[From Rick Marken (930617.0900)]
Hans Blom (930617) --
You can call the flock a social control system. No single
individual has a notion of what a flock is: it just controls
its own movements. Yet a (moving but stable) flock emerges.
The flock itself is not controlling anything; the behaviors of the
flock are the result of the controlling done by the individual
control systems in the flock. The variable that you call "flockness"
has to be better defined. If you mean the average distance between
all individuals in the flock, then this is NOT an emergent (uncontrolled)
variable; it is a controlled variable; a "piece" of this variable is being
controlled by each individual in the flock. The Test you propose would
show that "flockness" (my definition) IS a result of the controlling done
by each individual.
You can call the flock a social control system. No single
individual has a notion of what a flock is: it just controls
its own movements. Yet a (moving but stable) flock emerges.
"Flockness", defined as net inter-individual distance, is a
controlled variable -- controlled, in part by each member of
the flock. It is not "emergent" in my sense of the word. One
variable that IS emergent, by my defintion, is the SHAPE of the
flock as seen by an observer. If each member of the flock has
about the same reference for the distance between itself and
others the default shape of the flock will be a sphere (assuming the
individuals are not moving). If you put the flock in a box that is
smaller than the default sphere I predict that the flock will assume
the shape of the box -- with the individuals in the corners being
unable to keep the distance between themselves and their neighbors
at their reference level. What will NOT happen (and what shows that
the shape of the flock is NOT controlled) is an adjustment of the
inter-individual references so that the sphere shape is preserved.
If the emergent sphere were really a variable controlled by a "social"
control system, then that control system would adjust the references
of the individuals in the flock so that the controlled variable (shape)
was preseved in the face of the box disturbance; the result would be a
smaller sphere that fits in the box. Of course, you would still have to
show that "shape" was not a consequence of the controlling done by each
individual. And you would also have to identify the sensor and output
functions of the "social" control system, if you conclude that shape is
a variable that is NOT the result of the controlling done by the individuals
in the collective.
Martin Taylor (930616 19:30) --
I think this discussion of social control might be helped by an insightful
quote from an old posting by Bruce Nevin...
The problem with Bruce's analogy is that he is not clear about what
variables are controlled.
A statement like:
2. A cell cannot control another cell with these mechanisms.
is not clear. A cell can only control VARIABLES -- and there certainly
may be variable aspects of another cell (those not under control by that
cell) that can be controlled by "these [intracellular control] mechanisms".
The rest of Bruce's statement describes something like the PCT model
of the reorganization of a cellular control system.
Jeff Hunter (930617)--
Well here's an example of a collective reference signal
for you. If a number of women are confined together their
menstrual cycles will synchronize. This does not require the
conscious knowledge of the women, it seems to happen just by
pheromones.
This seems to be a pretty clear example of a "society"
setting the reference signals for individual humans, and it
hopefully will be a bit less of a " 'tis so, 'taint so" issue.
Assuming that this is a real phenomenon (I'm prepared to believe
that it is) it is not necessarily an example of a "society"
setting reference signals for individuals. This would imply that
there is a control system called "society" that, in this case, is
controlling for the relationship between the periods -- keeping
the perception of difference between cycles at zero. I think that
it is far more likely that each individual is controlling for the
relationship between her period and that of the others (unconsciously,
of course). We could test this by keeping the women together but
preventing each woman from perceiveing the periods of the others. If the
sychronicity of the periods is controlled by "society" then it should
still occur with this disturbance because the "society" control system
can still perceive all the women and can synchronize their periods. If it
doesn't occur (as I predict it won't) then the synchronized periods
result from the same processes as Hans' flocking -- individual
control systems controlling for their own perceptions. The sychronicity
is not an emergent phenomenon controlled relative to some "social"
reference level; it is a controlled variable that is controlled by the
individuals in the collective (just like Tom Bourbon's line pattern in
the cooperation experiment).
Martin Taylor (930617 09:50) correctly answered my three questions.
But he confuses me once again. In answer to:
3) Can you do it if I also give you the output function that was in effect at
the time? Yes or no?
Martin says:
Yes, if you assert that the output function includes the environmental
path between the ECS and the CEV to which d is applied, and that the
function is defined precisely and is monotonic.
This is true. But a control system that worked under these restrictions
would be working in a very contrived environment. In fact, it would
not even have to be a control system; a calculated output system
would do since the outputs that it generates are guaranteed to counter
the effects of the disturbing variable. We could call this kind of
environment an "artificial intelligence" world since most AI systems
ultimately work in this kind of environment (since they are organized
to produce preselected outputs, not inputs).
Martin goes on:
Given that the relation
between the output signal leaving the ECS and the effect of that output
on the CEV is dynamically variable, the perceptual signal also carries
information back to the output about the current relation between the
output signal and the CEV.
So you are saying that the perceptual signal carries information about the
relationship between output and CEV. Yet, in your answer to question 3)
you said that you could not reconstruct the disturbance unless you were
given information about the relationship between the output and CEV (as
part of the specification of the output function).
This is why I want to stop talking about information and start talking
about how control systems really work. I don't care if you guys want
to go on thinking that information is the greatest thing since sliced
bread and that it is the basis of control theory and all that stuff. But it
is
clear from your post that information in the perceptual signal about the
relation between output signal and CEV is of NO USE to you because
you cannot use it (as you admitted in your answer to 3) to reconstruct
the disturbance. If you can't use it, how do you know that the control
system uses it (as it presumably must in order to have effects on the
CEV that counter those of the disturding variable)?
you just don't
understand what is meant by information.
Apparently not; and somehow I feel like I just don't care. I know exactly
how control systems operate; I can build control systems that operate
in environments where the connection between output and CEV is
constantly changing and I can do it without understanding what is
meant be information. You're not getting me highely motivated to try
to understand information; it seems to me that, whatever information
is, it contributes more to mis-understanding than to understanding the
nature of control.
To show that the information
about the disturbance was available in the perceptual signal, given
the fact of control, all you need is to show that using a given language
it is shorter to describe the output function than it is to describe
the disturbance time series (or continuos waveform).
Ah. A new way to show that there's information in something. Ok Ok
There's information in perception. What's it good for? Keeping your
descriptions shorter?
Please try to remember: the ability to reconstruct is conclusive proof
that the information about the entity is available.
All an inability to reconstruct
can demonstrate is that the available information is not complete.
Well, then you can't lose, can you. Congratulations and enjoy all the
information in those perceptual signals. May all you descriptions
be brief.
Best
Rick