Evolution of higher levels of perceptual control

CSGnet Archive CSG2000
1

[From Chris Cherpas (2000.12.14.1010 PT)]

How did the higher levels of the human perceptual
control hierarchy evolve? One speculation is that
successively higher levels were added to the hierarchy
over the course of human evolution. So, the program
level would have been evident, say, 1,000,000 years ago,
but the principle level would not have been prevalent
until 500,000 years ago, with the system concept level
not being widespread until 100,000 years ago. (My numbers
are made up.)

While our definitions are not hard and fast at this point,
it appears that various species have the capacity for
category level (and possibly sequence level) control, but
few, if any, non-humans have program-, principle-, and
system concept perceptual control.

The heavy hammer of natural selection by survival (as a
prerequisite to reproduction) seems an inadequate basis,
for these changes, considering the cost of making such a
big brain. More likely is the operation of sexual selection,
in which differential reproduction rates can be driven by
factors that stray very far from (or even conflict with)
conferring a survival advantage. Up to a limit, sexual
selection can spiral into a runaway process that yields
novel and unique approaches to displaying and courting.

Could the male tendency to "show off" have been exaggerated
by sexual selection pressures in the relatively polygynous
organization of early human communities, to eventually lead
to humans' almost unique ability to imitate models? For example,
males who advertised their desirability as mates would be
more successful in mating -- and while other males may not
be as strong, as good at hunting, etc. as these models, those
that could imitate the displays (if not the other skills) of
"model" males, would also cash in on some of the extra mating
opportunities.

In typical "arms race" fashion, females who could better detect
such fraud, would be more likely to mix genes with the leading
males. And so on...

So one question is: does the ability to imitate a model require
system concept perceptual control? For mimicking various sounds
may not require it, but to imitate a dance might require some
ability to see one as "doing what another person (like myself)
is doing."

Speculatively,
cc

2

[From Bruce Abbott (2000.12.14.0750 EST)]

Chris Cherpas (2000.12.14.1010 PT) --

How did the higher levels of the human perceptual
control hierarchy evolve? One speculation is that
successively higher levels were added to the hierarchy
over the course of human evolution. So, the program
level would have been evident, say, 1,000,000 years ago,
but the principle level would not have been prevalent
until 500,000 years ago, with the system concept level
not being widespread until 100,000 years ago. (My numbers
are made up.)

While our definitions are not hard and fast at this point,
it appears that various species have the capacity for
category level (and possibly sequence level) control, but
few, if any, non-humans have program-, principle-, and
system concept perceptual control.

The heavy hammer of natural selection by survival (as a
prerequisite to reproduction) seems an inadequate basis,
for these changes, considering the cost of making such a
big brain. More likely is the operation of sexual selection,
in which differential reproduction rates can be driven by
factors that stray very far from (or even conflict with)
conferring a survival advantage. Up to a limit, sexual
selection can spiral into a runaway process that yields
novel and unique approaches to displaying and courting.

Could the male tendency to "show off" have been exaggerated
by sexual selection pressures in the relatively polygynous
organization of early human communities, to eventually lead
to humans' almost unique ability to imitate models? For example,
males who advertised their desirability as mates would be
more successful in mating -- and while other males may not
be as strong, as good at hunting, etc. as these models, those
that could imitate the displays (if not the other skills) of
"model" males, would also cash in on some of the extra mating
opportunities.

In typical "arms race" fashion, females who could better detect
such fraud, would be more likely to mix genes with the leading
males. And so on...

So one question is: does the ability to imitate a model require
system concept perceptual control? For mimicking various sounds
may not require it, but to imitate a dance might require some
ability to see one as "doing what another person (like myself)
is doing."

no.

ba

3

[From Bruce Gregory (2000.1215.1123)]

Bruce Nevin (2000.12.15 11:00 EST)

Bruce Abbott (2000.12.14.0750 EST)--

>Chris Cherpas (2000.12.14.1010 PT) --
>
[...] So one question is: does the ability to imitate a model require
> system concept perceptual control? For mimicking various sounds
> may not require it, but to imitate a dance might require some
> ability to see one as "doing what another person (like myself)
> is doing."

no.

Your reasoning?

I don't know his reasoning, but such imintative behavior is common
throughout the animal kingdom (learning bird songs, for example). Unless
you want to argue that systems level perceptual control is equally common,
the answer must be "no".

BG

···

At 07:52 PM 12/14/2000 -0500, Abbott, Bruce wrote:

4

[From Bruce Nevin (2000.12.15 11:00 EST)]

Bruce Abbott (2000.12.14.0750 EST)--

···

At 07:52 PM 12/14/2000 -0500, Abbott, Bruce wrote:

>Chris Cherpas (2000.12.14.1010 PT) --
>
[...] So one question is: does the ability to imitate a model require
> system concept perceptual control? For mimicking various sounds
> may not require it, but to imitate a dance might require some
> ability to see one as "doing what another person (like myself)
> is doing."

no.

Your reasoning?

         Bruce Nevin
         [still on the road -- RTP today]

5

[From Kenny Kitzke (2000.12.15.1400EST)]

<Chris Cherpas (2000.12.14.1010 PT)>

<How did the higher levels of the human perceptual
control hierarchy evolve? One speculation is that
successively higher levels were added to the hierarchy
over the course of human evolution.>

Another speculation is that the first human had the same perceptual hierarchy
available for control as a grown human has today.

<Could the male tendency to "show off" have been exaggerated
by sexual selection pressures in the relatively polygynous
organization of early human communities, to eventually lead
to humans' almost unique ability to imitate models?>

That is quite a mouthful of speculation built on imagination, assumption and
possibilities over time. I suppose it could. At just what level in the male
hierarchy would the tendency to "show off" be at, such that its sexual
selection pressure input could eventually lead and allow it to imitate models?

C'mon, are you pulling our leg again? Is this a corollary based on the sc
ientific theory of ENNEAGRAMS?

The Romantic

6

[From Chris Cherpas (2000.12.15.1230 PT)]

Bruce Gregory (2000.1215.1123)--

I don't know his reasoning, but such imintative behavior is common
throughout the animal kingdom (learning bird songs, for example). Unless
you want to argue that systems level perceptual control is equally common,
the answer must be "no".

No, I am specifically assuming that these other species do not
have perceptual control of system concepts.

I don't know of any non-humans that exhibit a natural tendency
to imitate a model, except a few species of birds (which seem
limited to mimicking certain sounds), nor am I aware of any that
have been trained to imitate.

Is "monkey see, monkey do" really true?

Bruce Nevin (2000.12.15 13:44 EST)--

Enter innateness. Distinguish unpredictably variant activities from
stereotyped patterns (such as birdsong) that change little from generation
to generation. Sparrows do not imitate wrens, but a monkey may imitate a
man, a man a monkey -- or for that matter a sparrow (albeit not in flight).

Sorry I wasn't clearer. I meant the more general skill, which
humans appear to perform spontaneously at an early age, and which
is rather easily developed to a high level of sophistication (e.g.,
as seen on Saturday Night Live).

Best regards,
cc

7

[From Bruce Gregory (2000.1215.1613)]

Chris Cherpas (2000.12.15.1230 PT)

I don't know of any non-humans that exhibit a natural tendency
to imitate a model, except a few species of birds (which seem
limited to mimicking certain sounds), nor am I aware of any that
have been trained to imitate.

It depends on what you mean by imitate. Take a look at _The Social Life of
Dogs: The Grace of Canine Company_ by Elizabeth Marshall Thomas , for
example. She provides examples of dogs adopting (mimicking) the behavior of
peers. I suspect this is not confined to canines, but is typical of most
social animals. If SNL is your standard, only humans qualify as mimics. I
have no data about the extent to which starring on SNL improves your sex life.

BG

8

[From Bruce Nevin (2000.12.15 13:44 EST)]

Bruce Gregory (2000.1215.1123)--

···

At 11:23 AM 12/15/2000 -0500, Bruce Gregory wrote:

I don't know his reasoning, but such imintative behavior is common
throughout the animal kingdom (learning bird songs, for example). Unless
you want to argue that systems level perceptual control is equally common,
the answer must be "no".

Enter innateness. Distinguish unpredictably variant activities from stereotyped patterns (such as birdsong) that change little from generation to generation. Sparrows do not imitate wrens, but a monkey may imitate a man, a man a monkey -- or for that matter a sparrow (albeit not in flight).

         Bruce Nevin

9

[From Bruce Abbott (2000.12.15.2130 EST)]

Bruce Nevin (2000.12.15 11:00 EST) --

Bruce Abbott (2000.12.14.0750 EST)

>Chris Cherpas (2000.12.14.1010 PT)
>
[...] So one question is: does the ability to imitate a model require
> system concept perceptual control? For mimicking various sounds
> may not require it, but to imitate a dance might require some
> ability to see one as "doing what another person (like myself)
> is doing."

no.

Your reasoning?

No reasoning, just my dry sense of humor. It _is_ funny, if you think about it.

Bruce A.

10

[From Bruce Gregory (2000.1216.0624)]

[From Bruce Abbott (2000.12.15.2130 EST)]

No reasoning, just my dry sense of humor. It _is_ funny, if you think
about it.

If you have to think about it, it isn't funny.

BG

11

[From Bruce Abbott (2000.12.16.0745 EST))]

Bruce Gregory (2000.1216.0624) --

Bruce Abbott (2000.12.15.2130 EST)

No reasoning, just my dry sense of humor. It _is_ funny, if you think
about it.

If you have to think about it, it isn't funny.

It isn't funny, if you don't get the joke.

Bruce A.

12

[From Bruce Gregory (2000.1216.0804)]

Bruce Abbott (2000.12.16.0745 EST))

It isn't funny, if you don't get the joke.

Apparently my wit is too dry for you.

BG

13

[From Bruce Abbott (2000.12.16.0840 EST)]

Bruce Gregory (2000.1216.0804) --

Bruce Abbott (2000.12.16.0745 EST))

It isn't funny, if you don't get the joke.

Apparently my wit is too dry for you.

I'm slower than most. You'll have to explain it to me.

Bruce A.

14

[From Bruce Gregory (2000.1216.0949)]

Bruce Abbott (2000.12.16.0840 EST)

I'm slower than most. You'll have to explain it to me.

If you explain it, it is no longer witty.

BG

15

[From Bruce Gregory (2000.1218.1541)]

Bruce Nevin (2000.12.18 14:05 EST

_Grooming, Gossip, and the Origin of Language_ by R. Dunbar

I'm familiar with the Dunbar book. One of its virtues is the emphasis
Dunbar places on man as a social animal. This dimension of human nature is
easy to loose sight of in constructing PCT-based explanations, since the
latter apply to all organisms and therefore, eo ipso, with equal force to
asocial organisms. The earlier exchanges on RTP reveal how easy it is to
build PCT models that ignore the social nature of human interactions.

BG

16

[From Bill Powers (2000.12.18.1415 MST)]

Bruce Gregory (2000.1218.1541)--

The earlier exchanges on RTP reveal how easy it is to
build PCT models that ignore the social nature of human interactions.

Is there anything about the social nature of human interactions that would
change the PCT or HPCT model? These models are inherently models of
individual organization. It seems to me that interactions have to be
modeled by modeling sets of individuals in an environment, each individual
being described by some appropriate model, and the interactions then
growing out of the properties of the individuals and their environments.

A very elementary example of this approach to social phenomena is the Crowd
program, where "social" regularities reliably appear even though they are
not properties of any individual (no individual, for example, has the goal
of forming arcs and rings). A more realistic model, which I'm not prepared
to develop yet (if ever) would have to include, in the models of the
individuals, each individual's conception of and goals for social variables
that involve other individuals. Since these conceptions would grow (through
reorganization) out of the same social interactions that they affect, the
behavior of such a model could get extremely complex. But it would not
involve a different model of the individuals, would it? I mean, while the
_content_ of various levels of perception and control would become complex,
the _nature_ of these levels as control systems in a hierarchical
relationship to other levels would presumably remain the same.

Best,

Bill P.

17

[From Bruce Gregory (2000.1218.1655)]

Bill Powers (2000.12.18.1415 MST)

A very elementary example of this approach to social phenomena is the Crowd
program, where "social" regularities reliably appear even though they are
not properties of any individual (no individual, for example, has the goal
of forming arcs and rings). A more realistic model, which I'm not prepared
to develop yet (if ever) would have to include, in the models of the
individuals, each individual's conception of and goals for social variables
that involve other individuals. Since these conceptions would grow (through
reorganization) out of the same social interactions that they affect, the
behavior of such a model could get extremely complex. But it would not
involve a different model of the individuals, would it? I mean, while the
_content_ of various levels of perception and control would become complex,
the _nature_ of these levels as control systems in a hierarchical
relationship to other levels would presumably remain the same.

I agree. My point was that existing models _do_ not incorporate goals
associated with social interactions, not that they _could_ not. Until the
data is sufficiently rich, there will be little incentive to make the
models more complex. This is perfectly understandable. In the RTP case, we
focused on models in which the teacher was controlling for a single
variable -- perceived order in the classroom. Since we had no data, there
was no way to assess how realistic such a simple model would prove to be.
Given the richness of social perceptions, some of us were inclined to
believe that the answer was "not very." But that remains to be demonstrated.

BG

18

[From Bruce Nevin (2000.12.18 14:05 EST]

Chris Cherpas (2000.12.14.1010 PT)

How did the higher levels of the human perceptual
control hierarchy evolve?

Two excellent discussions of social conditions for development of higher cognitive abilities in the evolution of early humanoids and humans:

_Lucy_, by Johannsen and Edey, especially the discussion involving Lovelace toward the end.

_Grooming, Gossip, and the Origin of Language_ by R. Dunbar

There are also these, which I have not read and so cannot judge:

_The Evolution of Culture : An Interdisciplinary View_ by R. I. M. Dunbar(Editor), et al

_Human Evolution, Language and Mind : A Psychological and Archaeological Inquiry_ by William Noble, et al

_Eve Spoke : Human Language and Human Evolution_ by Philip Lieberman

_Darwin Machines and the Nature of Knowledge_ by Henry Plotkin

Lieberman's work I know and value, I just haven't seen this book.

There is some very interesting discussion more specifically of the origins of language in

_Language and Information_ by Z. S. Harris

_A Theory of Language and Information_ by Z. S. Harris

Hope this helps.

         Bruce Nevin

···

At 10:38 AM 12/14/2000 -0800, Chris Cherpas wrote:

19

[From Rick Marken (2000.12.18.1510)]

Bruce Gregory (2000.1218.1655)--

My point was that existing models _do_ not incorporate goals associated
with social interactions, not that they _could_ not. Until the data is
sufficiently rich, there will be little incentive to make the
models more complex.

I don't think it's a "data richness" question. It's a controlled variable
question. Goals are references for controlled variables. Goals (including
those associated with social interactions) are incorporated into models
when there is evidence that people have goals for particular controlled
variables.

In the RTP case, we focused on models in which the teacher was
controlling for a single variable -- perceived order in the classroom.
Since we had no data, there was no way to assess how realistic such
a simple model would prove to be.

I think there is considerably more than "no data" on this. If teachers
consistently act to restore order to a classroom when there is a
disturbance to classroom order then that's pretty strong evidence
that the teachers are controlling a variable that can be called
"order in the classroom".

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
MindReadings.com mailto: marken@mindreadings.com
www.mindreadings.com

20

[From Bruce Gregory (2000.1218.1956)]

Rick Marken (2000.12.18.1510)

I think there is considerably more than "no data" on this. If teachers
consistently act to restore order to a classroom when there is a
disturbance to classroom order then that's pretty strong evidence
that the teachers are controlling a variable that can be called
"order in the classroom".

Stick to politics Marken, you're not cut out for this PCT stuff.

BG