cooperation and competition

[From Bruce Gregory (2000.05.10.1956)]

Rick Marken (2000.05.10.1500)

I hope you can see, now, that it is not necessary to avoid
disturbances that may have effects on a great many other,
likely controlled variables. To do the Test you just have
to apply disturbances to a variable you think is controlled
and watch for _lack of effect_ of the disturbance. The central
feature of the Test is monitoring the state of the hypothetical
controlled variable, _not_ the state of the outputs that are
presumably used to protect that variable from disturbance.
That's why conventional research is nearly always useless;
conventional researchers are watching (at best) disturbance-
output relationships, which don't necessarily tell you
anything about a variables that is actually under control.

Thanks for the illuminating tutorial.

BG

[From Rick Marken (2000.05.10.1800)]

Bjoern Simonsen (2000.05.10.22:40 GMT+1) --

You use the concept "intended" cooperation as apposed to
"unintended" cooperation and I am not quite sure if I
understand what you mean.

Very good questions!

Just think of cooperation as a perception, like the perception
of kicking a ball. If you intentionally produce (control) a
perception of "cooperation" or "ball kicking", then the
perception, when it occurs, is intended. If, however, you produce
the perception unintentionally (by accident) then the perception,
when it occurs, is unintended.

You have probably had the experience of both intentionally and
unintentionally producing the perception of kicking a ball. I
have intentionally kicked a ball while playing soccer; I have
unintentionally kicked a ball while walking near a soccer field
and not paying attention to where I was walking.

You have probably also had the experience of both intentionally
and unintentionally cooperating. I have intentionally cooperated
with people by kicking the soccer ball right in front of my
teammate, setting up his goal (an assist!). I have unintentionally
cooperated with people by kicking the soccer ball right in front
of an open opponent, who took it right down the field to score
for the bad guys.

The point I'm trying to make is that what we perceive as
"cooperation" could be intentionally produced (by all parties
to that perception), as it is in the case of soccer plays,
industrial production, marriage, etc., or it could be an
unintentional side effect of control system operation
that is not aimed at controlling cooperation (as in the case
of insect societies, playing into the hands of your opponents,
two dimensional tracking, etc).

Does this help at all?

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken/

[From Bruce Nevin (2000.05.10 1310 EDT)]

Rick Marken (2000.05.10.1500) --

Bruce Nevin (2000.05.10 1310 EDT)--

... how to isolate
a single variable in a domain of naturalistic observation
like this.

There is really no problem. Organisms control thousands
of variables simultaneously but this is not a factor in
doing the Test because the Test is based on observing the
behavior of possible controlled variables, not on the
actions that presumably keep these variables under control.

Yes, we've been talking about possible controlled variables. However, to
the extent that control of one variable (angle of dance from vertical same
as angel of target from solar azimuth) might be means of controlling
another variable (other bees go to the place the dancer has just come
from), it may *appear* as though we are talking about the actions that
might be keeping one of these variables (the same-angle variable) under
control.

To recap more explicitly, there are two possible controlled variables that
we have identified:

H1. Other bees go to the place that the forager has just returned from,
after being present when the forager does the "dance" inside the hive.

H2. In the "waggle dance" the middle segment of the figure-8 pattern is
angled relative to vertical just as the direction to the target site is
angled relative to the azimuth of the sun.

H1 is fairly consistently born out by observations. Failure of some
percentage of bees to go to the indicated place is easily explained:

- Only a limited number of them can have sufficient tactile contact with
the dancer in the dark of the hive.

- Some of the rest follow bees that are perceived as going on a "beeline"
instead of foraging more or less at random (these followers are perhaps
alerted to leave the hive and look for such "leaders" by the strong
vibrations generated by the dancer's flight muscles during the dance).

- The bees after all are autonomous control systems not programmed automata.

H2 is consistently born out by observations. I'm not sure about
substituting artificial light for the sun. It would be difficult to find a
large enough enclosed space, given the distances involved. A lot of work
has been done to try to disprove the "waggle dance" angle-to-sun theory.

You have proposed

H3. The forager/dancer is not controlling a perception of the variable H1,
in the sense that the execution of the dance is genetically hard-wired, and
the loop involving the behavior of other bees is closed through
evolutionary time only.

Turning now to your instructions about isolating controlled variables in a
complex situation like this:

Rick Marken (2000.05.09.0800)--

Hypothesis !H1:

Bees ... probably [don't] dance to show other bees where
to find pollen.

Hypothesis H2:

The first step is to define the hypothetical
controlled variable as precisely as possible. In this case,
you might start by saying that when a bee "dances" it is
controlling a perception of its direction of movement relative
to, say, the sun.

Here, we begin to test !H1 (the negative of H1), on the assumption that H2
is true, and apparently on the assumption that the two hypotheses are
mutually exclusive (if one is true, the other is not):

This definition [H2 assumed true] makes the presence of other
bees irrelevant. So you can test this hypothesis by removing
bees from the hive when the dancer returns. If the dance goes
on anyway, that is one little piece of evidence that the
hypothesis about the controlled variable is correct (or, at
least, not wrong).

Am I understanding you correctly?

Other variables that have been proposed:

- Supplies of nectar in the hive
- Supplies of water in the hive
- Some variable (perhaps population size or density) that correlates with
needing a place to build another hive

It has been observed that when supply of one of these is adequate, and a
forager returns from a place few or no bees go to the place indicated by
the forager's dance; but if supply is low, they do go.

In one case they released foragers from a supply of nectar in a rowboat out
in a lake. Bees went in the indicated direction, but stopped when they got
to the water, as if recognizing that a supply of nectar is not likely to be
found over a body of water. As they moved the boat closer to shore, a few
bees found them, and greater numbers the closer they got. (Whether this was
all on the same day with successive foragers released to return to the
hive, and therefore perhaps influenced by the number of foragers all
successively reporting the same thing, is not clear in the summary I read

If the forager does not dance

Did you ever consider publishing a revision of your 1981 _Methods in
Experimental Psychology_?

        Bruce Nevin

···

At 08:00 AM 05/09/2000 -0700, Richard S. Marken wrote:

[From Rick Marken (2000.05.11.0900)]

Bruce Nevin (2000.05.10 1310 EDT)--

To recap more explicitly, there are two possible controlled
variables that we have identified:

H1. Other bees go to the place that the forager has just
returned from, after being present when the forager does
the "dance" inside the hive.

H2. In the "waggle dance" the middle segment of the
figure-8 pattern is angled relative to vertical just as the
direction to the target site is angled relative to the azimuth
of the sun.

I'm afraid this is all my fault. I've been shooting hypotheses
about controlled variables "from the hip" and doing a piss poor
job of it.

Your H1 and H2 are not hypotheses about controlled variables. H1
and H2 are descriptions of bee behavior from the point of view
of an observer. Controlled variables must be defined _from the
point of view of the behaving organism_ (bees in this case). That
is, a controlled variable must be a variable aspect of the
organism's experience. For example, in tracking tasks, the
controlled variable is the perceived distance between cursor
and target _from the point of view of the subject_; in a tone
control experiment the controlled variable is the frequency of
the tone _heard by the subject_; in a baseball catching experiment
the controlled variable is vertical and horizontal velocity of
the image of the ball relative to a reference point in the image
_on the fielder's retinas_. All of these variables are aspects of
the controller's own perceptual experience.

In the case of the bees, H1 must be translated into a variable (or
variables) that might be controlled by a forager bee. The question
is "what perception might the forager be controlling by flying off
in a direction that is nearly aligned with the center path of
the dancer's dance"? It's not easy for me to think of what this
perception might be. The forager apparently senses the direction
of the dancer's dance in terms of a series of pressures. It might
sense its own direction of flight as a visual variable: the angle
of the brightest point in its visual field relative to the center
of this field. So what the forager might be controlling (H1),
then, is some kind of perceptual relationship between a temporal
pattern of pressures and a spatial pattern of light. A far more
precise definition of this variable is needed before we can do any
serious testing to see what variable(s) the forager is controlling.

The same kind of translation of the outside observer point of
view into a "bees eye" point of view is needed to properly
formulate a hypothesis (H2) about what perceptual variable(s)
the dancer bee might be controlling.

Did you ever consider publishing a revision of your 1981
_Methods in Experimental Psychology_?

Yes. I was going to write a PCT methods book. But my publisher
friend (to whom I sent a prospectus) knew a book without a
market when she saw it; she has been politely avoiding the issue
whenever we go to the opera together. But I think she's right.
Why should her company publish a book that won't sell enough
copies to pay for even a fraction of the production costs? And,
for that matter, why should I waste my time writing a book
that will be read by perhaps 10 people and understood by 2? I'm
happy doing my thing on CSGNet (with the occasional journal
article; the "Dancer and the Dance" paper is basically the short
version of my PCT methods text). I'll leave the psychological
methods text publishing to those who know what the market wants:
a description of methods aimed at studying non-existent cause-
effect systems rather than real living control systems.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates mailto: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[From Bruce Nevin (2000.05.011 2247 EDT)]

Rick Marken (2000.05.09.1100)]

···

At 05:35 PM 05/09/2000 -0700, Richard S. Marken wrote:

I don't really think it's necessary to limit use of
the word "cooperation" to those cases where cooperation is
intended.

I'm glad we're not proposing a restricted definition of a term that is in
common usage.

        Bruce Nevin

[From Bruce Nevin (2000.05.11 2249 EDT)]

Rick Marken (2000.05.11.0900)--

···

At 08:50 AM 05/11/2000 -0700, Richard S. Marken wrote:

I've been shooting hypotheses
about controlled variables "from the hip"

I wish you wouldn't do that. It confuses people.

        Bruce Nevin

[From Rick Marken (2000.05.12.1250)]

Me:

I'm afraid this is all my fault. I've been shooting hypotheses
about controlled variables "from the hip" and doing a piss poor
job of it.

Bruce Nevin (2000.05.11 2249 EDT)

I wish you wouldn't do that. It confuses people.

I try to learn from my mistakes as I continue to hone myself
into that dear perfection which will be my corpse;-) But I hope
I have learned (and maybe taught others) something from this
mistake.

My mistake was (and has been) my failure to consistently describe
controlled variables from the point of view of the organism. This
is particularly hard to do for high level perceptual variables like
"cooperation", thus the inclination to define such variables from
the observer's perspective.

I can't really say what I am perceiving when I perceive something
like "cooperation". If it is called to my attention, I can notice
_that_ I am perceiving cooperation. But it's hard to describe exactly
_what_ I am perceiving. It's easier to describe (and, thus, test for
control of) lower level perceptions (like muscle length, optical
velocity,
etc). That's why it's probably best to start a PCT research program
by developing and testing hypotheses about the low level perceptions
organisms control before trying to test for control of high level
variables (like cooperation) which, almost certainly, can be perceived
(and, thus, controlled) only by humans anyway.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates mailto: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[From Bruce Nevin (2000.05.12 1726 EDT)]

Rick Marken (2000.05.12.1250)--

···

At 12:53 PM 05/12/2000 -0700, Richard S. Marken wrote:

high level
variables (like cooperation) ..., almost certainly, can be perceived
(and, thus, controlled) only by humans anyway.

Until we know what we're talking about no conclusions are warranted.

        Bruce Nevin

[From Rick Marken (2000.05.12.1540)]

Me:

high level variables (like cooperation) ..., almost certainly,
can be perceived (and, thus, controlled) only by humans anyway.

Bruce Nevin (2000.05.12 1726 EDT)--

Until we know what we're talking about no conclusions are
warranted.

Of course. We would have to test to see whether non-human systems
control whatever perception we are controlling when we control for
"cooperation". But I do think it's worth encouraging people to
begin the search for explanations of behavior -- even behavior
that seems very complex -- by trying to explain (model) behavior
in terms of simple, low-level controlled perceptions.

I have found that people (myself included) tend to be seduced by
complex explanations of behavior. In particular, I've run into
this "over-complexization of behavior" phenomenon with my baseball
catching program. Nearly everyone who sees it or hears about it
assumes that anticipation and prediction must be involved in
the catching process. The fact that the model, without any
prediction or anticipation, acts almost exactly like a real
fielder, is completely ignored. It _looks like_ fielders are
predicting and anticipating so they must _be_ predicting and
anticipating -- no matter what the model says.

So, yes, we should not come to conclusions about what other
organisms are controlling until we have tested to see whether
or not they are actually controlling these variables. But I
think we should also make an effort to start our speculations
about controlled variables by assuming that the behavior we see --
even if it seems very complex -- can be explained in terms of
control of the simplest possible perceptions.

This is certainly the lesson of Bill's CROWD program, where
incredibly complex looking social behavior "emerges" when
individuals organisms control just three simple perceptions.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates mailto: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[From Bruce Gregory (2000.05.13.0600)]

Rick Marken (2000.05.12.1540)

So, yes, we should not come to conclusions about what other
organisms are controlling until we have tested to see whether
or not they are actually controlling these variables. But I
think we should also make an effort to start our speculations
about controlled variables by assuming that the behavior we see --
even if it seems very complex -- can be explained in terms of
control of the simplest possible perceptions.

This is certainly the lesson of Bill's CROWD program, where
incredibly complex looking social behavior "emerges" when
individuals organisms control just three simple perceptions.

I couldn't agree more, which is why I questioned the need to assume that
people must be controling a perception of "cooperation" in many situations
that we call cooperative (including carrying couches).

BG

[From Bruce Nevin (2000.05.12 1726 EDT)]

My mother in law gave us _Charlie's World: The improbable adventures of a
Hong Kong cockatoo and his American family_ (NY: Earth Times Books, 2000)
partly because the author, Audrey Topping, is a neighbor, but mostly
because it is a delightful book. In an early chapter she describes a
childhood encounter with birds. In Edmunton (Canada) the local theater gave
kids free admission for the price of two pairs of crows' feet or a gopher's
tail. Her brother one day gave her two baby crows that he had "found" (and
shortly went off to the theater). She raised them. An excerpt:

"When I fed the crows corn-on-the-cob, I held it so they could peck off the
kernels. But when I put the cob down and hid, Jill would stand on the cob
to hold it firm for Jack to peck. Then Jack took his turn, hodling it for
Jill. The boys thought this was clever; but what really amused them was
that when I reappeared, the crows pretended they could not eat the corn by
themselves. They played helpless and opened their mouths like babies to be
fed. This was the kind of trickery the boys could identify with."

Of course, crows are extremely intelligent animals. No comparison with
bees. But there was no verbal discussion between these two birds, I think.

        Bruce Nevin

[From Rick Marken (2000.05.13.0950)]

Me:

I think we should also make an effort to start our speculations
about controlled variables by assuming that the behavior we see --
even if it seems very complex -- can be explained in terms of
control of the simplest possible perceptions.

Bruce Gregory (2000.05.13.0600)--

I couldn't agree more, which is why I questioned the need to
assume that people must be controling a perception of
"cooperation" in many situations that we call cooperative
(including carrying couches).

OK. But what lower level perceptions might the independent
couch lifters be controlling that would produce this
coordinated activity?

When I think about modeling this process, I am always led to
the realization that it must involve control of higher
level perceptions than just "location of couch". For example,
if both parties are controlling for a perception of "couch in
next room" how did both come to have this reference at the same
time? We could assume that people are always walking around with
a reference for moving couches into the next room, but that
is clearly absurd because we rarely see people trying to
move couches into the next room all by themselves. So maybe
people are walking around with references for moving couches
into the next room when they see at least one other person
in a room with a couch. But, again, this doesn't seem to fit
the observations. What we usually see, when people move
couches from one room to another, is some verbal discussion
about what is to be done. At least, this is how I recall my
couch lifting experiences. That verbal discussion, I think,
is part of the process of controlling for a perception that
could be described as "cooperation".

I think forager bees could always be walking around with
a reference for perceiving a sequence of pressures from
a dancing bee and for perceiving a particular optic flow
in relationship to those pressures. So I'm pretty sure that
the "cooperative" dancing/foraging behavior of bees will be
completely understood once we understand the low level
perceptions that different bees control. But I don't think
people are always walking around with references for having
the perceptions they have when they are doing certain
cooperative activities, like moving couches into the next
room. I think we obviously will need to learn how to study
control of higher level perceptions (like system concepts)
in order to fully understand human behavior. I just don't
think we should _start_ a PCT science by trying to study
control of those kinds of perceptual variables.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken/

[From Rick Marken (2000.05.13.1150)]

Bruce Nevin (2000.05.12 1726 EDT)--

"When I fed the crows corn-on-the-cob, I held it so they could
peck off the kernels. But when I put the cob down and hid, Jill
would stand on the cob to hold it firm for Jack to peck. Then
Jack took his turn, hodling it for Jill.

Great story! It sure looks like these birds have worked out
a cooperative eating strategy non-verbally. I would really
like to know how the birds might have figured this out. Does
the book say anything about the history of the interaction
between Jack and Jill?

Very interesting!

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken/

[From Bruce Nevin (2000.05.13 1705 EDT)]

Rick Marken (2000.05.13.1150)--

I would really
like to know how the birds might have figured this out.

So would I.

Does
the book say anything about the history of the interaction
between Jack and Jill?

No, but when we visit Sarah's mother again I might be able to meet the
author and see what she remembers from this childhood experience.

They surely had some experience of turn-taking as she fed them. It may be
that while "mothering" them she insisted on feeding them in turns when one
that had just gulped something down would push forward with the other one
for more.

I recommend the book for stories of all sorts of fascinating interactions.

        Bruce

···

At 10:54 AM 05/13/2000 -0800, Rick Marken wrote:

[From Rick Marken (2000.05.15.0840)]

Bruce Nevin (2000.05.13 1705 EDT)--

They [the crows] surely had some experience of turn-taking
as she fed them.

The crows need more than experience with turn taking in order
to be able to take turns. They need to be able to _perceive_
turn taking and act in order to maintain a particular level
of this perception. I still have my doubts about whether
crows can perceive and control "turn taking", let alone be
able to perceive a reason for doing this. I'm still not
giving up on the possibility that the crows are controlling
some simple perception (like "cob movement") that results in
the appearance of turn taking behavior.

It would be nice if there was a video of the crows doing this
turn taking. When you actually see what's going on it seems
to be easier to develop hypotheses about what perception(s)
the organism might be controlling. This is what happened in
the case of the egg rolling behavior of the Herring Gull.
Ethologists report seeing "open loop" bill movement when the
egg being moved is taken away. But when you look at a video
(http://www.pigeon.psy.tufts.edu/psych26/fap.htm) of Herring
gull egg rolling behavior through "PCT glasses" it's pretty
clear that the Gull is controlling something like "pressure
against the back of the bill". Removal of the egg suddenly
reduces this pressure. The Gull seems to be trying to bring
the pressure back to the reference level by pressing harder
against the egg. Since the egg is no longer present, what is
seen by an observer is the bill accelerating towards where
the egg used to be.

Anyway, it may be that crows do control for taking turns in
order to help each other eat; but it may be that something
a bit simpler is going on. It's tough to know until someone
does the testing.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates mailto: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[From Bruce Nevin (2000.05.15 1705 EDT)]

The following essay was written by Clark McPhail for a book by David
Miller. Chuck Tucker sent it to me last January. He tells me that much of
it is a summary of a piece they wrote in an issue of _Behavioral Science_,
which I don't have, but I assume one could go there for a more detailed
consideration.

I send it because it addresses in a general way the different kinds of
references that might be set for what we have been rather vaguely referring
to as "cooperation". As Clark says: "My extension of Powers' perception
control theory of purposive individual action to purposive collective
action turns on one fundamental assumption. For two or more individuals to
engage in collective action - either parallel actions at the same time or
different actions taken simultaneously or sequentially-they must adjust
their respective individual actions to realize similar or related reference
perceptions." He talks about ways in which people come to have "similar or
related reference perceptions."

This essay doesn't talk about animals. A disadvantage of working with
animals is that they can't tell us about what they are doing. (But verbal
reports are not necessarily reliable.) We assume that with lower orders at
least, it was Darwin's hammer that tuned the reference perceptions. This
means that we could find a way to "fool" an animal in ways that we cannot
fool humans -- for example beavers building a dam over the sounds of
running water coming from a loudspeaker. But the advertising industry is
built largely on fooling people in just this way.

End of digression. Here's Clark:

............................. begin quote .............................

Collective action and perception control theory

Scattered throughout Miller's text are several references to perception
control theory, a perspective unfamiliar to most readers. Since my
application of this perspective to collective action is one effort to
address the shortcomings of both collective behavior and collective action
explanations of collective phenomena, the author and editors have
graciously invited me to offer the following synopsis,

The decade of the brain.

About the same time that students of collective action were becoming
disillusioned with the rational calculus model of purposive action, the
U.S. Congress and President George Bush proclaimed the 1990s "the decade of
the brain." While physical anthropologists tell us the structure of the
human brain and the central nervous system has not changed over the past
30,000 years, more has been learned about the brain's structure and
processes in the last two decades than in all previous history. The
trillions of connections between the billions of neurons in our brain form
a complex neural network. The neural net processes the primitive
information received from our various sensory receptors (e.g., eyes, ears,
nose, tongue, and skin). It makes possible both the storage of that
processed information in and its retrieval from multiple memory sites
distributed throughout neural net. It conveys the electrical and chemical
signals through the central nervous system and the spinal chord to the
several hundred voluntary muscle control systems distributed throughout the
body which are responsible for almost every verbal and nonverbal action in
which we engage. Revolutionary new brain scanning and other research
techniques document these general structures and processes. The scanning
technologies provide dramatic representations of the brain areas where and
while those processes take place. These also provide evidence for what I
refer to as the images or "pictures in the mind's eye" of purposive actors.

Because human beings have always lived in constantly changing environments,
their survival has depended on continuous interplay between the sensory
receptors providing information about that environment, the brain that
processes that information and the voluntary muscle systems that adjust the
actions necessary to deal with those changes. This was no less true for our
ancestors' survival in savannas filled with carnivorous wild beasts than it
is for contemporary confrontations between public order police armed with
shields and truncheons who face anarchist protesters armed with paving
stones and Molotov cocktails. Every perception, thought, feeling or
recollection of any one of those individuals, every action they take or
adjust, is a product of the connections and interactions, the
"feed-forward" and feedback loops, in their respective neural nets.

Perception control theory.

The theory I have found most useful for making sense of the ongoing
relationships between sensory experiences, cognitive processing and
purposive adjustments of action in volatile environments is William T.
Powers' (1973; 1998) perception control theory. There is considerable
support for this theory in the contemporary neuroscience literature (e.g.,
The work of Antonio Damasio), as well as much compatibility with the
earlier theoretical writings of John Dewey, G. H. Mead, and Erving Goffman,
and the contemporary work of sociologists Peter Burke, Kent McClelland and
Charles Tilly.

The basic premise of perception control theory is that individuals act, or
adjust their actions, in order to make their current perceptions correspond
to or approximate their reference perceptions. By reference perceptions, I
mean the image(s) or picture(s) "in the mind's eye" that the individual has
retrieved from memory to serve as the goal, target or intended outcome to
be accomplished by the adjustment of his or her actions.

Perception control is not about others controlling the perceptions of the
individual by propaganda or "spin." Rather, it is about the individual
adding layers of clothing in the wintertime and peeling them off in the
summertime in order to keep his or her body temperature around 98.6
degrees. It is about adding just enough salt or pepper to one's food to
make it taste "just right" according to the standard in the mind's
"palette." It is about raising one's voice loud enough (in the mind's
"ear") to be heard by oneself and one's companion over the ambient noise on
the subway, at the concert, the rally or the cocktail party. It is about
organizing - about planning, preparing and pulling off - the party, formal
dinner, wedding, funeral or protest event that corresponds to the picture
in the organizer's head. While some reference perceptions are innate or
hard wired (e.g., body temperature), most are acquired and modified in the
course of the individual's interaction with others and his or her environment.

It is the disparity between the reference perception and one's current
perception of the situation under consideration that ordinarily leads to
immediate or eventual adjustments in actions. Those purposive adjustments
almost always enter a dynamic environment containing random disturbances or
deliberate obstacles. These bumps and blockades take their toll on the
intended results of purposive actions. Consequently, the current
perceptions of outcome frequently don't initially match the reference
perceptions. Continuing disturbances yielding continuing disparities
require continuing adjustments to realize and maintain the desired
correspondence or approximation between current perception and reference
perception. Any driver can appreciate the difficulty of maintaining the
intended direction and path of his or her vehicle's movement on a bumpy or
slippery roadway surface when there are strong and gusting crosswinds not
to mention rain, sleet or snowfall. Continuous vigilance and adjustment are
essential to survival. The causal relationships between reference
perception, purposive action, dynamic environment, and current perception
are not linear, they are continuous and recursive. Thus, perception control
theory is often labeled a "closed-loop, negative feedback" model of
purposive action.

Insert closed loop figure here

Every individual brings to any point in space and time, an accumulated and
unique history of personal experiences stored in memories. Those memories
provide the basis for the individual's definition or framing of new as well
as familiar situations. Memories also constitute the repertoire from which
the individual draws reference perceptions for purposive actions in the
immediate or more distant future. The individual then adjusts his or her
actions in order to reproduce or approximate once again what was recalled
from his or her past experiences.

Alternating individual and collective actions.

Why must a theory of collective action in temporary gatherings include a
theory of individual action? It is because the most characteristic feature
of any temporary gathering is the ongoing alternation between individual
and collective actions! Individuals interact with their companions and then
act alone, they may then act collectively with a larger numbers of others
in the gathering, then interact with their companions, and again act alone.
An adequate theory of purposive action should explain both individual and
collective action with the same set of principles.

Sources of collective action.

How are two or more individuals, each with unique personal histories stored
in memory, able to interact with one another let alone engage in more
inclusive collective action? There are at least two compatible answers
although one is a bit more complicated. First, most all of us have had
headaches and stomach aches. From experience we know that all the headaches
are not alike but they have sufficient similarities among them, and
sufficient distinctions from stomachaches, that we place the former in a
distinctive category labeled "headaches." We do the same with our
experiences of stomach aches, toothaches, backaches and even heartaches.
While no one can experience another's aches, we can "share" the symbols or
names given those different categories of experiences, those symbols make
up the language that makes communication possible about aches as well as
assembling processes, actors, actions, and other objects and events.

Second, Charles Tilly has researched and written extensively about
"repertoires" of political collective action, an idea similar to what some
earlier scholars called "collective memory." Tilly compares repertoires of
political collective action to the repertoires of melodies, chords and keys
familiar to jazz musicians. Musicians who are total strangers, but familiar
with those repertoires, can "sit in" during jam sessions and make music
together. The contemporary repertoire of political collective action
includes vigils, picket lines, rallies, processions and a slightly
different form of "sit-in."

My extension of Powers' perception control theory of purposive individual
action to purposive collective action turns on one fundamental assumption.
For two or more individuals to engage in collective action - either
parallel actions at the same time or different actions taken simultaneously
or sequentially-they must adjust their respective individual actions to
realize similar or related reference perceptions. There are three or more
ways in which such similar or related reference perceptions can be
established.

Independently.

People who have interacted a great deal with one another, who are part of
the same daily rounds, social networks, groups and cultures, are more
likely to have similar named categories of experiences stored in memory
from which they can independently draw similar reference perceptions, for
example, in many cultures applause is one appropriate way of showing
approval or enjoyment of what one has seen or heard. In such cultures a
gathering of individuals does not have to be asked or told, nor does anyone
have to consult their neighbor about when to applaud when their team scores
the game winning point, when a speaker finishes a compelling speech or when
a musician concludes a thrilling performance. Two or more members of a
gathering can and do independently call up that reference perception as the
appropriate outcome to achieve by slapping their open palms together and
doing so at more or less the same time. Hence the collective action of
applause.

Interdependently.

We all have experienced something we did not initially understand clearly.
We all have been confronted with a task we could not complete by our
actions alone. Thus we frequently require the assistance or cooperation of
one or more additional individuals. This problem can be as simple as not
hearing or fully understanding what a speaker has just said to the
gathering of which we are a part. Thus we may ask the person next to us to
repeat what the speaker said or even what it meant. Or it could be the more
complex problem of moving a sofa or other piece of large furniture up or
down a stairwell or into an adjoining room. When one person requires the
assistance of another, or when two or more people are confronted with a
mutual problem to be solved, they can interact by signifying in words and
gestures what needs to be done, and who will do what, when, where and how.
Thereby they interdependently establish the similar or related reference
perceptions in relation to which, simultaneously or sequentially, they will
adjust their respective actions to solve the problem.

Adoption from a third party.

The more complex the problem to be solved, and the more people required for
that solution, the more important it becomes to have a single source of
reference perceptions. I refer to that single source as a "third party" who
addresses similar or different but related reference perceptions for
adoption by two or more other individuals. The third party cannot stick
those reference perceptions into the brains of the other individuals in
question. They must adopt them as their own and then adjust their similar
or different but related actions in order to realize the intended outcome
established by the reference perceptions. Familiar examples of third
parties include the principal organizer for large events (protests,
weddings, funerals, reunions), the coach of an athletic team, the director
of a church choir, the commanding officer of a military or police unit, the
chief of a construction crew, etc.

It is often the case that the third party has devised a solution for a
complex problem but requires the coordinated participation and cooperation
of many others in order to realize that solution. When a large number of
others are involved, all will not be able to see or hear the collective
outcome of the individual actions they have contributed. Only the third
party can perceive that outcome, compare it to the reference perception in
his or her head, and then ask others to make the adjustments in their
actions that will yield the collective outcome that approximates or
corresponds to that reference perception. Whether the third party is an
organizer, coach, choir director, commanding officer, or construction crew
chief, his or her feedback and proposals to (and the adjustments by) the
individual demonstrators, singers, rank and file officers, or construction
workers, respectively, are essential to bring the envisioned collective
outcome to fruition.

Third party sources of reference perceptions are often necessarily
supplemented by interaction between some individuals who do not hear or
understand or perhaps are reluctant to do what the third party asks. And it
is frequently the case that the actual implementation of what the third
party asks other individuals to do assumes that they can and will
independently draw upon their individual memories for additional bits and
pieces of cultural knowledge to supplement what the third party has
requested or proposed. Thus, the three sources of similar or related
reference perceptions-independent, interdependent, or third party can
operate separately or in various combinations.

My application of Powers' perception control theory to individual and
collective action is an unfinished work in progress, a product of the
individual and collaborative contributions of my students, colleagues and
myself. Perhaps this synopsis will encourage others to undertake their own
applications of perception control theory to the purposive individual and
collective phenomena in which they are interested and to which David
Miller's text provides an excellent introduction.

Contributed by Clark McPhail
Clark McPhail, Ph.D.
Professor Emeritus, University of Illinois at Urbana-Champaign, USA
Visiting Professor of Politics, University of Reading, UK, 1999-2000
Fulbright Senior Research Scholar, UK, 1999-2000
website: www.uiuc.edu/~cmcphail

Childs Hall, Flat #11
Upper Redland Road
Reading, Berkshire, RG1 5JW
United Kingdom (tel: 0118 931 8809)
(from the USA, tel: 011 44 118 931 8809)

.............................. end quote ..............................

        Bruce Nevin

[From Bruce Nevin (2000.05.15 1358 EDT)]
Rick Marken (2000.05.15.0840)

The crows need more than experience with turn taking in order
to be able to take turns. They need to be able to _perceive_
turn taking and act in order to maintain a particular level
of this perception. I still have my doubts about whether
crows can perceive and control "turn taking", let alone be
able to perceive a reason for doing this. I'm still not
giving up on the possibility that the crows are controlling
some simple perception (like "cob movement") that results in
the appearance of turn taking behavior.

Both crows control eating corn from the cob. Cob movement disturbs eating
(especially after one side is bare and the "good" side rolls to the bottom,
being heavier). Standing on the cob to resist disturbance to eating due to
cob movement itself disturbs eating. (A corncob is smaller than a crow. The
crow standing on it cannot reach it very well with its beak.) Jack hops on
to hold the cob steady with uneaten kernels on top and tries to eat, with
limited success; Jill eats with good success. Jill is munching a beakfull.
Jack hops off to reach the corn better. The cob rolls. Why should Jill now
hop on? Jack is the hungrier one.

Companion animals groom one another. Why not other "favors"? There is clear
evolutionary benefit.

        Bruce Nevin

···

At 08:35 AM 05/15/2000 -0700, Richard S. Marken wrote:

[From Rick Marken (2000.05.15.1345)]

Bruce Nevin (2000.05.15 1358 EDT)--

Both crows control eating corn from the cob. Cob movement
disturbs eating (especially after one side is bare and the
"good" side rolls to the bottom, being heavier). Standing
on the cob to resist disturbance to eating due to cob movement
itself disturbs eating. (A corncob is smaller than a crow.
The crow standing on it cannot reach it very well with its
beak.) Jack hops on to hold the cob steady with uneaten kernels
on top and tries to eat, with limited success; Jill eats with
good success. Jill is munching a beakfull. Jack hops off to
reach the corn better. The cob rolls.

It looks like my suspicion was right; "taking turns" does not
look like a controlled perception. The birds are both
controlling for eating corn, as you say. The one standing
on the cob is not able to control for eating very well but,
as a side effect, makes it easier for the other bird to eat.

Why should Jill now hop on? Jack is the hungrier one.

I presume Jill does hop on. Maybe she does this because
Jack gets in the way when he gets off the cob or because
the cob starts rolling.

Companion animals groom one another. Why not other "favors"?
There is clear evolutionary benefit.

What I question is whether the crows can actually control a
perception as complex as "doing a favor" or "taking turns"
I don't think a crow would know a "favor" or a "turn" if it
saw one. I bet (even more now than before) that the apparent
"favor giving" and "turn taking" are an anthropomorphic
interpretation of a side effect of the fact that the crows
are simply controlling for things like pecking corn (rather
than bare cobs or other crows) and seeing kernels in front
of the beak.

I can think of ways to test this but I would like to see
(or hear) much more about this behavior first. A good PCT
ethologist would spent some time observing the crows before
formulating hypotheses about controlled variables and testing
those hypotheses.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates mailto: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

Bruce Nevin (2000.05.15 2332 EDT)

Rick Marken (2000.05.15.1345) --

It looks like my suspicion was right; "taking turns" does not
look like a controlled perception.

No conclusion is warranted. I tried to reduce the situation to as simple a
level of perceptions as I could in imagination. Sticking to that level, I
was unable to see a motivation for the one crow to stay on the cob (so he
gets off) or for the other one to jump on (holding it steady) when the
other one gets off. Yet this is what they did, according to the
description, alternating back and forth.

        Bruce Nevin

···

At 01:45 PM 05/15/2000 -0700, Richard S. Marken wrote:

[From Rick Marken (2000.05.15.0820)]

Me:

It looks like my suspicion was right; "taking turns" does not
look like a controlled perception.

Bruce Nevin (2000.05.15 2332 EDT)

No conclusion is warranted.

Right. I didn't mean that to sound like a conclusion; just a
strong suspicion.

I was unable to see a motivation for the one crow to stay
on the cob (so he gets off) or for the other one to jump on
(holding it steady) when the other one gets off.

Yes. But in PCT we try to think of perceptions that, _if
controlled_, might result in the behavior we observe, in
this case "taking turns". We don't try to "see motivations";
we try to guess at controlled perceptions.

PCT tells us that any particular observed behavior might be
a side effect of controlling some perception(s) (like the
perception of cob stability and the perception of pecking at
kernels); or it might be part of the actions used to control
these perceptions; or it might be a controlled perception
itself (the crows may be controlling for a perception of
"taking turns"). We can't tell by just looking at the behavior;
we have to do tests and try to build models to see if, when
these models control particular perceptions, their observed
behavior matches that of the system under study.

In the case of the crows, Jack and Jill, one simple test to
see if they are actually controlling for "taking turns" would
be to drop a new cob into the cage close to Jill just as Jill
is about to take her turn holding the cob for Jack. If Jill is
controlling for taking turns, the new cob will have no effect
on Jill's perception of "taking turns"; Jill will ignore the
new source of food and take her turn standing on the cob to
keep her perception of "taking turns" at its reference. My
guess, however, is that Jill isn't controlling for "taking
turns" at all and when the opportunity for free food presents
itself she will abandon Jack for the new cob faster than an
adulterous Republican can accuse a Democrat of being "immoral".

In PCT, we don't try to see (or guess at) a "motivation" for
an observed behavior. That's the conventional approach, which is
based on the assumption that any particular observed behavior
(like "taking turns") is caused by internal stimuli (motives).
In PCT, we try to understand behavior in terms of the perceptual
variables that are being controlled.

Best

Rick

···

---
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates mailto: rmarken@earthlink.net
http://home.earthlink.net/~rmarken