# What is a disturbance

[Martin Taylor 950531 19:20]

Bruce Nevin 950531:1013

Nice to hear from you again.

Movement is a
disturbance or it is not. If it isn't, why?

Think about the circuit that forms a scalar control system of the kind
discussed on CSG-L most of the time. What is it perceiving
and therefore controlling? The answer is that it controls the level
of the signal that results from whatever its Perceptual Input Function
does with the input variables.

A disturbance is ANYTHING that would alter the level of that perceptual
signal in the absence of an output from the control system. Movement
is not a disturbance unless the perception is of the POSITION of the
thing moved, and the movement is perceived as a change of position. One
can perceive movement, but the perception of movement itself is not the
disturbance. A change of movement could be a disturbance to a control
system that worked on the perceived movement, but the movement itself
could not.

It is quite possible to dissociate the perception of movement from the
perception of change of position, in normal people, as opposed to the
pathological condition you mentioned. If you watch a slowly rotating
disk for a while, then when it stops you can see movement without change
of position, or more dramatically, if you have a knob to adjust its speed
you can adjust it to have no movement, but definite change of position.
The latter has quite a disorienting effect the first time you experience it.

Another case--in a 1960's type disco, with strobe lighting, you can often
see people changing position substantially from flash to flash, while
perceiving them also as moving very slowly. My guess is that this mimics
the pathological condition you describe, but that's only a guess. The point
is that the motion perception system and the system that perceives change
of position are two separate systems, which don't always follow the
mathematically logical condition that one is the derivative of the other.

Anyway, back to the notion of "disturbance." A disturbance is what would
change the value of ANY perceptual signal in the absence of output from
the control system that affects that perceptual signal. Anything you
perceive could potentially be a disturbance, under appropriate conditions.
Whether to say that a changing perception is ALWAYS a disturbance, because
that perception MIGHT on some occasion be controlled, that's a matter of
linguistic convention. I lean toward restricting the term to perceptions
actually being controlled, but the boundary has to be very fuzzy, since
the negative gain of a control system can be anything from zero upward.

Maybe gain in those systems for some reason cannot be set high enough for
movement to be a disturbance.

That would be a linguistic fact, not an observable fact. In any case, "gain"
usually refers to the output function, or at least the path between
reference and CEV, rather than to the perceptual input function, which
is what you seem to be suggesting.

Undisturbed, but in rapid motion...

Martin