Using Information

[Martin Taylor 960625 18:40]

Rick Marken (960625.1500)

I guess you'll be glad to know I'm about to go home :slight_smile:

The perceptual signal carries information about the disturbance, and if
control is good, this information can be used to reconstruct the disturbance
pretty accurately--as accurately as the perceptual signal is controlled.

Ok. The perceptual signal carries information about the disturbance. But the
control system doesn't use this information;

Let's put it this way. You always interpret "uses the information about the
disturbance" as if it means "extracts and uses independently the information
about the disturbance", which is false. So I have to say that it doesn't
"use" the information in the perceptual signal about the disturbance.
I keep "use" in the same quotes Bruce used.

In no way does the control system segregate the information about the
disturbance that is carried by the perceptual signal. But if the amount
of this information was zero, the control system wouldn't work. If you
take _that_ sense, then the control system does use the information from
the disturbance. But because you always take "use" to have a connotation
of "segregate" or "extract" or "use to reconstitute", I try to avoid the
term, and to say that "use" is not what the control system does with that
information.

So--if the perceptual signal does not carry information about the disturbance,
control is impossible. But that information is never segregated (except
perhaps in one of Hans Blom's adaptive systems, but that's quite another
story:-), so, taking your interpretation, I choose to say the control
system doesn't "use" it.

Martin