[Mary Powers 9303.28]
Ken Hacker
The problem I see with with the external feedback loop between
people interacting is that what looks like one loop is really
two.
Here is how I visualize interaction in your terms:
feedfoward
Control --------------> Control
feedback
System 1 <-------------- System 2
I think of it like this
Control .-.-.-.-.-.-.-.> Control
System 1 <-.-.-.-.-.-.-.- System 2
which dissects out as
output
Control ---------------> Control
input
<--------------
output
System 1 <.............. System 2
input
...............>
System 1 outputs and inputs being dashes, and System 2 being
dots.
The point is that the first diagram makes it look as though 1 &
2, in conversation or whatever, are a single system. I think they
are two seperate systems using the same external events in two
different ways - much like the rubbwr band demo, in which both
parties are using the same rubber band, one to keep the knot on
the dot, and the other to move the first one's hand to a mark.
Each system should also have a loop like this:
CS 1 ---*----------->
···
*
*
*
<--*-----------
Now when CS 1 speaks, his own output is fed back, which is how he
knows what he's saying. He also perceives the answer from CS 2,
which is also feedback from what he's saying. But CS 1's feedback
is also CS 2's output. This makes it look like CS 2 is "giving"
feedback to CS 1. I think not. CS 2's output is input to CS 1,
and part of that input is CS 1's own feedback, a consequence of
his own actions.
Feedforward has nothing to do with it.
I agree with you and Wiener that social interaction is indeed
very interesting. I also think it can be dealt with in PCT terms,
using the PCT model of individuals.
I'm glad Klaus Krippendorf thinks control theory explains
something. But does he understand that movements are the ultimate
means by which control systems at every level control perception?
Mary Powers