[Martin Taylor 2005.12.02.17.48
[From Bjorn Simons (2005.12.02,23:10 EAST)]
Martin Taylor 2005.12.02.10.17
That's not right at all. Every component of each feedback loop is
entirely causal. The output of the Perceptual Input Function is
determined exactly by its input, and so on all around the loop.
It's like saying that the farmers wife is causal for sowing the corn, the
winter is causal for late growing, the farmer for harvesting , the
marketing consultant for the sale ,etc., that is OK, but I think it's
wrong to say that the winter is responsible for the entirely production. I
don't think the winter is responsible for the sale neither. But maybe I am
wrong.
I fail to see the analogy. There are lots of different influences that affect the sowing of the corn, and for all of your "like"s. That's not true of the functions of a control loop. The output of each is completely determined causally by its input and by the function it embodies.
When I said: "I have the
understanding that PCT is quite independent of any causal relations", I said
it in a connection with teleology = "The study of design or purpose in
natural phenomena". I don't think it's correct to say that the control of
perceptions are dependent of Purposes in the same way the behaviorists say
behavior is dependent of stimuli.
It could hardly be, could it? Control of perception (in PCT) is a _process_ that operates on the purpose, whereas in behaviourism, the response is a consequence of the stimulus. To make the ideas at least comparable, you would have to say something like: "I don't think it's correct to say that the outputs of the control of perceptions are dependent of Purposes" and naturally, PCT wouold say you are correct, because the outputs depend jointly on the purposes (reference values) and the disturbances.
Still I think control is something else than determinism.
I suppose you could introduce magic, but I doubt that many PCT theorists would agree.
To me, determinism means that if you know the current states of all relevant variables, and the functions and processes that act on them, you can forecast their future states.
The fact that chaotic systems diverge over time doesn't mean they are non-deterministic. It just means that if you want to forecast accurately, you have to know the initial states and the processes very accurately. They are just as determinstic as normal linear processes are.
>Even considering the loop as a whole, the system is entirely causal.
If you mean that the system of nerves is causal for human behavior, I of
course agree.
No, I'm talking about the canonical PCT loop.
My question is: "Shall we be careful within PCT and not express what an
error is result of?" (cause-effect).
No. If you act as an external analyst, you can measure both the
disturbance and the reference signal (or postulate them), and you can
say explicitly what the error is the result of. "External" here means
"outside the loop being analyzed", not necessarily outside the body
that incorporates that loop.
I am not sure I understand what you say, and I disagree.
From here on, I think your reponse is at cross-purposes with my comments, so I won't address what you say directly. I think what didn't understand was "external analyst".
An external analyst is someone who can see the whole circuit and can make measurements on any part of it. I'm not sure whether you were involved in CSGnet when there was a thread on the different observer possibilities, but it's quite important to keep them straight when you are discussing PCT.
One important "observer" is the one in the control loop, in other words, the perceptual input function (the "observation" being the perceptual signal. That observer cannot see the disturbance or the output (or, for that matter, the reference signal).
A quite different observer is one who looks AT the control loop. This is the "external analyst" who can put measuring probes on any component (or all of them). It's the external analyst who can say things like p = P(s) = P(o + d) etc., because all those signals, p, s, o, d, e, r are observables.
If you grasp a
glass of water, you control those perceptions in one loop (the loop being
analyzed). External this loop you can control another perception (another
loop). I can count from 10 to 1 at the same time I grasp a glass of water.
Which, as you may now understand, is irrelevant to my comment.
>...in the absence of
>external influences (sensory input), the brain activity would, like
>that of any physical system, relax toward an attractor. If that's
>true, then the effects of much prior experience would eventually
>diminish below the level of quantum uncertainty. Only those
>experiences that had pushed the brain into different attractor basins
>would have a permanently retained influence.
If the brain is, after the experience in question, open to further
sensory input, those inputs could move the brain into different
attractor basins, which would have the effect of eliminating (in the
long term) the influence of the earlier experience.
I am not sure I agree.
How can you not agree? It's a question of simple physics. To say "I am not sure I agree" is like saying "I'm not sure I agree that F = ma". To make sense of "I am not sure I agree" you have to demonstrate that the brain activity would NOT approach any attractor, and so far as I am aware, there is not physically possible system for which this would be true (it would be true, for example, of balls rolling without friction on an absolutely flat surface in a vacuum).
>So, taking your second possible meaning, noting the word "always", it
is not true that "all our experiences are always represented in our
brain".
Well, I read your conclusion and I don't agree. I know some people forget
what they have remembered earlier. But if they have forgotten something, it
is no longer an experience, is it? I talk about experiences that exist, of
course.
So, you diddn't mean "always". You meant "at that moment". A language confusion!
Martin