[From Rick Marken (930322.2100)]
Chuck Tucker -- I've been wrestling with you private posts all
day. Let me just respond quickly to one of your comments; it took
me some time to finally get it right (I think).
It seems to me that I can state with great confidence and you would agree
with it that statements which appear to you to transform PCT into an
S-R like formulation are disturbing to you (I can think like you so I
have an advantage).
This is a colloquial meaning of disturbance. The statements
which "appear to me to transform PCT into an S-R like formulation"
are just perceptions (or the cause of perceptions, if you like);
They are not necessarily disturbances in the sense that they tend to
push a controlled perception away from a reference. The perceptions
resulting from these statements are probably uncontrolled, in the
sense that they depend ONLY on the statements, and not on any of
my own output effects. That is, the resulting perceptions might
be described as p(t) = d(t) -- where d(t) is the statements -- but
not p(t) = d(t) + o(t). If there is no effect of the organism on a
perception VIA THE EXTERNAL ENVIRONMENT -- that is, if there is no
contribution of o(t) to a perception, then that perception CANNOT be
controlled.
I think the perceptions resulting from the statements you describe
are uncontrolled -- my outputs have no effect on them. This
just means that there is no output I can produce that will change
the way I perceive what Martin, Allan or anyone else is saying;
I just perceive what they say (or have perceptions based on what they
say). But these perceptions do cause a "disturbance" in the colloquial
sense; that is, they do make me feel uncomfortable -- like something
is amiss. I think this is because we have REFERENCES FOR UNCONTROLLED
PERCEPTIONS (Bill P. mentioned this possibility long ago). All
this means is that certain states of our perceptions (uncontrolled)
produce error that we can do nothing about; this must lead to
reorganization (trying to find variables to control that might
eliminate this error). I think this is what we are seeing a lot of
in this "information about the disturbance" discussion -- on all
sides.
But it is important to understand that, with uncontrolled perceptions,
the effect of the distrubance (technical sense) is completly "visible"
as the perception (p(t) = d(t)) -- and any discrepency between this
uncontrolled perception and a reference for its preferred value (r) is
due completely to variations in d(t); these discerpencies (r-d(t)) are
disturbances in the colloquial sense because they are unpleasant (if we
perceive these discrepencies, they are what feels "amiss") and there is
nothing we can do to reduce this error. So with uncontrolled perceptions,
disturbances in the technical sense are highly correlated with disturbances
in the colloquial sense; it's the correlation between d(t) and r-d(t).
Of course, this also means that all the information about uncontrolled
variables is represented in p(t) and maintained in the error "signal"
(there may be no explicit signal), r-p(t).
···
----
Before Chuck sent his note, my sermon today was going to be
from p. 56, paragraph 2 of "The psychology of everyday things"
by D. Norman (Basic Books, 1988). Norman is a leader in the
field of cognitive psychology and this book is about how to
design products that cognitive systems (people) can use. The
section I wanted to discuss is called " Information is in the
world". Now that I think of it, maybe I need say no more --
for now.
------
Finally, a quick note to Ken Hacker (932203) --
The problem with PCT as you describe it is that it denies the social
aspects of what is social: human INTERaction.
Could you explain how it does that? Seems like from the relationship
level on up there are perceptions to be controlled that could
reasonably be described as involving human interaction (relationships
like "husband", "father"; programs like "helping a co-worker";
principles like "do unto others"; system concepts like "society").
Remember, now, from your argument, if you are taken back by
any of my comments, it has nothing to do with what I am saying; it is
all inside of you!!
This is partially correct. Let's say that your comments are a disturb-
ance, d(t), that contributes to some perception I am controlling. The
perception (if it is controlled -- see above) MUST be a simultaneous
result of both d(t) -- you comments -- and o(t) -- some effect I have on
p(t) via the external environment; let's say that being "taken back" is
one value of the output, o(t). Since
o(t) = r - d(t)
you can generate an o(t) = "taken aback" if you generate the
appropriate level of d(t). That is, you can get me to do "taken
back" if you say the appropriate comment. But the appropriate comment
(value of d(t)) depends on r, which is, indeed, inside of me (and its
value is determined by me). So the same comment, d(t), might lead to a
"taken back" value of o(t) when my reference is set at one
level and to no response (or a more positive one) when it is
set at another. But your comment, d(t), has its influence on o(t)
via p(t) -- which has no information in it about your comment, d(t).
Best
Rick