B:CP Course Summary of CH. 6 A Hierarchy of Control Systems

CSGnet Archive CSG2013
1

[From Rick Marken (2013.08.18.1130)]

Summary, Ch 6 Hierarchy of Control.doc (32.5 KB)

···

Thanks to Rupert Young for being the only student in class who raised his hand and gave his answers to the leading questions. I am attaching the summary of CH 6 which contains my answers to those questions and I think you’ll see that I basically agree with Rupert’s.

I hope there are people still listening in on the course. I think the lack of participation may be because everyone is on Facebook or Twitter (whatever that is) or because it’s Summer and they are sipping Vodka tonics beside their pools. But David and I will soldier on, hoping to spark some discussion. The study guide for the next chapter (7) will be posted soon.

Any suggestions about things we might to to get more involvement in the course – or regarding the content of the course – would be most welcome.

Best regards

Rick

–
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

2

[John Kirkland 20130819]

It sure is helpful having the little summaries. BTW Rick for some of us it’s now winter/spring. I know it’s hard to accept New Year festivities may include swimming at the beach or river.

My niggle about the course thus far is it seems to have adopted a pre-packaged GLM approach: everybody’s to read the set text and answer fixed questions. I was anticipating a more radical approach by way of illustrating PCT in action. Some time back I had raised the question as to how a PCT pedagogy might differ from what I’ll call the bulk default standard. In that traditional S-R dominant format what tended to happen when a reader n(student) got stuck or asked a question was they were ignored and/or got blamed for wasting everyone’s time, and/or they were classified as stupid and summarily dismissed. It still happens.

As an aside, thanks Martin for helping elucidating some matters I was pondering over, not made any easier with my cryptic style. Oh, yes, it was the ‘Royal we’ as you’d have fathomed.

For what it’s worth here are a few examples of some difficulties I encountered when reading Chapter 6.

pg 70 end of para #1, I’m unsure what a ‘control organisation’ is in this context. For instance is this a set of non-closed boxes as at the top of Fig 6.1 as mentioned in para 2, pg 72. Or is it the whole kit and caboodle?

pg 70 a three-level model is illustrated in Fig 6.1. It took me a good hour to label each box/arrow of Fig 6.1 and relate these tags to the text. So far I understand it each level is characterised by three boxes (input function - comparator - output function); yes? So, what’s the relationship between a level and a step (as in three steps of visual information processing)? Does each of these steps aim at and point to a different level? ‘The signal representing the actual relationship comes in from the left’ – where exactly is that again as I read ‘perceived’ in several places but not ‘actual signal’?

Can a high-level system operate as a low-level one, be passed down the line? For instance with a proficient sport-player’s performance, or those of a concert pianist? I quite liked the Australian study when drivers returning from a ski field were breath-analysed for illicit substances. Even at the bottom of a twisty, winding road they were accident free yet many well over legal limits. On a familiar road they could perform almost automatically. But when encountering an unexpected obstacle on the road (a cone for instance) they could not execute evasive actions successfully.

pg72 line 2 has the word ‘First’ but I cannot find a ‘second’ in the subsequent text. It’s probably staring me in the face…

Pg 75 para 2 ‘… yet the feedback loop for all systems passes through the environment of the whole collection’. Is this like upper-levels of a house of cards built upon those bottom ones resting on the tabletop which is where contact occurs? Is this ‘radical reductionism’, the golden bullet, the gravity of living organisms?

Pg 78 para 2, I’d be interested in hearing how and by whom the proposed model been modified over the past 40 years.

With kind regards

JohnK

···

On Mon, Aug 19, 2013 at 6:34 AM, Richard Marken rsmarken@gmail.com wrote:

[From Rick Marken (2013.08.18.1130)]

Thanks to Rupert Young for being the only student in class who raised his hand and gave his answers to the leading questions. I am attaching the summary of CH 6 which contains my answers to those questions and I think you’ll see that I basically agree with Rupert’s.

I hope there are people still listening in on the course. I think the lack of participation may be because everyone is on Facebook or Twitter (whatever that is) or because it’s Summer and they are sipping Vodka tonics beside their pools. But David and I will soldier on, hoping to spark some discussion. The study guide for the next chapter (7) will be posted soon.

Any suggestions about things we might to to get more involvement in the course – or regarding the content of the course – would be most welcome.

Best regards

Rick

–
Richard S. Marken PhD

rsmarken@gmail.com
www.mindreadings.com

3

[Martin Taylor 2013.08.18.23.06]

[John Kirkland 20130819]

...
For what it's worth here are a few examples of some difficulties I encountered when reading Chapter 6.

pg 70 a three-level model is illustrated in Fig 6.1. It took me a good hour to label each box/arrow of Fig 6.1 and relate these tags to the text. So far I understand it each level is characterised by three boxes (input function - comparator - output function); yes? So, what's the relationship between a level and a step (as in three steps of visual information processing)? Does each of these steps aim at and point to a different level?

Shortly, the answer to your question is "Not necessarily".

This is something that is seldom if ever mentioned on CSGnet. I discussed it face to face with Bill at CSG93, asking him specifically whether there could be several stages (your steps) within a level. Put briefly, Bill's answer was effectively: "I don't see why not". In other words, the model does not require that the upgoing perceptual signal that is output from a perceptual function at one level needs to serve as input to the different kind of perceptual function that defines the next higher level. Relationships of relationships or sequences of sequences, for example, could be perfectly accommodated within the HPCT structure. We tend to slide over that when discussing the hierarchic structure, and talk as though there could be only one stage at each level, which could easily lead people to assume that "one stage per level" is actually part of the definition of HPCT.

Martin

···

On 2013/08/18 6:39 PM, John Kirkland wrote:

4

[From Rick Marken (2013.08.19.2210)]

···

[John Kirkland 20130819]

JK: It sure is helpful having the little summaries.

RM: Thanks John. Glad they help.

JK: BTW Rick for some of us it’s now winter/spring. I know it’s hard to accept New Year festivities may include swimming at the beach or river.

RM: Ah, you’re in that hemisphere. G’day.

JK: My niggle about the course thus far is it seems to have adopted a pre-packaged GLM approach: everybody’s to read the set text and answer fixed questions. I was anticipating a more radical approach by way of illustrating PCT in action. Some time back I had raised the question as to how a PCT pedagogy might differ from what I’ll call the bulk default standard. In that traditional S-R dominant format what tended to happen when a reader n(student) got stuck or asked a question was they were ignored and/or got blamed for wasting everyone’s time, and/or they were classified as stupid and summarily dismissed. It still happens.

RM: I’m open to suggestions. But I did want to go trough the book chapter by chapter; the class is an homage to Powers’ main opus, which should sit up on the shelf with the Principia and Origin of Species, if books are still around (to say nothing of coastal cities) in 100 years;-) I was hoping there would be more back and forth about each chapter; those back and forths could be based on stuff we know from reading the hole book. I would want to keep that back and forth aimed at reaching some consensus that we could all agree was “correct”. I certainly don’t want anyone to be dissed but I also do want this to be a course where everyone is right; I still believe that there are right and wrong answers to scientific questions; well answers that are more right than wrong. But if you have some better ideas about how tyo conduct this “class”; I would especially like it if you had concrete ideas about how to get more participation, even for those in this hemisphere where it is still Summer (though school does start this week here in LA).

JK: For what it’s worth here are a few examples of some difficulties I encountered when reading Chapter 6.

pg 70 end of para #1, I’m unsure what a ‘control organisation’ is in this context. For instance is this a set of non-closed boxes as at the top of Fig 6.1 as mentioned in para 2, pg 72. Or is it the whole kit and caboodle?

RM: I think Powers is referring to the closed loop “organization” of a control system – the feedback loop – when he talks about control organization (or the organization of behavior). In figure 6.1 there are several closed loops but the overall architecture of this model is a closed loop control organization.

JK: pg 70 a three-level model is illustrated in Fig 6.1. It took me a good hour to label each box/arrow of Fig 6.1 and relate these tags to the text. So far I understand it each level is characterised by three boxes (input function - comparator - output function); yes?

RM: Martin gave a good answer to this. I would just say that a control system at a higher level than another is one whose perceptual inputs are based on perceptions coming from systems below it and whose outputs contribute to the references of the systems below it. The three boxes representing a control system is at some level relative to the control systems (including at the same level); but a control system is not a level.

JK:So, what’s the relationship between a level and a step (as in three steps of visual information processing)?

RM: In this case I would say that each step up corresponds to going up a level (of one system relative to another).

JK: Does each of these steps aim at and point to a different level? ‘The signal representing the actual relationship comes in from the left’ – where exactly is that again as I read ‘perceived’ in several places but not ‘actual signal’?

RM: Yes, each step represents a different level of perceptual processing; Th perception of spot-target relationship is based on perceptions of spot and target position; perceptions spot and target position are based on perceptions of the intensity of light reflected onto the retina by spot and target.

JK: Can a high-level system operate as a low-level one, be passed down the line? For instance with a proficient sport-player’s performance, or those of a concert pianist? I quite liked the Australian study when drivers returning from a ski field were breath-analysed for illicit substances. Even at the bottom of a twisty, winding road they were accident free yet many well over legal limits. On a familiar road they could perform almost automatically. But when encountering an unexpected obstacle on the road (a cone for instance) they could not execute evasive actions successfully.

RM; I don’t see how this is an example of a high level system operating as a low level one. My quick analysis is that this is an example of a higher level system controlling for the program or sequence of turns that get them done the twisty road; the alcohol probably slows neural conduction so this would affect the low level (fast acting) systems more than the high level systems that use these low level systems to control their perceptions; so the low level systems can’t correct for unexpected disturbances (like the cones) as well as they could when alcohol free. But the higher level systems can still work pretty. This would explain why Dylan Thomas, for example, could recite a poem from memory when drunk but slur the words all the way through; the low level systems that control for proper articulation don’t work very well due to the slowing of the neural conduction but the higher level systems that control for the high level perception of the poem are still working pretty well.

JK: pg72 line 2 has the word ‘First’ but I cannot find a ‘second’ in the subsequent text. It’s probably staring me in the face…

RM: I don’t think there is a “second”; he just assumes that you can see the two steps by looking at the diagram. The first step is the perception of the sate of the arm via the Effort receptors; the second is the Kinesthetic-Visual position detector above that.

JK: Pg 75 para 2 ‘… yet the feedback loop for all systems passes through the environment of the whole collection’. Is this like upper-levels of a house of cards built upon those bottom ones resting on the tabletop which is where contact occurs? Is this ‘radical reductionism’, the golden bullet, the gravity of living organisms?

RM: No, it’s more like complex perceptions being built from the sensory intensities – the results of stimulation of the rods, cones, hair cells (or the ear), taste buds, force sensors in the muscles, etc – that are our only interface with physio-chemical world (environment) outside our control systems.

JK: Pg 78 para 2, I’d be interested in hearing how and by whom the proposed model been modified over the past 40 years.

RM: I would say “no”, I have not seen any research that requires a change in the model. But that’s just my impression and as Kent demonstrated today I am not all knowing;-) Maybe others listening in can tell you about modifications they know of. I guess I would say that I don’t know of any modifications that I have had to make in order to get the model to fit the data I’ve collected in my research.

Best

Rick

–
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

On Mon, Aug 19, 2013 at 6:34 AM, Richard Marken rsmarken@gmail.com wrote:

[From Rick Marken (2013.08.18.1130)]

Thanks to Rupert Young for being the only student in class who raised his hand and gave his answers to the leading questions. I am attaching the summary of CH 6 which contains my answers to those questions and I think you’ll see that I basically agree with Rupert’s.

I hope there are people still listening in on the course. I think the lack of participation may be because everyone is on Facebook or Twitter (whatever that is) or because it’s Summer and they are sipping Vodka tonics beside their pools. But David and I will soldier on, hoping to spark some discussion. The study guide for the next chapter (7) will be posted soon.

Any suggestions about things we might to to get more involvement in the course – or regarding the content of the course – would be most welcome.

Best regards

Rick

–
Richard S. Marken PhD

rsmarken@gmail.com
www.mindreadings.com

5

[From Rick Marken (2013.08.19.1020)]

I’ve really got to remember to proof read before sending. In my answer to John’s question about the class I said:

···

What I meant to say was:

I would want to keep that back and forth aimed at reaching some consensus that we could all agree was “correct”. I certainly don’t want anyone to be dissed but I also DON’T want this to be a course where everyone is right.

I’m liberal, but to a degree!

Best

Rick

–
Richard S. Marken PhD
rsmarken@gmail.com

www.mindreadings.com

I would want to keep that back and forth aimed at reaching some consensus that we could all agree was “correct”. I certainly don’t want anyone to be dissed but I also do want this to be a course where everyone is right

6

Hi All,

JK: Pg 78 para 2, I’d be interested in hearing how and by whom the proposed model been modified over the past 40 years.

RM: I would say “no”, I have not seen any research that requires a change in the model. But that’s just my impression and as Kent demonstrated today I am not all knowing;-) Maybe others listening in can tell you about modifications they know of. I guess I would say that I don’t know of any modifications that I have had to make in order to get the model to fit the data I’ve collected in my research.

HB:

I would say “yes”. Contrast it’s with no doubt just Rick’s impression. But I doubt that he doesn’t know anything about possible modifications to PCT. I would say that problem is more in the question : "Why would scientist build a “theoretical construct” if not for trying to show with data and researches and testing how their “theoretical construct” is perfect, as it is probably in their minds. Why would they trouble themselves with proving that their model is not perfect.

I’m quite sure that there are myriad of researches that could change the PCT model. Of course not inside PCT model and not with purely PCT research tools as that is probably the most important condition for “change” or “modifictaion” in PCT.

It seems to me, at least for the last 13 years, that nobody was wellcome to change anything if he wasn’t close “PCT’er”. I got impression (as Rick) that the only “one” who can upgrade PCT are those who are fully devoted to PCT model and to the preffered way of testing the model. It’s usually like that. People are modeling and researching in the borders of their knowledge and habituated way of thinking. It’s probably hard to include other researches and knowledge into own theoretical model in which you invested so much time and effort. If you are not acquainted with others terminology and research strategy, it seems everything to complicated. It’s easier to persist in “used thinking”. It’s probably the way how HPCT works.

I think that Bill was aware of possible changes to PCT, if I remember right and if I can conclude from his first inspired wrtings. So there was an attempt to modify PCT model some years ago, but later disagrements with “upgrade” to PCT intensify till the end of conversation about modifications.

The whole conversation about possible changes is probably still somewhere on the CSGnet.

But anyway I assume that if PCT would want to reveal “full scientific power”, some critical changes will have to be made sooner or later. Just “abstract thinking” and testing what fit into “wanted system theory” will by my oppinion not give any significant progress.

RY :

  1. When we get to the chapter on the Reorganization System, let’s see if you still think this.

HB :

Well I assume that specialy in the mentioned Chapter will appear most of the troubles.

Best,

Boris

···

----- Original Message -----

From:
Richard Marken

To: CSGNET@LISTSERV.ILLINOIS.EDU

Sent: Tuesday, August 20, 2013 7:12 AM

Subject: Re: B:CP Course Summary of CH. 6 A Hierarchy of Control Systems

[From Rick Marken (2013.08.19.2210)]

[John Kirkland 20130819]

JK: It sure is helpful having the little summaries.

RM: Thanks John. Glad they help.

JK: BTW Rick for some of us it's now winter/spring. I know it's hard to accept New Year festivities may include swimming at the beach or river.

RM: Ah, you’re in that hemisphere. G’day.

JK: My niggle about the course thus far is it seems to have adopted a pre-packaged GLM approach: everybody's to read the set text and answer fixed questions. I was anticipating a more radical approach by way of illustrating PCT in action.  Some time back I had raised the question as to how a PCT pedagogy might differ from what I'll call the bulk default standard. In that traditional S-R dominant format what tended to happen when a reader n(student) got stuck or asked a question was they were ignored and/or got blamed for wasting everyone's time, and/or they were classified as stupid and summarily dismissed. It still happens.

RM: I’m open to suggestions. But I did want to go trough the book chapter by chapter; the class is an homage to Powers’ main opus, which should sit up on the shelf with the Principia and Origin of Species, if books are still around (to say nothing of coastal cities) in 100 years;-) I was hoping there would be more back and forth about each chapter; those back and forths could be based on stuff we know from reading the hole book. I would want to keep that back and forth aimed at reaching some consensus that we could all agree was “correct”. I certainly don’t want anyone to be dissed but I also do want this to be a course where everyone is right; I still believe that there are right and wrong answers to scientific questions; well answers that are more right than wrong. But if you have some better ideas about how tyo conduct this “class”; I would especially like it if you had concrete ideas about how to get more participation, even for those in this hemisphere where it is still Summer (though school does start this week here in LA).

JK: For what it's worth here are a few examples of some difficulties I encountered when reading Chapter 6.

pg 70 end of para #1, I'm unsure what a 'control organisation' is in this context. For instance is this a set of non-closed boxes as at the top of Fig 6.1 as mentioned in para 2, pg 72. Or is it the whole kit and caboodle?

RM: I think Powers is referring to the closed loop “organization” of a control system – the feedback loop – when he talks about control organization (or the organization of behavior). In figure 6.1 there are several closed loops but the overall architecture of this model is a closed loop control organization.

JK: pg 70 a three-level model is illustrated in Fig 6.1. It took me a good hour to label each box/arrow of Fig 6.1 and relate these tags to the text. So far I understand it each level is characterised by three boxes (input function - comparator - output function); yes?

RM: Martin gave a good answer to this. I would just say that a control system at a higher level than another is one whose perceptual inputs are based on perceptions coming from systems below it and whose outputs contribute to the references of the systems below it. The three boxes representing a control system is at some level relative to the control systems (including at the same level); but a control system is not a level.

JK:So, what's the relationship between a level and a step (as in three steps of visual information processing)?

RM: In this case I would say that each step up corresponds to going up a level (of one system relative to another).

JK: Does each of these steps aim at and point to a different level? 'The signal representing the actual relationship comes in from the left' -- where exactly is that again as I read 'perceived' in several places but not 'actual signal'?

RM: Yes, each step represents a different level of perceptual processing; Th perception of spot-target relationship is based on perceptions of spot and target position; perceptions spot and target position are based on perceptions of the intensity of light reflected onto the retina by spot and target.

JK: Can a high-level system operate as a low-level one, be passed down the line? For instance with a proficient sport-player’s performance, or those of a concert pianist? I quite liked the Australian study when drivers returning from a ski field were breath-analysed for illicit substances. Even at the bottom of a twisty, winding road they were accident free yet many well over legal limits. On a familiar road they could perform almost automatically. But when encountering an unexpected obstacle on the road (a cone for instance) they could not execute evasive actions successfully.

RM; I don’t see how this is an example of a high level system operating as a low level one. My quick analysis is that this is an example of a higher level system controlling for the program or sequence of turns that get them done the twisty road; the alcohol probably slows neural conduction so this would affect the low level (fast acting) systems more than the high level systems that use these low level systems to control their perceptions; so the low level systems can’t correct for unexpected disturbances (like the cones) as well as they could when alcohol free. But the higher level systems can still work pretty. This would explain why Dylan Thomas, for example, could recite a poem from memory when drunk but slur the words all the way through; the low level systems that control for proper articulation don’t work very well due to the slowing of the neural conduction but the higher level systems that control for the high level perception of the poem are still working pretty well.

JK: pg72 line 2 has the word ‘First’ but I cannot find a ‘second’ in the subsequent text. It’s probably staring me in the face…

RM: I don’t think there is a “second”; he just assumes that you can see the two steps by looking at the diagram. The first step is the perception of the sate of the arm via the Effort receptors; the second is the Kinesthetic-Visual position detector above that.

JK: Pg 75 para 2 ‘… yet the feedback loop for all systems passes through the environment of the whole collection’. Is this like upper-levels of a house of cards built upon those bottom ones resting on the tabletop which is where contact occurs? Is this ‘radical reductionism’, the golden bullet, the gravity of living organisms?

RM: No, it’s more like complex perceptions being built from the sensory intensities – the results of stimulation of the rods, cones, hair cells (or the ear), taste buds, force sensors in the muscles, etc – that are our only interface with physio-chemical world (environment) outside our control systems.

JK: Pg 78 para 2, I’d be interested in hearing how and by whom the proposed model been modified over the past 40 years.

RM: I would say “no”, I have not seen any research that requires a change in the model. But that’s just my impression and as Kent demonstrated today I am not all knowing;-) Maybe others listening in can tell you about modifications they know of. I guess I would say that I don’t know of any modifications that I have had to make in order to get the model to fit the data I’ve collected in my research.

Best

Rick

On Mon, Aug 19, 2013 at 6:34 AM, Richard Marken rsmarken@gmail.com wrote:

[From Rick Marken (2013.08.18.1130)]

  Thanks to Rupert Young for being the only student in class who raised his hand and gave his answers to the leading questions. I am attaching the summary of CH 6 which contains my answers to those questions and I think you'll see that I basically agree with Rupert's.

  I hope there are people still listening in on the course. I think the lack of participation may be because everyone is on Facebook or Twitter (whatever that is) or because it's Summer and they are sipping Vodka tonics beside their pools. But David and I will soldier on, hoping to spark some discussion. The study  guide for the next chapter (7) will be posted soon.

  Any suggestions about things we might to to get more involvement in the course -- or regarding the content of the course --  would be most welcome.

Best regards

Rick

–
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

–
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

7

[Martin Taylor 2013.08.20.10.11]

If you are following what Bill Powers said he thought might be the

case, RM is using “level” in two different senses. The first sense
is the 11 levels defined by different kinds of perceptual function
– intensities, relationships, events, princlples, etc. The second
sense is what JK called a “step”, a piece of the hierarchy in which
perceptual functions of one kind produce signals that contribute to
the inputs of higher in the hierarchy perceptual functions, which
might be of the same kind – sequences of sequences, relationships
of relationships, programs of subprograms, etc. – or of a different
kind (a higher one of the 11 levels). RM’s answer seems to conflate these two kinds of “level”. As I
remember it, Bill contemplated maintaining a strict hierarchy, but
without limiting the number of steps within a level, level being
defined by the kind of perceptual function involved. Future
discussion might be easier if we keep a distinction between what we
might call a “type level” or simply 'level" and what JK calls a
“step”. “Level” is like a floor in a house, “step” like a step in
the stairway between levels.
Martin

···

[From Rick Marken (2013.08.19.2210)]

[John Kirkland 20130819]

…

                JK:So, what's the relationship between a level

and a step (as in three steps of visual information
processing)?

            RM: In this case I would say that each step up

corresponds to going up a level (of one system relative
to another).

                JK: Does each of these steps aim at and point to

a different level? ‘The signal representing the
actual relationship comes in from the left’ – where
exactly is that again as I read ‘perceived’ in
several places but not ‘actual signal’?

            RM: Yes, each step represents a different level of

perceptual processing;

8

[From Rick Marken (2013.08.20.1000)]

···

On Tue, Aug 20, 2013 at 6:16 AM, Boris Hartman boris.hartman@masicom.net wrote:

Hi All,

JK: Pg 78 para 2, I’d be interested in hearing how and by whom the proposed model been modified over the past 40 years.

RM: I would say “no”, I have not seen any research that requires a change in the model. But that’s just my impression and as Kent demonstrated today I am not all knowing;-) Maybe others listening in can tell you about modifications they know of. I guess I would say that I don’t know of any modifications that I have had to make in order to get the model to fit the data I’ve collected in my research.

HB: I would say “yes”. Contrast it’s with no doubt just Rick’s impression. But I doubt that he doesn’t know anything about possible modifications to PCT. I would say that problem is more in the question : "Why would scientist build a “theoretical construct” if not for trying to show with data and researches and testing how their “theoretical construct” is perfect, as it is probably in their minds. Why would they trouble themselves with proving that their model is not perfect.

RM: They would do it because that is how science is done. Every experimental test of the PCT model can be viewed as an attempt to see whether or not it is wrong.

Like any scientific theory, PCT purports to be an explanation of some observable phenomenon. As we have seen (from Ch. 4 in B:CP on Feedback and Behavior) the phenomenon that PCT purports to explain is control as seen in the controlling done by living organisms. So every attempt to use PCT to explain an example of such control subjects the model to possible proof of it’s incorrectness (or, at least, it’s need for modification).

My answer to John’s question about whether there has been modification of the model described in B:CP over the last 40 years was based on the fact that I know of no example of a test of PCT – where PCT is used to explain some example of controlling by a living organism – where the PCT model has failed the test in such a way that there was a need for modification. As I said, there may have been such tests that required modifications but I am not aware of them; but I would love to hear about them.

PCT is a scientific theory and I am sure it is no more “true” than was Newtoniam physics or Einsteinian physics for that matter. But PCT is clearly a better theory of the behavior of organisms – which is a process of control – than is any version of the current theory of behavior – the causal model. So right now I just think PCT is the best model of the behavior of living systems that we have.

By the way, I’m sure there will be a need to modify some aspects of the theory described in B:CP, particularly the nature of the hierarchy of control and the nature of the learning process. Indeed, that’s what research in PCT should be about, I think: getting a better picture of the relationship between control systems in behavior (for example, research may show that the relationship between control systems is not even strictly hierarchical, as Martin has pointed out). But I do think that the basic organizational component of the model – the negative feedback control loop described in Ch. 5 – will hold up against whatever tests you can throw at it. But I encourage throwing those tests at it. I’ve been doing it for 30 years; I wish some more people would join me.

Best

Rick

–
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

9

[Martin Taylor 2013. 20. 13.49]

[From Rick Marken (2013.08.20.1000)]

By the way, I'm sure there will be a need to modify some aspects of the theory described in B:CP, particularly the nature of the hierarchy of control and the nature of the learning process. Indeed, that's what research in PCT should be about, I think: getting a better picture of the relationship between control systems in behavior (for example, research may show that the relationship between control systems is not even strictly hierarchical, as Martin has pointed out).

That wasn't what I understood Bill to say. What I understood was that he was sticking with "strictly hierarchic" but that there might well be sub-levels (what JK called "steps") within any level. Maybe I misunderstood Bill 20 years ago (or misremember now), but that's the structure I have been "pointing out".

Whether future research will indicate that the structure is or is not strictly hierarchic is something on which we have no evidence. Non-hierarchic organizations are quite possible. But that's not what I've been talking about these two days.

But I do think that the basic organizational component of the model -- the negative feedback control loop described in Ch. 5 -- will hold up against whatever tests you can throw at it. But I encourage throwing those tests at it. I've been doing it for 30 years; I wish some more people would join me.

I totally agree.

Martin

10

Hi Rick,

this time I put my text into yours…

···

----- Original Message -----

From:
Richard Marken

To: CSGNET@LISTSERV.ILLINOIS.EDU

Sent: Tuesday, August 20, 2013 7:03 PM

Subject: Re: B:CP Course Summary of CH. 6 A Hierarchy of Control Systems

[From Rick Marken (2013.08.20.1000)]

On Tue, Aug 20, 2013 at 6:16 AM, Boris Hartman boris.hartman@masicom.net wrote:

Hi All,

JK: Pg 78 para 2, I'd be interested in hearing how and by whom the proposed model been modified over the past 40 years.

RM:  I would say "no", I have not seen any research that requires a change in the model. But that's just my impression and as Kent demonstrated today I am not all knowing;-) Maybe others listening in can tell you about modifications they know of. I guess I would say that I don't know of any modifications that I have had to make in order to get the model to fit the data I've collected in my research.

HB: I would say "yes". Contrast it's with no doubt just Rick's impression. But I doubt that he doesn't know anything about possible modifications to PCT. I would say that problem is more in the question : "Why would scientist build a "theoretical construct" if not for trying to show with data and researches and testing how their "theoretical construct" is perfect, as it is probably in their minds. Why would they trouble themselves with proving that their model is not perfect.

RM: They would do it because that is how science is done. Every experimental test of the PCT model can be viewed as an attempt to see whether or not it is wrong.

HB :

The problem I see in science and in PCT is that results are mostly interpreted as scientists wants. How could you estimate that “experimental tests of the PCT model” are wrong, if theory is wrong in some parts ? With which knowledge do you “match” the results of tests ? Or on what bases you design tests ?

RM :

Like any scientific theory, PCT purports to be an explanation of some observable phenomenon. As we have seen (from Ch. 4 in B:CP on Feedback and Behavior) the phenomenon that PCT purports to explain is control as seen in the controlling done by living organisms. So every attempt to use PCT to explain an example of such control subjects the model to possible proof of it’s incorrectness (or, at least, it’s need for modification).

HB :

I also beleive it’s true that PCT purports “control in living organism” but there is no garanty that “inside control model” is right and that test on that bases are right. If PCT is partly wrong about how internal control works, also the interpretation of “observable phenomenon” are more likely to be partly wrong. I’m not saying that PCT is fundamentaly wrong. It has some “holes”, that can be seen if comparing it to some other knowledge and experiments.

I’m also trying to say that if those experiments and some knowledge would be included in PCT it would show more right results and provide better explanation of “observable phenomenon”.

RM :
My answer to John’s question about whether there has been modification of the model described in B:CP over the last 40 years was based on the fact that I know of no example of a test of PCT – where PCT is used to explain some example of controlling by a living organism – where the PCT model has failed the test in such a way that there was a need for modification.

HB :

As I see it, the problem are not tests of PCT but your interpretation of the tests on the bases of the theory that is partly wrong and maybe you sometimes make wrong conclusions. Like for example in your first simulation of “baseball catch”.

RM :

As I said, there may have been such tests that required modifications but I am not aware of them; but I would love to hear about them.

HB:
I’m sure you would like to hear about those experiments and knowledge, but that will not be possible in this moment if you are expecting me to do that. As I said many times before first some changes has to be made also to CSGnet and to “human equal approach”. But you can help yourself a little if you go through our past conversations…:)). And maybe there are some other friends of Bill who wouldn’t advise “hearing” about those facts as that could possibly mean some crucial modifications to the model and diagrams in B:CP.

RM :

PCT is a scientific theory and I am sure it is no more “true” than was Newtoniam physics or Einsteinian physics for that matter. But PCT is clearly a better theory of the behavior of organisms – which is a process of control – than is any version of the current theory of behavior – the causal model. So right now I just think PCT is the best model of the behavior of living systems that we have.

HB :

I agree with you. But would you like to see that PCT is as good model and is equaly good accepted in the World as “Newtoniam physics or Einsteinian physics” ? I never doubt about Bill’s geniality.

RM :

By the way, I’m sure there will be a need to modify some aspects of the theory described in B:CP, particularly the nature of the hierarchy of control and the nature of the learning process. Indeed, that’s what research in PCT should be about, I think: getting a better picture of the relationship between control systems in behavior (for example, research may show that the relationship between control systems is not even strictly hierarchical, as Martin has pointed out).

HB :

I agree that there is need to modify some aspects of the theory in B:CP. But if, as you said before, tests of the PCT are showing the “right and wrong” sides of the theory, so why wasn’t that done before ? Why now there are nedded researches in PCT that should show some “errors” and corrections or in what direction researches should be about ?

As the “strictly hierarchical relations between control systems” are concerned you could find that out aproximately 50 yeras ago if you would read some extra books. “Cooperation” among control units in living organisms is amazing. I really can’t think that you lived in “dark” with one possible organization of control units and possibly with “one” hierarchical organization.

Isn’t that a little bit naive thinking when we know what possibilities are given with practically endless connections between neurons ?

RM :

But I do think that the basic organizational component of the model – the negative feedback control loop described in Ch. 5 – will hold up against whatever tests you can throw at it. But I encourage throwing those tests at it. I’ve been doing it for 30 years; I wish some more people would join me.

HB :

The “basic organizational component of the model - the negative feedback loop” is very good, excelent. But I think that some improvements will still be necesary before it will “hold up against whatever tests you can throw at it”. There is still some mistery. And as I pointed before we don’t know how much of the control knowledge will be “stolen” by psychologist and how much scientific interest will be in tests with PCT control loop, to be “throwing at it”.

But by my oppinion that is not the main problem. I think that the main problem is how control units are organized to enable life. And actual hierarchical organization of control units in PCT is by my oppinion quite away from that goal.

So one of the basic Ashby’s conclusions that difference between dead and alive horse is obviuous enough, will be hard to prove scientifically with PCT.

Martin T. kindly remainded me (and I’m greatfull to him for that) that we have a little bit different concepts about how to use organization of control units and what kind of experimental support should we have. Some are satisfyed in testing the control loop for 30 or more years and it’s good that not more people joined, because it wouldn’t make necesary imrpovements to come to the man point : how do organisms “really” work and how do they “adapt” to physical and social environments. That’s the goal of some other people like me, and there are probably even more different posibilities of use of control units organization.

I think the main problem is in organizational concept of control units and the question : what is genetic control or what is life, although we could probably join that in paradigm : control is to live. But in which way the right answer could be obtained ? Only with PCT “abstract thinking” and testing or with “full” cooperation among sciences and other “different thinking people” with tolerance?

But I’m glad for one thing. You Rick weren’t “agresive” this time (by my control of perception) and you left quite “opened door” for possible cooperation (constructive conversation) between “two worlds”. I must admitt that I was pleasantly surprised.

Best

Rick

Best, Boris

–
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

11

Hi there!

Martin, Bill explained to me about ‘steps’ in an email, and I remember reading about them in the program level chapter. Is there a clear reference where Bill has made the statement that you make here as it would be VERY helpful to cite this?

Boris, did you read my article on feedforward?

Warren

···

On Tue, Aug 20, 2013 at 3:27 PM, Martin Taylor mmt-csg@mmtaylor.net wrote:

[Martin Taylor 2013.08.20.10.11]

[From Rick Marken (2013.08.19.2210)]

If you are following what Bill Powers said he thought might be the

case, RM is using “level” in two different senses. The first sense
is the 11 levels defined by different kinds of perceptual function
– intensities, relationships, events, princlples, etc. The second
sense is what JK called a “step”, a piece of the hierarchy in which
perceptual functions of one kind produce signals that contribute to
the inputs of higher in the hierarchy perceptual functions, which
might be of the same kind – sequences of sequences, relationships
of relationships, programs of subprograms, etc. – or of a different
kind (a higher one of the 11 levels).

RM's answer seems to conflate these two kinds of "level". As I

remember it, Bill contemplated maintaining a strict hierarchy, but
without limiting the number of steps within a level, level being
defined by the kind of perceptual function involved. Future
discussion might be easier if we keep a distinction between what we
might call a “type level” or simply 'level" and what JK calls a
“step”. “Level” is like a floor in a house, “step” like a step in
the stairway between levels.

Martin

–
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

[John Kirkland 20130819]

…

                JK:So, what's the relationship between a level

and a step (as in three steps of visual information
processing)?

            RM: In this case I would say that each step up

corresponds to going up a level (of one system relative
to another).

                JK: Does each of these steps aim at and point to

a different level? ‘The signal representing the
actual relationship comes in from the left’ – where
exactly is that again as I read ‘perceived’ in
several places but not ‘actual signal’?

            RM: Yes, each step represents a different level of

perceptual processing;

12

Hi Warren,

I read it. But as usualy I have some troubles with understanding a language concept. I’ll return it in day or two. Did you let also Martin to read it ?

Best,

Boris

13

Warren,

I don't know if the idea of steps within a level was ever put into

writing by Bill. I had used steps within a level in something I was
doing, and I asked him about whether that was kosher during CSG93.
He said he didn’t see why not. I can still remember the visual
circumstances of our discussion. That’s the only time I can remember
the subject ever coming up.

However, whether or not Bill endorsed the concept or mentioned it in

print, it is still a reasonable option that is totally within HPCT,
and one that is intuitively reasonable. After all, we do perceive
relationships of relationships: “A is to B as X is to ?” is a
favourite kind of test. Likewise sequences of sequences “a triple
alternation may be followed by a repetitive sequence and back to the
triple alternation”, events that are complexes of coincident events,
programs built of sub-programs, etc. etc.

Sorry I can't be of more help.

Martin
···

On 2013/08/21 4:49 AM, Warren Mansell
wrote:

Hi there!

        Martin, Bill explained to me about 'steps' in an email, and

I remember reading about them in the program level chapter.
Is there a clear reference where Bill has made the statement
that you make here as it would be VERY helpful to cite this?

Boris, did you read my article on feedforward?

Warren

      On Tue, Aug 20, 2013 at 3:27 PM, Martin

Taylor mmt-csg@mmtaylor.net
wrote:

          [Martin Taylor

2013.08.20.10.11]

[From Rick Marken (2013.08.19.2210)]

[John Kirkland 20130819]

…

                            JK:So, what's the relationship

between a level and a step (as in three
steps of visual information processing)?

                        RM: In this case I would say that each

step up corresponds to going up a level (of
one system relative to another).

                            JK: Does each of these steps aim at

and point to a different level? ‘The
signal representing the actual
relationship comes in from the left’ –
where exactly is that again as I read
‘perceived’ in several places but not
‘actual signal’?

                        RM: Yes, each step represents a different

level of perceptual processing;

          If you are following what Bill Powers said he thought

might be the case, RM is using “level” in two different
senses. The first sense is the 11 levels defined by
different kinds of perceptual function – intensities,
relationships, events, princlples, etc. The second sense
is what JK called a “step”, a piece of the hierarchy in
which perceptual functions of one kind produce signals
that contribute to the inputs of higher in the hierarchy
perceptual functions, which might be of the same kind –
sequences of sequences, relationships of relationships,
programs of subprograms, etc. – or of a different kind (a
higher one of the 11 levels).

          RM's answer seems to conflate these two kinds of "level".

As I remember it, Bill contemplated maintaining a strict
hierarchy, but without limiting the number of steps within
a level, level being defined by the kind of perceptual
function involved. Future discussion might be easier if we
keep a distinction between what we might call a “type
level” or simply 'level" and what JK calls a “step”.
“Level” is like a floor in a house, “step” like a step in
the stairway between levels.

              Martin

    --

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](http://www.psych-sci.manchester.ac.uk/staff/131406)


        See [teamstrial.net](http://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](http://www.amazon.co.uk/Transdiagnostic-Approach-Method-Levels-Therapy/dp/0415507642/ref=sr_1_1?ie=UTF8&qid=1351756948&sr=8-1) is available now.

        Check [www.pctweb.org](http://www.pctweb.org)
        for further information on Perceptual Control Theory
14

No worries! W

···

On 2013/08/21 4:49 AM, Warren Mansell
wrote:

Hi there!

        Martin, Bill explained to me about 'steps' in an email, and

I remember reading about them in the program level chapter.
Is there a clear reference where Bill has made the statement
that you make here as it would be VERY helpful to cite this?

Boris, did you read my article on feedforward?

Warren

      On Tue, Aug 20, 2013 at 3:27 PM, Martin

Taylor mmt-csg@mmtaylor.net
wrote:

          [Martin Taylor

2013.08.20.10.11]

[From Rick Marken (2013.08.19.2210)]

[John Kirkland 20130819]

…

                            JK:So, what's the relationship

between a level and a step (as in three
steps of visual information processing)?

                        RM: In this case I would say that each

step up corresponds to going up a level (of
one system relative to another).

                            JK: Does each of these steps aim at

and point to a different level? ‘The
signal representing the actual
relationship comes in from the left’ –
where exactly is that again as I read
‘perceived’ in several places but not
‘actual signal’?

                        RM: Yes, each step represents a different

level of perceptual processing;

          If you are following what Bill Powers said he thought

might be the case, RM is using “level” in two different
senses. The first sense is the 11 levels defined by
different kinds of perceptual function – intensities,
relationships, events, princlples, etc. The second sense
is what JK called a “step”, a piece of the hierarchy in
which perceptual functions of one kind produce signals
that contribute to the inputs of higher in the hierarchy
perceptual functions, which might be of the same kind –
sequences of sequences, relationships of relationships,
programs of subprograms, etc. – or of a different kind (a
higher one of the 11 levels).

          RM's answer seems to conflate these two kinds of "level".

As I remember it, Bill contemplated maintaining a strict
hierarchy, but without limiting the number of steps within
a level, level being defined by the kind of perceptual
function involved. Future discussion might be easier if we
keep a distinction between what we might call a “type
level” or simply 'level" and what JK calls a “step”.
“Level” is like a floor in a house, “step” like a step in
the stairway between levels.

              Martin

    --

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](http://www.psych-sci.manchester.ac.uk/staff/131406)


        See [teamstrial.net](http://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](http://www.amazon.co.uk/Transdiagnostic-Approach-Method-Levels-Therapy/dp/0415507642/ref=sr_1_1?ie=UTF8&qid=1351756948&sr=8-1) is available now.

        Check [www.pctweb.org](http://www.pctweb.org)
        for further information on Perceptual Control Theory