[Martin Taylor 2010.11.23.12.37]
[From Rick Marken (2010.11.23.0910)]
It's the existence of the controlled variable -- not the environment
event itself -- that determines whether an environmental event appears
to be an incentive, reinforcement or stimulus.
Yes.
Seeing the
environmental event itself as having the ability to incent, reinforce,
stimulate or, for that matter, afford is simply to succumb to the
behavioral illusion: the illusion that environmental events cause
behavior.
I think you misconstrue the behavioural illusion -- or if you don't, I do.
My take on it is that in order to oppose the effect of a disturbance on the controlled perceptual variable, the output must be such that after the effect of the output has been transmitted through the environmental feedback path, it closely matches the disturbance magnitude, but has the opposite sign.
In symbolic form, we can write:
p = P(qi). For the sake of argument, let's take P to be the identity function, as is often done in these exercises, which gives
p = qi
Again for simplicity of writing, without losing generality, lets take the reference value of p to be zero.
If the output is qo, we can write the effect of the output through the environmental feedback path to be E(qo), so, if control is good
E(qo) = qd approximately.
From which, qo = E^-1(qd) where E^-1(.) is the inverse of the environmental feedback function.
The observed behaviour is qo. Many different internal structures might lead to the same degree of control, and therefore to the same behaviour, qo. My concept of the behavioural illusion is the illusion that the behaviour can be used to determine important aspects of the internal brain mechanisms, whereas in fact the behaviour is determined by the fact of control and the nature of the environmental feedback pathway.
Is that different from your idea of what constitutes the behavioural illusion?
Martin
When I
