[From Rick Marken (960919.0900)]
Bruce Abbott:
if a control system is to stabilize a heavy gun...it must be built of
components...along with many other such properties dictated by the task
environment.
Bill Powers:
Not by the task environment; by the intentions of the people designing the
gun.
Bruce Abbott:
I know that; Simon knows that.
Then why, pray tell, didn't he say it?
Powers:
You can't know whether the system has been "properly" (I would say
"adequately") designed unless you know what job it was designed to
do. That means you have to know what the designer _wanted_ it to do.
Abbott:
Exactly Simon's point, too.
I would have found it easier to get Simon's points if he had written them
down. He wrote down so many other things I don't see why he didn't write
down his points as well.
Powers:
This whole paragraph never once refers to the person who is responsible for
all of this.
Abbott:
We're talking about things people have designed to suit some purpose or
purposes, but it is also true that "artifacts" resulting from the process of
evolution have the same quality, although in the latter case it is much more
difficult to determine to what purposes some structure has become suited via
variation and selective retention, and must remain a matter of scientific
inference. I chose the passive voice to allow for this possibility -- after
all, there is no "designer" in the latter case to refer to actively.
There is, indeed, no designer in Darwinian "selection". This means that the
"artificats" (like flying organisms) resulting from this process were _not_
designed to suit the purposes of a designer. This is quite different than
what goes on when people design things (like gun control systems) to suit
some purpose (their own). Using the passive voice implies that there is no
designer involved in the both of these processes; it suggests that active
human and passive evolutionary selection are the same. In fact, the process
involved when humans "select" a gun control system is quite different than
the process involved in Darwininan selection of a bird flight system.
Human "selection" of the gun control system involves purposeful behavior
(control); a person has references for various characteristics of this system
and acts to produce an artifact with those characteristics; if the gun
control system isn't working properly (certain characteritics of the system
are not at the person's reference level) the person will take steps (vary the
slew rate, change the gun materials, do more research, etc) to get it to work
properly.
Darwininan "selection" of the bird flight system involves passive filtering;
if, after 30 billion years of evolution, there is no bird that can fly, the
evolutionary selection process will _not_ take steps to produce one.
Darwininan "selection" doesn't _care_ what artifacts make it through the
survival filter; what gets through, gets through. Human selection (control)
does care what artifacts result from human action; what results from a
person's action (when control is good) are the perceptions (like that of a
gun control system) that the person wants.
The fact that Simon would conflate the processes of Darwininan and human
"selection" -- basically confounding passive (cause-effect) filtering with
active control -- suggests a serious lack of understanding of the nature of
purposeful behavior. It even suggests that Simon thought of purposeful
behavior as a passive (cause-effect) process. Simon may be preaching to the
choir but it's not the one in the church of PCT;-)
Abbott:
Observing external behavior doesn't tell you much about the inner
system.
Not unless you know _how_ to observe external behavior properly. As my "mind
reading" demo shows, the proper application of The Test can reveal the
perceptions that a person's inner system is controlling. In this demo,
observable external behavior alone provides no hint of what is going on
inside the subject's mind. The only way to tell what is going on inside the
subject is by monitoring the effect of disturbances (which are not visible in
the behavior of the possible controlled objects on the screen or the
subject's actions) on visible behavior.
By the way, have you thought of any conventional experiments where the
independent variable is not an actual or potential disturbance variable?
Best
Rick