[From Bill Powers (960214.1800 MST)]
Hans Blom, 960214c --
I said:
This is one reason I don't place the reorganizing system at the
top of the hierarchy, or anywhere else in the hierarchy.
I don't see the connection between my remark and yours.
I was a bit cryptic. I assume that the process of reorganization (E.
coli style) is driven by errors in systems that monitor critical
variables having to do with the life support systems.
···
-----------------------------
We have a real problem here, I think. Maybe these two types of
uncertainty reduce to one fundamental process. How can there be an
"underlying" perceptual variable y if there is no corresponding
"real world" variable x?
Easily. Think of a relative humidity detector, which receives a wet-bulb
and a dry-bulb temperature and emits a signal proportional to relative
humidity. That signal does not correspond to any one variable in the
environment; it is constructed from two observable variables, but does
not correspond to either of them.
Many human perceptions are of this nature: they are constructed from
sensory input signals, but do not correspond to any one input. In many
cases (color, smell, beauty) they do not have direct physical
counterparts.
On a loftier plane, we are talking about the basic problem of
espistemology: we experience perceptions of a world, but we have no
direct way to check to see if those perceptions have one-to-one
correspondence with entities in the world. That's the kind of
uncertainty I was talking about.
-------------------------------------
Rick suggested that you apply a disturbance and see if this
system resists its effects on progress up the gradient. It would
be interesting to know how effective the control is.
See my answer to Rick. What kind of disturbance would satisfy you?
And what kind of outcome ought to occur before you pronounce the
verdict "control"?
I should think that a constant applied force would be an acceptable kind
of disturbance. You could compare the effect of the disturbance on mean
velocity with and without the change-of-size term that you use to create
control. We can judge the degree of control by seeing how much the
effect of the disturbance on the velocity is reduced when the control is
turned on. You can use small disturbances so as not to exceed the
capacities of the putative control system.
This is a quantitative question. I'm not disputing that control may
exist in your simulation. If the effect of the disturbance is reduced by
any amount, then control technically exists. However, if the effect of
the disturbance is reduced only by, say, 0.1%, I don't think we would
consider this a very useful kind of control system.
This is why I asked you to compare the mean velocity with the up-
gradient component of velocity. In E. coli, the up-gradient component
averages 50% to 70% of the swimming speed, so while the control is not
spectacularly good, it is quite detectable. In the Brownian movement
simulation, you should be able to compute the mean speed of the particle
and the component of that speed up the gradient, and come up with a
relevant number.
----------------------------------
P.S. Hans, when you reply to one of my posts, I get two copies: one sent
directly to my e-mail address, and the other via CSGnet.
-----------------------------------------------------------------------
Only brief replies today. I'm happily exhausted from getting the
observing platform joists in place.
-----------------------------------------------------------------------
Best,
Bill P.