[From Bill Powers (961005.0830 MDT)]
Fred Nickols (961005.0745 EST)--
I also asked Kent if anyone had actually conducted tests
with two people using joysticks and their merged signals to
control cursor position. He was unaware of any. So, here's
a general question to the list: Have such experiments been
conducted? Let me explain also why I ask.
Yes, Tom Bourbon and his students have done an extended series of
experiments in which two to four subjects engaged in cooperative or
conflictive control over a common display. The same degree of predictivity
was obtained using the same basic model as in the single-person tracking
experiments. These results could be obtained with two or more people, or
with the control model substituting for one or more (up to all) of the
people, or even with a single person's left hand playing one part and his
right hand playing the other. I suggest that you get in touch with Tom
directly at
bourbon@sprynet.com
Tom's busy chasing Ed Ford all around the country right now, but he has a
laptop and will get around to responding eventually.
I read Kent's paper, too, and while I'm tempted to agree with
Bill's observation that "there is really nobody in charge of a
society, nobody to complain to about it," I'm mindful of a set
of folks who think they are in charge of it.
That's a good way of putting it -- "who think they are in charge of it." As
one who has always worked in the lower echelons of big businesses (and
universities), I have been struck by the differences between what managers
think the people in the company are doing and what those people are actually
doing. Micromanagement is generally ignored, although in reports on
progress, attention is always paid to assuring managers that their detailed
instructions have been carried out. The nice thing about control of
perceptions is that once the perception has been brought to the target
state, nobody can tell what actions were used to bring it to that state.
Of course managers do have "big picture" functions which are essential to
running a large enterprise, and the smart ones confine their efforts to that
level of organization, leaving those whose jobs concern the details to
accomplish what needs to be done in whatever way will do it best.
Re the COMDEX demonstration of collective control:
Ed Ford has designed a neat example using eight rubber bands and four
people. Connect four of the rubber bands into a circle, and attach the other
four at the four joints. Give the free end of each of the extra four rubber
bands to the four people. Tell them "Make a square." They will immediately
do so, without consultation and without much delay. You can also tell them
"Make it bigger" or "make it smaller", or "rotate it to the right by 45
degrees." They will do that, too, keeping it square. I shudder to think of
making a working model of that four-person process, especially considering
the under-specification of the goals.
Each person obviously has a reference image of a square in mind, but
considering that the others probably have different images, each person must
also adjust the specific configuration image to accomodate what the others
are doing, to accomplish the _categorical_ goal of "a square." This is
obviously collective control because no one person could make the square appear.
It's also interesting to add another rubber band and another person to
"help" in the process. Or you might have the manager (who says "make a
square") also start giving advice about _how_ to make the square, and see
how much better 
the performance is.
Best,
Bill P.
the performance is.