkable rate... it is just that processing power has more than kept
pace (most of the time).

I am "overly belaboring" the point here except that I think that it is
worthwile to consider that what we often perceive as an improvement in
engineered control systems is in fact a system with vastly greater
processing power that is only able to make marginal improvements in
control system capability. The "real" improvement is in the change to
the environment that the DESIGNER is working in. People do things in
control system design today as a matter of course that would not have
been much more than dreamed of even as recently as 10 years ago. The
reason that they would not have done these things was not because they
were impossible but rather because they were impractical (I will admit,
of course, that there are things that are done today that were, at least
perceived, to be impossible at one time).

What do you mean by this? Higher level engineered control structures
don't have propagation and processing lags?

No, but intra-node communication can be faster than inter-node, due to
physical distance if nothing else.

For the technology in actual use. Indeed, we have deployed communcations
systems whose throughput would "crush" the sort of controllers that you
are talking about but in general it is not practical to use such
communications systems for your application.

An IC based CPU is a might faster (and more reliable) than the
equivalent made of tubes or transistors.

I'm not sure of the significance of this statement... is Bailey still
using tubes? :slight_smile:

As a team grows in size the increased communication and coordination
overhead makes the increase in capacity diminish on the order of np^2
(a.k.a an NP complete problem). Centralizing computation or logic
minimizes communication at the expense of complexity. All I am saying is
that in a given situation a 'logic chain' and single control loop may be
more efficient than a hierarchy of control loops.

This is true but its' relevence is dependent upon how control is
structured. A biological control hierarchy seems to be an example in the
extreme for what you are saying is NOT a good control methodology. If
you think you have ever seen a "large" distributed control system design,
just ponder for a moment the sheer number of "controllers" that a human
might have.

The "trick" there is that each level also "removes" raw detail in a
perceptual signal so that each "neural signal" in successively higher
levels represents a greater amount of information. Likewise, each high
level reference "decomposes" into many lower level references on the way
down. It is sort of like the idea that a node in a binary tree
represents all of the nodes below it (in its' chain). The datum is (at
least close) to the same size and duration as datum "at the skin". Thus,
no NP problem).

-bill

4

[Paul George 940722 10:20]

[Bill Leach 940721.20:16 EST(EDT)]

OTOH, if what you mean by "and that a neural
mechanism can produce it." is that "we" can show that biological beings
exhibit control system behaviour then that is another matter altogether.

Kind of both, if the B:CP discussion of spinal reflexes is correct (which to my
level of neurological knowledge it is). The latter point is what I meant by:

The first is that you can demonstrate that behavior is a hierarchical control
process...

If you reword that first line to "... is THE major ...", I believe that
you will have phrased it correctly.

:slight_smile: Actually, I did phrase it that way the first time and later toned it down
out of politeness.

As I tried to indicate above, the physical evidence of structure is not
at issue and would not be even if it appeared to be conflicting. If the
iological evidence indicated something other than closed loop control
systems, their would still be the problem of explaining how something
that is not a closed loop control system behaves as though it is one.

Want to run that one by me again? You are saying that if I could prove that the
CNS anatomically did not use closed loop control structures that the burden of
proof would be on me to show how and why it worked??
that closed loop control is the mechanism for organismic behavior. It would
still be useful for analyzing or predicting it. Two different mathematical
functions that give the same results within a given range are equivalent within
that range.

A biological control hierarchy seems to be an example in the
extreme for what you are saying is NOT a good control methodology........
It is sort of like the idea that a node in a binary tree
represents all of the nodes below it (in its' chain). The datum is (at
least close) to the same size and duration as datum "at the skin". Thus,
no NP problem).

I'm not saying it is bad, just not always the most efficient or simple (in
terms of number of components). The situation I am talking about is where there
are a large number of nodes on the same level which must interact in order to
'evaluate' (not necessarily control) enough perceptions to 'make decisions',
'recognize patterns', or make plans. Computers use memory and logic structures
other than trees (particularly in AI) for a reason. The levels 1-3 of a
biological system are clearly hierarchical anatomically . Hierarchical
structures to filter, summarize, or compress information are common. However,
the hierarchy is often in terms of 'strata' rather than levels. As Bill B
indicated in his response to my post:

[From Bill Powers (940721.0815 MDT)]
I keep thinking of cases where what I call a third-level control systems seems
to operate via a fifth-level control system, and so on. So far I've been able
to resolve most of these possibilities, but the question is still open
whether we should think of these "levels" as "dimensions" instead. ... I
opted for the hierarchy, but have wondered ever since whether there
isn't something to the other view, too.

At some point the 'cannonical model' will likely break down due to the
explosion of the number of nodes required. It takes a lot of control loops to
control something like writing this post. When you get to the level of the
sensory processors and upper cortex functions such as language,I strongly
suspect that newtwork rather than hierarchical structures predominate in order
to generate the constructs we usually term 'perceptions' or 'ideas'. There will
be more 'horizontal' connections than 'vertical'. I also suspect that control
functions will involve more 'algorithmic' processing. However, I am not sure
how you could tell empirically, though the stuff being done with PET scanners
might help.

Paul

···

From my point of view it would simply show PCT to be incorrect in asserting

5

<[Bill Leach 940723.16:23 EST(EDT)]

[Paul George 940722 10:20]

As I tried to indicate above, the physical evidence of structure is not

Want to run that one by me again? You are saying that if I could prove
that the CNS anatomically did not use closed loop control structures
that the burden of proof would be on me to show how and why it worked??

First, let me say that this was my opinion and not necessarily shared by
anyone else here. In particular, I know that Bill P. HAS studied anotomy
to try to look for evidence in either direction and I'm sure that he
cares very much what sort of new evidence shows up.

In a sense yes. The issue with PCT as I understand it IS NOT the
representation of the structure, the possible levels and hierarchy. The
issue is closed loop negative feedback control system is the only known
mode of operation of living systems that explains what is actually seen.

That PCT explains S-R behaviour is not even the point though the fact
that it explains exceptions to S-R behaviour is getting closer.

If you "come up with proof" that a person's neurological struction CAN
NOT be a control system, then yes you indeed have a great burden of proof
on your hands -- how does a non-control system consistently exhibit
control system behaviour?

From my point of view it would simply show PCT to be incorrect in
asserting that closed loop control is the mechanism for organismic
behavior. It would still be useful for analyzing or predicting it. Two
different mathematical functions that give the same results within a
given range are equivalent within that range.

I don't challenge the last sentence of your statement there at all. Show
me the function that provides the equivalent results. Short of
metaphysics there is no known phenomenon that can make a non-control
system behave exactly like a control system. If you do find such a
phenomenon maybe the second sentence would still be true.

I'm not saying it is bad, just not always the most efficient or simple
(in terms of number of components).

I agree and will even add that when enough is known about the limits of
environmental disturbances possible, even the choice of controlled
perceptions may be poor.

At some point the 'cannonical model' will likely break down due to the
explosion of the number of nodes required.

No, I don't think so. Even the discussions and models currently existing
in the PCT world do major consolidation of control loops and even
hierarchy. Where such "would break down" is in any attempt to exactly
model the structure of the living being as opposed to modeling its
behaviour for specific instances of control.

-bill

6

[Paul George 940722 10:20]

[Bill Leach 940721.20:16 EST(EDT)]

. . .
Bill L:

A biological control hierarchy seems to be an example in the
extreme for what you are saying is NOT a good control methodology........
It is sort of like the idea that a node in a binary tree
represents all of the nodes below it (in its' chain). The datum is (at
least close) to the same size and duration as datum "at the skin". Thus,
no NP problem).

Paul:

I'm not saying it is bad, just not always the most efficient or simple (in
terms of number of components). The situation I am talking about is where there
are a large number of nodes on the same level which must interact in order to
'evaluate' (not necessarily control) enough perceptions to 'make decisions',
'recognize patterns', or make plans. Computers use memory and logic structures
other than trees (particularly in AI) for a reason. The levels 1-3 of a
biological system are clearly hierarchical anatomically . Hierarchical
structures to filter, summarize, or compress information are common. However,
the hierarchy is often in terms of 'strata' rather than levels. As Bill B
indicated in his response to my post:

[From Bill Powers (940721.0815 MDT)]
I keep thinking of cases where what I call a third-level control systems seems
to operate via a fifth-level control system, and so on. So far I've been able
to resolve most of these possibilities, but the question is still open
whether we should think of these "levels" as "dimensions" instead. ... I
opted for the hierarchy, but have wondered ever since whether there
isn't something to the other view, too.

Paul:

At some point the 'cannonical model' will likely break down due to the
explosion of the number of nodes required. It takes a lot of control loops to
control something like writing this post. When you get to the level of the
sensory processors and upper cortex functions such as language,I strongly
suspect that newtwork rather than hierarchical structures predominate in order
to generate the constructs we usually term 'perceptions' or 'ideas'. There will
be more 'horizontal' connections than 'vertical'. I also suspect that control
functions will involve more 'algorithmic' processing. However, I am not sure
how you could tell empirically, though the stuff being done with PET scanners
might help.

Paul, I don't know for certain _how_ the questions raised in this exchange
will be resolved in the future, but I will not be putting my money on PET
studies (positron-emission-tomography studies of "higher," "cognitive"
functions). The widely held and promulgated idea that PET allows us to
"image the mind" (the brave new "neuro-cognitive" mind) is another sorry
feature on the bleak landscape of contemporary
behavioral-cognitive-neuro science. Another item behind my claim that those
sciences are often dangerous to the species.

Later,

Tom

···

In Message Fri, 22 Jul 1994 10:28:06 EDT, Paul George <pgeorge@BAILEY.COM> writes: