HPTC and Tacit Knowledge

[FredNickols]

Over in another world that I sometimes inhabit the subjects of explicit vs tacit knowledge are often center stage in some heated discussions. One of the examples used to illustrate the distinction is riding a bicycle. We can’t fully articulate how to ride a bicycle in ways that someone new to riding a bicycle could study and then successfully ride a bicycle. Hence, Michael Polanyi’s assertion that “we know more than we can tell.”

In the course of reading a paper about the epistemologies of knowledge, it occurred to me that what we often mean by tacit knowledge (e.g., being able to ride a bicycle) consists of reference signals that are established down at lower levels in the HPCT hierarchy. Thus, we are able to successfully ride a bicycle but we can’t say exactly just how it is we do it because those reference signals at lower levels are not accessible in ways that we can express in linguistic form.

All that said, I have a question for list members: What are the levels from the hierarchy that are involved in riding a bicycle?

P.S. I have a hunch that this is NOT an inconsequential question.

Regards,

Fred Nickols, CPT

Managing Partner

Distance Consulting LLC

“Assistance at A Distance^TM”

The Knowledge Workers’ Tool Room

Blog: Knowledge Worker Tools

www.nickols.us | fred@nickols.us

[chadtgreen]

[From Chad Green (2013.02.05.1619 EST)]

Great idea, Fred. Let's have some fun with this. Imagine PCT and HPCT as key components of a bicycle metaphor. How would they connect and interact? More importantly, would the metaphor be incomplete? If so, what's missing?

Best,
Chad

Chad Green, PMP
Program Analyst
Loudoun County Public Schools
21000 Education Court
Ashburn, VA 20148
Voice: 571-252-1486
Fax: 571-252-1633

"If you want sense, you'll have to make it yourself." - Norton Juster

Fred Nickols <fred@NICKOLS.US> 2/5/2013 1:05 PM >>>

Over in another world that I sometimes inhabit the subjects of explicit vs
tacit knowledge are often center stage in some heated discussions. One of
the examples used to illustrate the distinction is riding a bicycle. We
can't fully articulate how to ride a bicycle in ways that someone new to
riding a bicycle could study and then successfully ride a bicycle. Hence,
Michael Polanyi's assertion that "we know more than we can tell."

In the course of reading a paper about the epistemologies of knowledge, it
occurred to me that what we often mean by tacit knowledge (e.g., being able
to ride a bicycle) consists of reference signals that are established down
at lower levels in the HPCT hierarchy. Thus, we are able to successfully
ride a bicycle but we can't say exactly just how it is we do it because
those reference signals at lower levels are not accessible in ways that we
can express in linguistic form.

All that said, I have a question for list members: What are the levels from
the hierarchy that are involved in riding a bicycle?

P.S. I have a hunch that this is NOT an inconsequential question.

Regards,

Fred Nickols, CPT

Managing Partner

Distance Consulting LLC

"Assistance at A DistanceTM"

The Knowledge Workers <http://www.nickols.us/toolroom.html&gt; ' Tool Room

Blog: Knowledge Worker Tools

www.nickols.us <http://www.nickols.us/&gt; | fred@nickols.us

[rsmarken]

[From Rick Marken (2013.02.06.1015)]

FN: Over in another world that I sometimes inhabit the subjects of explicit vs tacit knowledge are often center stage in some heated discussions. One of the examples used to illustrate the distinction is riding a bicycle. We can’t fully articulate how to ride a bicycle in ways that someone new to riding a bicycle could study and then successfully ride a bicycle. Hence, Michael Polanyi’s assertion that “we know more than we can tell.”

In the course of reading a paper about the epistemologies of knowledge, it occurred to me that what we often mean by tacit knowledge (e.g., being able to ride a bicycle) consists of reference signals that are established down at lower levels in the HPCT hierarchy. Thus, we are able to successfully ride a bicycle but we can’t say exactly just how it is we do it because those reference signals at lower levels are not accessible in ways that we can express in linguistic form.

RM: I’d say that PCT is a model of all “tacit knowledge”, which includes how to ride a bike as well as our incomplete and inaccurate “explicit” knowledge of how we think we ride a bike (because we are tacitly organized to be able to control for producing such explanations, which are themselves high level controlled perceptions). So I would say what is called “explicit knowledge” is included in the tacit knowledge that is the whole PCT hierarchy. And the PCT hierarchy is itself explicit knowledge that may not correctly represent the true tacit knowledge that is the way out nervous system operates in a way that let’s us have this discussion of tacit and explicit knowledge;-)

How’s that for tacitly being confusing about my explicit knowledge of what’s going on with tacit and explicit knowledge.

FN: All that said, I have a question for list members: What are the levels from the hierarchy that are involved in riding a bicycle?

RM: I think a first step at figuring it out would be to look at Bill’s model of balancing an inverted pendulum. I believe he first two levels are velocity and position perceptions. The model you come up with would be your current “explicit” knowledge of the neural basis of riding a bike. The mental processes that let you control for producing that model represent you tacit knowledge of how to control much higher level perceptions – perceptions of mathematical relationships, principles of model construction, etc.

Best

Rick

···

On Tue, Feb 5, 2013 at 10:05 AM, Fred Nickols fred@nickols.us wrote:

P.S. I have a hunch that this is NOT an inconsequential question.

Regards,

Fred Nickols, CPT

Managing Partner

Distance Consulting LLC

“Assistance at A Distance^TM”

The Knowledge Workers’ Tool Room

Blog: Knowledge Worker Tools

www.nickols.us | fred@nickols.us

–
Richard S. Marken PhD
rsmarken@gmail.com

www.mindreadings.com

[FredNickols]

[From Fred Nickols (2013.02.06.1241 AZT)]

Thanks, Rick.

Fred Nickols

···

From: Control Systems Group Network (CSGnet) [mailto:CSGNET@LISTSERV.ILLINOIS.EDU] On Behalf Of Richard Marken
Sent: Wednesday, February 06, 2013 11:17 AM
To: CSGNET@LISTSERV.ILLINOIS.EDU
Subject: Re: HPTC and Tacit Knowledge

[From Rick Marken (2013.02.06.1015)]

On Tue, Feb 5, 2013 at 10:05 AM, Fred Nickols fred@nickols.us wrote:

FN: Over in another world that I sometimes inhabit the subjects of explicit vs tacit knowledge are often center stage in some heated discussions. One of the examples used to illustrate the distinction is riding a bicycle. We can’t fully articulate how to ride a bicycle in ways that someone new to riding a bicycle could study and then successfully ride a bicycle. Hence, Michael Polanyi’s assertion that “we know more than we can tell.”

In the course of reading a paper about the epistemologies of knowledge, it occurred to me that what we often mean by tacit knowledge (e.g., being able to ride a bicycle) consists of reference signals that are established down at lower levels in the HPCT hierarchy. Thus, we are able to successfully ride a bicycle but we can’t say exactly just how it is we do it because those reference signals at lower levels are not accessible in ways that we can express in linguistic form.

RM: I’d say that PCT is a model of all “tacit knowledge”, which includes how to ride a bike as well as our incomplete and inaccurate “explicit” knowledge of how we think we ride a bike (because we are tacitly organized to be able to control for producing such explanations, which are themselves high level controlled perceptions). So I would say what is called “explicit knowledge” is included in the tacit knowledge that is the whole PCT hierarchy. And the PCT hierarchy is itself explicit knowledge that may not correctly represent the true tacit knowledge that is the way out nervous system operates in a way that let’s us have this discussion of tacit and explicit knowledge;-)

How’s that for tacitly being confusing about my explicit knowledge of what’s going on with tacit and explicit knowledge.

FN: All that said, I have a question for list members: What are the levels from the hierarchy that are involved in riding a bicycle?

RM: I think a first step at figuring it out would be to look at Bill’s model of balancing an inverted pendulum. I believe he first two levels are velocity and position perceptions. The model you come up with would be your current “explicit” knowledge of the neural basis of riding a bike. The mental processes that let you control for producing that model represent you tacit knowledge of how to control much higher level perceptions – perceptions of mathematical relationships, principles of model construction, etc.

Best

Rick

P.S. I have a hunch that this is NOT an inconsequential question.

Regards,

Fred Nickols, CPT

Managing Partner

Distance Consulting LLC

“Assistance at A Distance^TM”

The Knowledge Workers’ Tool Room

Blog: Knowledge Worker Tools

www.nickols.us | fred@nickols.us

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

[Bill_Powers1]

[From Bill Powers O(2013.02.07.1330 MST)]

Fred Nickols (2013.02.06.1241 AZT)–

Rick Marken (2013.02.06.1015)

BP: I think that what is called “knowledge” or “explicit
knowledge” can probably be equated to verbal knowledge, knowledge as
expressed in language. That is very different from what is represented by
the parameters of control systems as forms of perceptual input functions,
gain, delay, integration constant, and so on. Knowledge expressed in
words is about the processes that actually convert one kind of
signals into another, or that generate muscle contractions and physical
forces. But the words are not what make things happen. You can say or
think “hit the nail with the hammer” as often as you like, and
the hammer will just lie there on the workbench and the nail won’t sink
in one micron more.

To me, this means that “implicit knowledge” doesn’t actually
exist. If there’s no language involved, then what is going on is not what
we call knowledge at all – no descriptive language is needed to make a
control process happen.

Best,

Bill P.

P.S. Moving day is Saturday. This is Thursday, so I won’t be posting much
for a while.

[rsmarken]

[From Rick Marken (2013.02.07.1400)]

Bill Powers O(2013.02.07.1330 MST)–

Fred Nickols (2013.02.06.1241 AZT)–

Rick Marken (2013.02.06.1015)

BP: I think that what is called “knowledge” or “explicit
knowledge” can probably be equated to verbal knowledge, knowledge as
expressed in language. That is very different from what is represented by
the parameters of control systems as forms of perceptual input functions,
gain, delay, integration constant, and so on. Knowledge expressed in
words is about the processes that actually convert one kind of
signals into another, or that generate muscle contractions and physical
forces. But the words are not what make things happen. You can say or
think “hit the nail with the hammer” as often as you like, and
the hammer will just lie there on the workbench and the nail won’t sink
in one micron more.

To me, this means that “implicit knowledge” doesn’t actually
exist. If there’s no language involved, then what is going on is not what
we call knowledge at all – no descriptive language is needed to make a
control process happen.

RM: This is exactly how I see it, though I was being more permissive than you inasmuch as I was willing to count “implicit knowledge” as the parameters of control that make it possible to perform behaviors like ridding a bike.

BP: P.S. Moving day is Saturday. This is Thursday, so I won’t be posting much
for a while.

RM: Good luck on the move.

Best

Rick

···

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

[FredNickols]

[From Fred Nickols (2013.02.07.1517 AZT)]

A few points of clarification. Explicit or codified knowledge does indeed hinge on language. That is commonly referred to as “know-about.” The other kind – tacit knowledge – is reflected in and inferred from the doing of something (e.g., riding a bike, repairing a radar, solving an equation, etc.). We see someone riding a bike and we can legitimately say “That person knows how to ride a bike.” But, chances are, that person can’t fully articulate how that is done. Hence, Polanyi’s notion of tacit knowledge (i.e., we know more than we can tell). I coined a third category about a dozen years ago (“implicit” knowledge) to refer to knowledge that can be articulated but hasn’t yet been articulated. That is the kind I and many others can “dig out” of someone we typically call a “subject matter expert.” Using the generally accepted definition of tacit knowledge, it seemed to me that what we are really referring to is that hierarchy of control systems making up the human being, some of which (at the higher levels I suspect) can be described/articulated but many of which (e.g., staying upright on a bike) can’t be adequately described in ways that an individual who can’t ride a bike can simply pick up, read and then go ride a bike.

So I’m not inclined to tell anyone that “tacit” knowledge doesn’t really exist; instead, I’m more inclined to suggest to them that it exists in an embedded form in a hierarchy of control systems.

Fred Nickols

···

From: Control Systems Group Network (CSGnet) [mailto:CSGNET@LISTSERV.ILLINOIS.EDU] On Behalf Of Richard Marken
Sent: Thursday, February 07, 2013 3:04 PM
To: CSGNET@LISTSERV.ILLINOIS.EDU
Subject: Re: HPTC and Tacit Knowledge

[From Rick Marken (2013.02.07.1400)]

Bill Powers O(2013.02.07.1330 MST)–

Fred Nickols (2013.02.06.1241 AZT)–

Rick Marken (2013.02.06.1015) –

BP: I think that what is called “knowledge” or “explicit knowledge” can probably be equated to verbal knowledge, knowledge as expressed in language. That is very different from what is represented by the parameters of control systems as forms of perceptual input functions, gain, delay, integration constant, and so on. Knowledge expressed in words is about the processes that actually convert one kind of signals into another, or that generate muscle contractions and physical forces. But the words are not what make things happen. You can say or think “hit the nail with the hammer” as often as you like, and the hammer will just lie there on the workbench and the nail won’t sink in one micron more.

To me, this means that “implicit knowledge” doesn’t actually exist. If there’s no language involved, then what is going on is not what we call knowledge at all – no descriptive language is needed to make a control process happen.

RM: This is exactly how I see it, though I was being more permissive than you inasmuch as I was willing to count “implicit knowledge” as the parameters of control that make it possible to perform behaviors like ridding a bike.

BP: P.S. Moving day is Saturday. This is Thursday, so I won’t be posting much for a while.

RM: Good luck on the move.

Best

Rick

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

[Bob_Hintz]

from bob hintz 2-7-13

I would simply like to point out that the activity of picking up the hammer and hitting the nail is very similar to the activity of saying “pick up the hammer and hit the nail” from the standpoint of the hierarchical control system producing that behavior. It is only when the saying is directed toward the hammer that it becomes problematic. How we talk is quite as “tacit” as how to ride a bike.

bob

···

On Thu, Feb 7, 2013 at 4:25 PM, Fred Nickols fred@nickols.us wrote:

[From Fred Nickols (2013.02.07.1517 AZT)]

A few points of clarification. Explicit or codified knowledge does indeed hinge on language. That is commonly referred to as “know-about.” The other kind – tacit knowledge – is reflected in and inferred from the doing of something (e.g., riding a bike, repairing a radar, solving an equation, etc.). We see someone riding a bike and we can legitimately say “That person knows how to ride a bike.” But, chances are, that person can’t fully articulate how that is done. Hence, Polanyi’s notion of tacit knowledge (i.e., we know more than we can tell). I coined a third category about a dozen years ago (“implicit” knowledge) to refer to knowledge that can be articulated but hasn’t yet been articulated. That is the kind I and many others can “dig out” of someone we typically call a “subject matter expert.” Using the generally accepted definition of tacit knowledge, it seemed to me that what we are really referring to is that hierarchy of control systems making up the human being, some of which (at the higher levels I suspect) can be described/articulated but many of which (e.g., staying upright on a bike) can’t be adequately described in ways that an individual who can’t ride a bike can simply pick up, read and then go ride a bike.

So I’m not inclined to tell anyone that “tacit” knowledge doesn’t really exist; instead, I’m more inclined to suggest to them that it exists in an embedded form in a hierarchy of control systems.

Fred Nickols

From: Control Systems Group Network (CSGnet) [mailto:CSGNET@LISTSERV.ILLINOIS.EDU] On Behalf Of Richard Marken
Sent: Thursday, February 07, 2013 3:04 PM

To: CSGNET@LISTSERV.ILLINOIS.EDU
Subject: Re: HPTC and Tacit Knowledge

[From Rick Marken (2013.02.07.1400)]

Bill Powers O(2013.02.07.1330 MST)–

Fred Nickols (2013.02.06.1241 AZT)–

Rick Marken (2013.02.06.1015) –

BP: I think that what is called “knowledge” or “explicit knowledge” can probably be equated to verbal knowledge, knowledge as expressed in language. That is very different from what is represented by the parameters of control systems as forms of perceptual input functions, gain, delay, integration constant, and so on. Knowledge expressed in words is about the processes that actually convert one kind of signals into another, or that generate muscle contractions and physical forces. But the words are not what make things happen. You can say or think “hit the nail with the hammer” as often as you like, and the hammer will just lie there on the workbench and the nail won’t sink in one micron more.

To me, this means that “implicit knowledge” doesn’t actually exist. If there’s no language involved, then what is going on is not what we call knowledge at all – no descriptive language is needed to make a control process happen.

RM: This is exactly how I see it, though I was being more permissive than you inasmuch as I was willing to count “implicit knowledge” as the parameters of control that make it possible to perform behaviors like ridding a bike.

BP: P.S. Moving day is Saturday. This is Thursday, so I won’t be posting much for a while.

RM: Good luck on the move.

Best

Rick

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

[Bill_Powers1]

BP: My argument is that being able to talk about how to ride a bike doesn't establish the kind of perceptions or the synaptic connections and forms of connection needed actually to do it. That can be done only by reorganization at all the relevant nonverbal levels.

Also, I'm, generally against overstretching metaphors. "Knowing" at one level is not just an extension of "knowing" at a lower level. We have a theory that distinguishes pretty sharply between what one level can do and what another, higher or lower, can do. Calling two very different processes by the same name doesn't make them the same process. It's not a question of being able to "fully" articulate a skill. At the level of doing, the process doesn't require articulation at all. Doing is doing, articulation is imagining and talking about what is to be done.

Best,

Bill P.

···

At 03:25 PM 2/7/2013 -0700, Fred Nickols wrote:

[From Fred Nickols (2013.02.07.1517 AZT)]

A few points of clarification. Explicit or codified knowledge does indeed hinge on language. That is commonly referred to as �know-about.� The other kind � tacit knowledge � is reflected in and inferred from the doing of something (e.g., riding a bike, repairing a radar, solving an equation, etc.). We see someone riding a bike and we can legitimately say �That person knows how to ride a bike.� But, chances are, that person can�t fully articulate how that is done.

[FredNickols]

[From Fred Nickols (2013.02.08.0905 AZT)]

I agree with your argument, Bill. I guess I'm not making myself clear.

Fred Nickols

From: Control Systems Group Network (CSGnet)
[mailto:CSGNET@LISTSERV.ILLINOIS.EDU] On Behalf Of Bill Powers
Sent: Friday, February 08, 2013 8:47 AM
To: CSGNET@LISTSERV.ILLINOIS.EDU
Subject: Re: HPTC and Tacit Knowledge

>[From Fred Nickols (2013.02.07.1517 AZT)]
>
>A few points of clarification. Explicit or codified knowledge does
>indeed hinge on language. That is commonly referred to as
>"know-about." The other kind - tacit knowledge - is reflected in and
>inferred from the doing of something (e.g., riding a bike, repairing a
>radar, solving an equation, etc.). We see someone riding a bike and we
>can legitimately say "That person knows how to ride a bike." But,
>chances are, that person can't fully articulate how that is done.

BP: My argument is that being able to talk about how to ride a bike

doesn't

establish the kind of perceptions or the synaptic connections and forms of
connection needed actually to do it. That can be done only by

reorganization

at all the relevant nonverbal levels.

Also, I'm, generally against overstretching metaphors. "Knowing" at one

level

is not just an extension of "knowing" at a lower level. We have a theory

that

distinguishes pretty sharply between what one level can do and what
another, higher or lower, can do. Calling two very different processes by

the

same name doesn't make them the same process. It's not a question of
being able to "fully" articulate a skill. At the level of doing, the

process

doesn't require articulation at all. Doing is doing, articulation is

imagining and

···

-----Original Message-----
At 03:25 PM 2/7/2013 -0700, Fred Nickols wrote:
talking about what is to be done.

Best,

Bill P.

[chadtgreen]

[From Chad Green (2013.02.08.1134 EST)]

BP: My argument is that being able to talk about how to ride a bike
doesn't establish the kind of perceptions or the synaptic connections
and forms of connection needed actually to do it. That can be done only
by reorganization at all the relevant nonverbal levels.

CG: Does a visual exist that depicts the integration of PCT,
reorganization, and HPCT all in one?

Best,
Chad

Chad Green, PMP
Program Analyst
Loudoun County Public Schools
21000 Education Court
Ashburn, VA 20148
Voice: 571-252-1486
Fax: 571-252-1633

"If you want sense, you'll have to make it yourself." - Norton Juster

Bill Powers <powers_w@FRONTIER.NET> 2/8/2013 10:46 AM >>>

[From Fred Nickols (2013.02.07.1517 AZT)]

A few points of clarification. Explicit or
codified knowledge does indeed hinge on
language. That is commonly referred to as
“know-about.�? The other kind � tacit knowledge
� is reflected in and inferred from the doing of

something (e.g., riding a bike, repairing a
radar, solving an equation, etc.). We see
someone riding a bike and we can legitimately
say “That person knows how to ride a
bike.�? But, chances are, that person can’t fully articulate how

that is done.

BP: My argument is that being able to talk about
how to ride a bike doesn't establish the kind of
perceptions or the synaptic connections and forms
of connection needed actually to do it. That can
be done only by reorganization at all the relevant nonverbal levels.

Also, I'm, generally against overstretching
metaphors. "Knowing" at one level is not just an
extension of "knowing" at a lower level. We have
a theory that distinguishes pretty sharply
between what one level can do and what another,
higher or lower, can do. Calling two very
different processes by the same name doesn't make
them the same process. It's not a question of
being able to "fully" articulate a skill. At the
level of doing, the process doesn't require
articulation at all. Doing is doing, articulation
is imagining and talking about what is to be done.

Best,

Bill P.

···

At 03:25 PM 2/7/2013 -0700, Fred Nickols wrote:

[rsmarken]

[From Rick Marken (2013.02.08.1550)]

Chad Green (2013.02.08.1134 EST)–

CG: Does a visual exist that depicts the integration of PCT,

reorganization, and HPCT all in one?

RM: HPCT and PCT are the same thing; the H in HPCT just calls attention to the fact that PCT posits a hierarchy of control systems as the basis of purposeful behavior. I think the best visual of HPCT and reorganization – ie. just about the whole schmear – is Fig. 14.1 on p. 191 of the second edition of Behavior: The Control of Perception.

Best

Rick

···

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

[MartinT]

Martin Taylor 2013.02.09.12.48]

PCT and HPCT are not the same thing. HPCT is a specific hypothetical

organization of perceptual control systems. An indefinite numbers of
other organizations of perceptual control systems can be
hypothesized that would be properly described as PCT. Even if they
postulated a hierarchical organization of perception in many levels
(as psychologists have done for nearly 200 years), they would not be
HPCT if they didn’t conform to the structure Bill Powers presented.
PCT is, I think, required by the laws of thermodynamics to be a
necessary property of life (I argued for its necessity in my
Editorial for the PCT issue of the International Journal of
Human-Computer studies). HPCT must be demonstrated by experiment
either to be required or to be a better match to the observed
behaviour of living organisms than are other structures either
already postulated or to be proposed in the future.
Just to riff off a couple of possible non-HPCT variants of PCT
(without any claim to whether they are actually viable), one could
postulate that the individual control units control vector-valued
perceptions rather than scalar-valued perceptions; one could suggest
(and I have) that the categorical boundary between analogue and
logical perceptions may be not a level, but may instead be a
boundary between parallel systems, each of which have many levels.
There are doubtless many other possibiltiies.
None of these possibilities would be HPCT, though both of the
suggested ones would be PCT and both presuppose multiple levels of
perception. If either of them were to be fleshed out to provide
specific predictions, and if experiments could be devised in which
they and HPCT gave different predictions, it is possible that HPCT
might not give the better prediction and the other would be
preferred. Also, it is quite possible that different organisms, all of which
must control perceptions if they are to be considered alive, may
structure their control systems differently. Maybe humans operate as
HPCT says, and maybe trees or ants do not. We have different
physical structures from them; it seems a bit arrogant to assert
without experimental evidence that we must have the same behavioural
structures as they do.
It is true that on CSGnet, HPCT is often taken for granted when
“PCT” is being discussed, but that doesn’t mean they are the same
thing, any more than all tissues are Kleenex, in the last century
all cameras were Kodaks, or all photocopies are Xeroxes.
Martin

···

On 2013/02/8 6:51 PM, Richard Marken
wrote:

[From Rick Marken (2013.02.08.1550)]

      Chad Green

(2013.02.08.1134 EST)–

CG: Does a visual exist that depicts the integration of PCT,

      reorganization, and HPCT all in one?

  RM: HPCT and PCT are the same thing; the H in HPCT just calls

attention to the fact that PCT posits a hierarchy of control
systems as the basis of purposeful behavior. I think the best
visual of HPCT and reorganization – ie. just about the whole
schmear – is Fig. 14.1 on p. 191 of the second edition of
Behavior: The Control of Perception.

[Kent]

[From Kent McClelland 2013.02.10.1430]

Chad, another visual that does a nice job of laying out the relationship between PCT and HPCT is Ted Cloak’s video/PPT presentation, which is available for download from his website:
www.tedcloak.com

Ted has produced some pretty cool graphics.

Kent

···

On 2013/02/8 6:51 PM, Richard Marken wrote:

[From Rick Marken (2013.02.08.1550)]

Chad Green (2013.02.08.1134 EST)–

CG: Does a visual exist that depicts the integration of PCT,

reorganization, and HPCT all in one?

RM: HPCT and PCT are the same thing; the H in HPCT just calls attention to the fact that PCT posits a hierarchy of control systems as the basis of purposeful behavior. I think the best visual of HPCT and reorganization – ie. just about the whole schmear –
is Fig. 14.1 on p. 191 of the second edition of Behavior: The Control of Perception.

[rsmarken]

[From Rick Marken (2013.02.09.1500)]

Martin Taylor (2013.02.09.12.48)–

RM: HPCT and PCT are the same thing;
MT: PCT and HPCT are not the same thing. HPCT is a specific hypothetical
organization of perceptual control systems. An indefinite numbers of
other organizations of perceptual control systems can be
hypothesized that would be properly described as PCT.

RM:Yes there are other possible organizations. But why go off hypothesizing other organizations before you have found the one Bill proposed wanting? So far PCT/HPCT, as proposed by Powers, has passed all tests; no need for change so far; not that there won’t be in the future.

MT: Even if they

postulated a hierarchical organization of perception in many levels
(as psychologists have done for nearly 200 years), they would not be
HPCT if they didn’t conform to the structure Bill Powers presented.

RM: Right. So why worry about those alternatives until we have rejected the one that has worked so far?

MT: PCT is, I think, required by the laws of thermodynamics to be a

necessary property of life (I argued for its necessity in my
Editorial for the PCT issue of the International Journal of
Human-Computer studies).

RM: Ah, so PCT is the part of PCT/HPCT that doesn’t have to be tested;-)

MT: HPCT must be demonstrated by experiment

either to be required or to be a better match to the observed
behaviour of living organisms than are other structures either
already postulated or to be proposed in the future.

RM: I think this is true of what you think of as PCT as well as HPCT. It’s all got to be tested.

MT: Just to riff off a couple of possible non-HPCT variants of PCT

(without any claim to whether they are actually viable), one could
postulate that the individual control units control vector-valued
perceptions rather than scalar-valued perceptions;

RM: Of course, one could postulate that. But why do so if the scaler version of the model accounts for all the data just fine? It seems to me that I would have to see some data that cannot be handled by the current version of PCT/HPCT before I’d start postulating changes like that to the model.

MT: one could suggest

(and I have) that the categorical boundary between analogue and
logical perceptions may be not a level, but may instead be a
boundary between parallel systems, each of which have many levels.
There are doubtless many other possibiltiies.

RM: I know. Some like to suggest; some like to test. I think suggestions like those you mention are valuable only if you can also suggest ways to test them.

MT: None of these possibilities would be HPCT, though both of the

suggested ones would be PCT and both presuppose multiple levels of
perception. If either of them were to be fleshed out to provide
specific predictions, and if experiments could be devised in which
they and HPCT gave different predictions, it is possible that HPCT
might not give the better prediction and the other would be
preferred.

RM: Yes, the suggestions are worthwhile, but only if you can test them. If the data demand that the new suggestions are better than those in the current PCT/HPCT model then you can call the model something else if you like; maybe MTPCT – Martin Taylor’s perceptual control theory. But I think it’s a lot less confusing to just consider the current version of PCT/HPCT to be ground truth and then devise tests to see if we have to revise it.

Best

Rick

···

Also, it is quite possible that different organisms, all of which

must control perceptions if they are to be considered alive, may
structure their control systems differently. Maybe humans operate as
HPCT says, and maybe trees or ants do not. We have different
physical structures from them; it seems a bit arrogant to assert
without experimental evidence that we must have the same behavioural
structures as they do.

Martin

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

[MartinT]

[Martin Taylor 2013.02.09.23.35]

Good. I'm glad we agree with no need for argument this time. PCT and

HPCT are different beasts. HPCT is one possible instantiation of
PCT. Others are possible. I just wanted to correct the
misapprehension you might have caused in some readers when you said,
obviously inadvertently, that HPCT and PCT are the same thing.
Martin

···

On 2013/02/9 5:59 PM, Richard Marken
wrote:

[From Rick Marken (2013.02.09.1500)]

        Martin Taylor

(2013.02.09.12.48)–

RM: HPCT and PCT are the same thing;
MT: PCT and HPCT are not the same thing. HPCT is a specific
hypothetical organization of perceptual control systems. An
indefinite numbers of other organizations of perceptual
control systems can be hypothesized that would be properly
described as PCT.

      RM:Yes there are other possible organizations.

[rsmarken]

[From Rick Marken (2013.02.09.2030)]

[Martin Taylor 2013.02.09.23.35]

[From Rick Marken (2013.02.09.1500)]

        Martin Taylor

(2013.02.09.12.48)–

RM: HPCT and PCT are the same thing;
MT: PCT and HPCT are not the same thing. HPCT is a specific
hypothetical organization of perceptual control systems. An
indefinite numbers of other organizations of perceptual
control systems can be hypothesized that would be properly
described as PCT.

      RM:Yes there are other possible organizations.

MT: Good. I'm glad we agree with no need for argument this time. PCT and

HPCT are different beasts. HPCT is one possible instantiation of
PCT. Others are possible. I just wanted to correct the
misapprehension you might have caused in some readers when you said,
obviously inadvertently, that HPCT and PCT are the same thing.

RM: Sure. We agree. PCT and HPCT are completely different;-) Now I’ll just back in my armchair and think about all those other non PCT theories and see which one turns out to be best: APCT, BPCT, CPCT… Yawn

Best

Rick

···

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

[_Boris_Hartman]

Hi all,

well I think this is the same old story : who is the enemy and who is the friend of PCT (HPCT), who is red and who is white. But is this really the question PCT’ers should talk about ?

RM: Sure. We agree. PCT and HPCT are completely different;-) Now I’ll just back in my armchair and think about all those other non PCT theories and see which one turns out to be best: APCT, BPCT, CPCT… Yawn

HB :

Rick I think that you are missing the point Martin tries to tell you, because it seem to me, that you are too emotional (irational) again about PCT, probably working on some lower level of HPCT, where some “local unstabilities” are occuring

Sometimes you give me the imperesiion of a “robot-car” which is blindly bumping into the wall (of course with no feed-back). You said it nicely for youself

RM earlier :

I think I get into arguments (conflicts) so often on CSGNet because I am typically controlling for some perceptions (like PCT and liberalism) with such high gain that I tend to push back against disturbances to these perceptions rather forcefully (which just magnifies the conflict). But I can purposefully lower the gain when I want….

HB :

Maybe it’s really time that you change something. Although I must admitt from time to time that you are workung on yourself.

If I understood Martin right ….

MT earlier :

PCT and HPCT are not the same thing. HPCT is a specific hypothetical organization of perceptual control systems. An indefinite numbers of other organizations of perceptual control systems can be hypothesized that would be properly described as PCT. Even if they postulated a hierarchical organization of perception in many levels (as psychologists have done for nearly 200 years), they would not be HPCT if they didn’t conform to the structure Bill Powers presented.

HB :

Martin is probably trying to tell us that PCT is a general control unit (Bill’s) and one of the structural blocks that is used in specific organization of PCT units (Bill’s HPCT) which appears in specific part of nervous system (behavioral part).

Now the question, as I understand everything, is : do all parts of nervous system behave as PCT predicts and are all parts following logic of Bill’s HPCT ? Or can all parts of organism be described with “PCT tools” and different HPCT, what could mean different organizations of PCT units ?

As I understand PCT and specific Bill’s HPCT, all the testings are done on one level of the “Grand output” as Ashby would put it. So as I see it (and nobody says it’s right) the PCT testings are done in range of “outer behavior” of living organisms and of course everything could be 100% O.K. because it’s in the range of PCT theory. Usually experiments and other researches are done so to confirm experimenters expectations.

Although I think that on this generalized level Rick’s testing could be right (to 99,9% probability). But the problem I see here is what is tested all the time.

The question and problem I see here is, what is happening in parts of organism which compose the “Grand output” – behavior ? Could everything what is going on in organism be modeled or described in Bill’s HPCT language and thus give the right bases for Rick’s testing and his perfect results ?

So it’s somehow obvious to me that Martin doesn’t doubt in PCT, but doubt in specific HPCT that make construction of organism (“real machine”).

So it seems to me, that Martin is trying to tell us something else. That maybe there are parts of nervous system which could have some other principles or mechanisms of working, different from HPCT, although it seems to me that he thinks that on the general level (behavior) all theories and physiological experiments should show in the direction of PCT or final results which Rick see through his testings as right.

Here I’m not sure, so I’m sorry Martin if I didin’t understand you well.

Although I think that PCT is a very good mainframe, some details still have to be analyzed again inside it.

MT earlier :

I don’t know if the paper I attach is the paper to which you refer, but it uses the mouse vibrissae. It’s one I had intended to bring up in this forum anyway, for two reasons. Firstly, from time to time questions have been raised about the computational abilities of neurons, and this paper seems to show that individual dendrites do their own computations, and secondly because the HPCT structure shows only one place in which sensory and motor “input” come together, and that is the comparator, where input comes from the higher level outputs in the form of a reference signal. Could this kind of dendritic computation be some element of comparator function? It’s not easy for me to see how, but if that’s not it, then there must be another place in some revised HPCT circuitry where motor and sensory inputs are used in coordination.

HB :

My personal oppinion is that PCT, if it wants to be general theory of nervous system or even general theory of how organisms work, should be able to “incorporate” all results of experiments and specific PCT testings. So it has to be able to include and explain all the experiments about organisms “structure and organization” in any HPCT organization in any language (also in physiological).

If that is not so, PCT is not a general theory but just one of specific theoires of nervous system, although it wants to be much more.

My personal oppinion is that PCT shows on the most general level right results of contol structure of organisms, what could be the confirmation of Rick’s precise results. But I also think that this results have to be synchronised also with all parts of organisms structure and organization, which maybe don’t show PCT properties, but serve the “final purpose”.

What’s the participation of dendrites in the whole “Grand reorganization” of organisms structure in PCT sense ? What’s the end goal of all proceses in organism if it is alive and how these knowledge about dendrites could be included ?

My personal oppinion is that Ashby gave correct general answer with “diagram of immediate effects” and Bill gave correct specific answer with HPCT, but incorrect answer in general sense. I think even, that Bill’s HPCT could be incorporated in Ashby’s “diagram of immediate” effects and results could be very interesting.

I think also that Rick is all the time looking on organisms working through the specific “PCT glases” and can’t see the whole picture, which Bill really briliantly tried to fuse from all kind of knowledges and theories. I have to repeat again that I admire him for clarifying so many “black holes” in human knowledge. But nobody is perfect and so I take him as the pioneer on the field of HPCT control systems which has to be improved.

And as I see debates on CSGnet Martin is trying all the time to improve PCT, but Rick tries somehow to stop him. And so “never-ending” story is going on. As I’m concerned in wrong direction.

So I think that Rick is probably right at his point of view and his testing (results) are probably right in the range of general PCT explanation. But we must consider also other perceptual researches (testing) of organisms part (“real machine”) and their results, which maybe are not expressed in PCT language, but nevertheless they can be right.

For example : pyramidal cells are real “pearl” in nervous system, mostly working on the highest levels of nervous system and probably Bill’s HPCT. Their abilities for reorganization (connection properties among thousands of them) are probably really incredible. So researches of these cells and their “jungle” branch structures on input and output and their layers could be really interesting for PCT, as Martin by my oppinion tried to show.

They can maybe clarify some details in PCT and in specific HPCT’s.

Layers of pyramidal and other nerv and glial cells could be interesting for HPCT as they resemble to “levels”, although their functions seem to be far more complex.

Best,

Boris

P.S. oh just that I don’t forget question about Carver and Scheier. What makes you think Rick, that they are “input-output thinkers” or as it’s put in “never-ending” story here on CSG-net - “S-R”. Could you show me, where in their books or their arcticles you made your “S-R” conclusion about them ???

···

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

From:
Richard Marken

To: CSGNET@LISTSERV.ILLINOIS.EDU

Sent: Sunday, February 10, 2013 7:27 AM

Subject: Re: HPTC and Tacit Knowledge

[From Rick Marken (2013.02.09.2030)]

[Martin Taylor 2013.02.09.23.35]

[From Rick Marken (2013.02.09.1500)]

    Martin Taylor (2013.02.09.12.48)--

    RM: HPCT and PCT are the same thing;

    MT: PCT and HPCT are not the same thing. HPCT is a specific hypothetical organization of perceptual control systems. An indefinite numbers of other organizations of perceptual control systems can be hypothesized that would be properly described as PCT.

RM:Yes there are other possible organizations.

MT: Good. I'm glad we agree with no need for argument this time. PCT and HPCT are different beasts. HPCT is one possible instantiation of PCT. Others are possible. I just wanted to correct the misapprehension you might have caused in some readers when you said, obviously inadvertently, that HPCT and PCT are the same thing.

RM: Sure. We agree. PCT and HPCT are completely different;-) Now I’ll just back in my armchair and think about all those other non PCT theories and see which one turns out to be best: APCT, BPCT, CPCT… Yawn

Best

Rick

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

[MartinT]

[Martin Taylor 2013.02.10,09.24]

I don't see myself as trying to improve PCT, so much as trying to

follow the implications of understanding PCT. If PCT is correct, as I think it must be, what are the implications?
As I said (I believe over on the ECACS forum) I always look first to
see whether an observation or experiment can be interpreted, or
better, predicted, within HPCT. Usually it can. Even the dendrite
study you referenced can be interpreted within HPCT, since HPCT does
require an energy system and some way of time-multiplexing what
perceptions are controlled out of the myriads that could be
controlled at any one moment.
I don’t think ANYONE understands the full implications of the HPCT
structure. And even within that structure there are lots of
questions. For example, how are patterns of higher-level outputs
converged to form reference values for the control units at the next
lower level? That question has been simmering on CSGnet ever since I
first joined 20 years ago. It is finessed in the simulations by
making the reference input a simple summation of the converging
output values from a higher level. but this cannot be the whole
answer at all levels of the hierarchy, even if it is the correct
answer at the lowest levels. If you have your feet, a bicycle, and a
car as effordances for getting you to where you want to go, an
output that produces 40% car and 60% bicycle doesn’t leave you with
any means of transport.
A related question is the role of associative memory. In B:CP Bill
suggested a possibility – that rather than reference signals being
derived directly from the outputs at higher levels, they come from
the outputs of associative memories that are addressed by the
higher-level control unit outputs. The effect is to set the
lower-level control units with reference values that would
regenerate a perception that existed when control had been
successful on a previous occasion. Such a system would produce a
“car” reference or a “bicycle” reference, but would never produce
“40% car and 60% bicycle”. I think that Bill’s suggestion has many possible ramifications that
solve not only the question of how reference values are derived from
higher-level outputs, but also addresses some other issues, such as
time-binding and context-dependent learning independent of the
reorganization system that forms part of HPCT. But one can’t study
these aspects of HPCT (they aren’t additions to HPCT) without being
able to figure out theoretical predictions that could be tested
against experiments designed to reject them. That track has hardly
been followed, even by Bill.
HPCT does not require that output functions of control units be
leaky integrators. within HPCT, output functions could be any
function at all, and could be functions that vary over time (for
example, under control of a “Gain” signal, as has been tried in some
experiments) or that have different characteristics in different
control units of levels of perception. Leaky integrators work well
in the low-level tracking studies, but they show characteristic
patterns of error when compared against real tracking data when the
target has abrupt changes of position or velocity. What form of
output function might track equally well in the global sense while
eliminating these tiny but consistent patterns of mismatch with
human performance? Just as in physics, in PCT it is the small
consistent discrepancies between theory and observation that usually
lead to big insights. Here’s a place where such an insight is
needed.
None of what I have said, nor of what I might add in the same vein,
in any way suggests deficiencies in HPCT. They are questions
within HPCT. The point I raised in my earlier message was that
HPCT isn’t the only possible structure of control systems that would
legitimately be called PCT. No matter what the science, if it is
science, it must accept the possibility that its suppositions might
need radical alteration, just as Galileo destroyed Aristotle’s
theory of motion and Einstein superseded Newton. To say that there
are other possibilities is not to say that the current set of
beliefs is wrong. The greater the variety of other possibilities to
be tried and found wanting, the stronger the likelihood that the
current beliefs are near being correct.
Martin

···

On 2013/02/10 7:32 AM, boris_upc wrote:

      And as I see

debates on CSGnet Martin is trying all the time to improve
PCT, but Rick tries somehow to stop him.

[rsmarken]

[From Rick Marken (2013.02.10.1000)]

HB :Rick I think that you are missing the point Martin tries to tell you

If I understood Martin right ….

Martin is probably trying to tell us that PCT is a general control unit (Bill’s) and one of the structural blocks that is used in specific organization of PCT units (Bill’s HPCT) which appears in specific part of nervous system (behavioral part).

Now the question, as I understand everything, is : do all parts of nervous system behave as PCT predicts and are all parts following logic of Bill’s HPCT ? Or can all parts of organism be described with “PCT tools” and different HPCT, what could mean different organizations of PCT units ?

RM: Yes, I agree that that’s pretty much what Martin is trying to say. And I agree with it; there are other possible ways for control systems to be organized than the one Bill proposed, which we are calling HPCT. All I’m trying to do is encourage people to test the existing proposal – HPCT – which was not just pulled out of thin air but is based on a lot of data and careful reasoning (documented in B:CP) before proposing all the other possibilities.

My beef with Martin is and has always been over how best to move the science of PCT forward. I think the best course is testing; Martin thinks it’s theorizing. I know I’m not going to change Martin’s mind about this; I just want to put in my two cents about it whenever I see the opportunity and Martin’s post about PCT and HPCT being different just provided that opportunity. I don’t think there is anything wrong with Martin’s approach, by the way. Indeed, it’s probably the one that most people find most interesting, not least because you can do it in the comfort of your own armchair. I just think it’s a waste of time. But then what isn’t?

HB: So it’s somehow obvious to me that Martin doesn’t doubt in PCT, but doubt in specific HPCT that make construction of organism (“real machine”).

RM: I know that Martin doesn’t doubt PCT. That’s the problem (from my perspective). I think it’s important to doubt PCT and subject those doubts to empirical test. Simply speculating about alternative ways the nervous system might be organized does not count (for me) as a worthwhile way to doubt the current theory.

BH: So it seems to me, that Martin is trying to tell us something else. That maybe there are parts of nervous system which could have some other principles or mechanisms of working, different from HPCT,

RM: Yes, I too think that’s what Martin is saying. But we already know that. I’m just saying that it makes more sense to test the existing model before going off and speculating about alternatives. If the tests show that a revision of the model is needed then that is the time to try one of these other possibilities.

HB :My personal oppinion is that PCT, if it wants to be general theory of nervous system or even general theory of how organisms work, should be able to “incorporate” all results of experiments and specific PCT testings.

RM: The problem is that research results that are obtained without an understanding that the system under study may be a control system are difficult to make sense of in terms of closed loop models. The main missing variable is, of course, the controlled variable. So we can interpret these results in terms of PCT but it is always just wild guesses (about the controlled variable(s)) unless the researchers themselves have collected the data in the context of a control model of the system under study.

BH: P.S. oh just that I don’t forget question about Carver and Scheier. What makes you think Rick, that they are “input-output thinkers” or as it’s put in “never-ending” story here on CSG-net - “S-R”. Could you show me, where in their books or their arcticles you made your “S-R” conclusion about them ???

RM: In their first book, forgot the title, that came out in 1982, they gave a wonderful description of Powers’ theory but then went on to describe research, all of which was based on the causal model. So they themselves may not think in input-output terms but their research is done in input-output terms; and such research (as I noted above) neglects the possibility that any observed relationship between independent and dependent variable is mediated by controlled variable(s). So it as great that Carver and Scheier saw the merits of PCT early on; they just ended up understanding PCT in the context of the causal model of psychology. So they talk a lot about goals and goal attainment; they just don’t do their studies in a way that recognizes the nature of a goal attaining system – a control system.

Best

Rick

···

On Sun, Feb 10, 2013 at 4:32 AM, boris_upc boris.hartman@masicom.net

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