[Martin Taylor 2016.06.10,10.55]
[From Vyv Huddy 9.06.2016 8:07]
Remember that in the quoted section, Bill is introducing a simple
control loop in the context of a tracking task. He doesn’t get into
the hierarchy of levels until later in the book. So, when you are
considering only the simple loop, all the lower levels are subsumed
into “the output function” and into the “Perceptual Input Function”.
Later, when you look more closely and discover the hierarchy, each
elementary control system that contributes to what was earlier seen
as a simple loop
has its own output function.
But the lower level output functions don't independently contribute
to the output function of the original simplified loop, because they
are parts of their own simple loops whose reference values are
determined from the level above them in the hierarchy. The output
function described in LCS III p31 includes the functioning of these
entire supporting control loops, as in this example of ringing a
doorbell, shown first as two levels and then as three.
To which we can easily add higher-level "Whys" as in the next
figure, and lower level “Hows”.
The "skin barrier" is just under the lowest boxed functions in each
figure. The nature of the output function for that level is
described separately in each output box.
The point is that the control loop consists of the control unit
shown in the first figure plus the feedback connection through the
environment of the unit. The “environment” of a control unit
contains all the levels below it. The skin surface determines only
what an outside observer can and cannot see directly, though that
barrier can now be breached by technology.
If I wanted, in a tracked task, to move the mouse with my nosethen that would mean the output function changes?
Taken as a simple
one-level control loop, yes it would. Why? Because if you break it
down into levels of control, the supporting control loops in the
environment of the “tracking” elementary control unit that use the nose have different
characteristics than the control loops using the hand. (That’s
ignoring the point that it would be very hard to see the screen if
you were moving the mouse with your nose).
Martin
···
…
VH: I
just read in LCS III p. 31 Bill says the output
function “includes the hand holding the mouse (not
the mouse itself)” which lines up well with what
you say. I guess I had thought of the controlling
of an organism to be - in animals - primarily
about the nervous system. If I wanted, in a
tracked task, to move the mouse with my nose then
that would mean the output function changes? So
the whole animal to the skin would be on the
system side of the loop?
…