[From Bill Powers (930621.0840 MDT)]
Bob Clark (930620.2015 EDT) --
From me:
I guess that as an engineer, I am pained when people cooperate
in an attempt to achieve common goals using means that will
almost certainly, in the end, frustrate the attempt.
You:
I too, Bill, am concerned when tasks are not well done -- but I
still struggle to avoid having my own "perfectionism" interfere
with moving ahead.
"Tasks not well done" does not mean the same as "Tasks doomed to
failure," which is what I was talking about. I don't mind joining
in tasks that are being done suboptimally. It's the other kind I
have a hard time getting enthusiastic about.
We're close indeed about social control systems. I've been
thinking lately about how social cooperation toward controlling
variables of mutual interest (how's that?) is actually carried
out. I was musing about a few halcyon weeks in Boot Camp (now you
have me doing it: I mean boot camp) in the Navy when I suddenly
became a 17-year-old platoon leader with about 40 people to march
around in the compound, practicing close-order drills. Boot camp
is basically boring, so we turned this into a form of
entertainment. The idea was that the people in the platoon would
do EXACTLY what I ordered, and I would suffer the consequences
when I got confused. My commands would go like this (MARCH on one
foot, execute on the next foot if possible):
FIRST column half-left, MARCH , SECOND column to-the-rear MARCH,
FOURTH column half-right MARCH ...
At this point all the columns (of 10 men each) would be parallel.
The outer columns would be moving diagonally left and right, the
second column would be heading opposite the original direction,
and the third column would be going straight on. The challenge
was to get them all back together, marching in four columns again
in parallel. This means counting off intervals, and then going
through other commands to direct the columns around the field
until on the last MARCH they were formed up again. If I goofed it
up, the guys would happily crash into each other and fall dead
laughing, or continue marching off into the distance where I had
forgotten about them. We got very good at this (we had to
practice every day) and it was a heck of a lot more fun than just
tromping around in columns. The officers used to come out to
watch.
I, of course, got a bang out of having control of all these
people and seeing them do what I wanted. But that soon became
secondary to thinking up complex series of commands that would
leave groups scattered all over the field, and then figuring out
a way of getting them back together. It was sort of like
programming (which didn't exist then); there is a finite set of
operations that have to be used to create an infinite series of
different patterns.
It was interesting, however, that the platoon members thought it
was great fun, too. We often crashed, but quite often we would go
through a really involved set of maneuvers that resolved with a
series of consecutive commands into a perfect final formation,
and a great cheer would go up: WE DID IT! The best final move was
when a series of four commands would be issued, one to each
group, without the final "MARCH." Then, at the last second, the
command "MARCH!!: would be issued, and on the next step the crash
would be averted and order be restored. Usually. After a lot of
practice.
In retrospect, it's clear that none of this would have worked if
the individuals hadn't had a clear idea of what was going on and
even predicted my intentions. After an extended set of maneuvers
involving dozens of commands, there's no way that the columns
could have remained in proper relationship just by blindly
following the orders. The individuals in the groups had to be
making fine adjustements all the time, particularly near the end,
to keep the lines straight and to be sure that the final step
wouldn't occur halfway between the positions of people in other
columns. So while I was nominally in control, I was actually
delegating detailed control to the individuals.
All this does give some notion of how a social control project
could be set up to work. The control isn't arbitrarily imposed
the way it is inside a person. It requires all the people
involved to understand most of what is going on, at least to the
extent that it affects something they can perceive and control by
themselves. It was up to me to think up the pattern and the
strategy in the marching drill, but it was up to the individuals
to make it work.
Also, the speed of control is very slow in such a group effort.
The marching cadence and the fixed meaning of the commands,
thoroughly understood by each individual and translated reliably
into shift of directional reference levels, left a minimum time
resolution of four steps, about two to three seconds depending on
the pace (one thing that the platoon and I often disagreed about;
I ended up following them on that point). The commands had to be
issued over a period of at least four counts in advance of the
actual execution, even so, to allow for processing time. "First
column tatharear HAR!" takes four steps, and is executed on the
fifth step (assuming I started on the correct foot for the
direction of pivot). Some commands like "First column column-
half-left MARCH" were too long to do in four steps and had to be
started another step in advance to be understood. At each word,
all the people had to (a) recognize that they were in the group
to which the command applied, (b) interpret exact technical
meaning of the term and turn it into a reference level for an
actual physical act, and finally (c) wait for the MARCH command
and execute the procedure already set up with a delay of one step
(if correctly given) or two steps (if issued on the wrong foot,
blush). At the same time, each individual was maintaining a
cadence in synchronism with the nearest visible people,
maintaining spacing and line straightness, and imagining the move
that would result in the correct perception when the MARCH
occurred. Furious mental activity in each person.
I'm impressed with how much practice is needed to create a
precise group control process, and how simple it has to be to
work at all.
···
--------------------------------------------------------------
Your comments on defining levels of control are noted. I'm not
satisfied with them, not because because I disagree that the
things you mention are controlled, but because I'm trying to get
beneath the surface manifestations to basic controlled variables.
When you speak of controlling "interpersonal relations", I
immediately wonder what basic capacities of perception and
control are required in order to control that sort of thing. I
see controlling interpersonal relations as an empirical
classification, but not as one that defines the underlying
control structure that makes it possible. Your category just
doesn't seem general enough. In my attempt to define levels, I
tried to define them so it didn't matter what particular things a
person was controlling. Control of relationships occurs in all
contexts, whether involving other people or not. Ditto, I hope,
for all my other levels. Also, each of my levels involves a type
of perception, independent of modality, that, when analyzed into
its components, proves to be made of elements of the next lower
level. And control of one level of perception REQUIRES varying
perceptions of the next lower level. My levels are supposed to
meet all these criteria. And they're all supposed to be context-
free.
Well, I can't spend all day at the keyboard. The rest can wait.
-------------------------------------------------------------
Oh, one last thing. I don't think any cells in the body are
merely passive participants. Kidney cells control independently
for concentrations of many substances in the blood by varying the
rate of their removal; one kidney cell contains many control
systems. Liver cells control the sensed level of glycogen;
raising the circulating glycogen has a negative feedback effect
on glycogen production. Liver cells also control for the
circulating concentration of cholesterol. If the concentration
falls below the current reference level, the liver cells begin
manufacturing it. This is why attempts to alter the cholesterol
level are so ineffective; it takes rather violent means to change
it by as much as 15 percent, and the means have undesired side-
effects. If there is a problem, it's in the reference level, not
the production machinery or the circulating level. As doctors
found out after only about 20 years of effort, trying to alter a
controlled variable like circulating thyroxin is no answer to any
problem. The body fights any attempt to alter variables it has
under control, and in the attempt can damage itself (atrophy of
the thyroid gland, for example, when thyroxin is administered
over a long period). I haven't yet seen any evidence of cells
that are purely passive. It seems safest to assume that
everything is under active control.
----------------------------------------------------------------
Greg Williams (930621) --
I don't want to get into another long war of words, but it seems
to me that "stress" is treated as if it's like a measles germ
that gets into you and causes internal problems. It seems to me
that stress is merely the condition of the system when it's
trying to resist a large disturbance. Mental stress, I think, is
largely caused by internal conflict. External events can seem to
produce it when they bring a person's conflicts strongly into
play: when the person is prevented from doing something about an
external situation by an internal conflict that prevents any
effective action from being carried out.
What seems stressful depends on what your goals are and how
effectively you can maintain control. Stressful sexual
harrassment for one person is a titillating sexual game for
another. Of course in my ethical system, you don't play mind-
games with emotional cripples anymore than you go around making
life difficult for paraplegics. But the law is a pretty blunt
instrument for enforcing ethics.
---------------------------------------------------------------
Bruce Nevin (Mon 930621 09:44:18 EDT) --
It would help if we could just focus on one issue. I'm raising a
question as to whether "contrast" is the right perceptual
variable to be thinking of, while you're continuing to miss my
question and answer other ones.
Contrast itself can be a perception that is specifically under
control, as in adjusting the contrast on a TV set (the difference
in brightness between highlights and the darkest areas). Here you
are perceiving several different inputs at the same time and
perceiving a relationship between them, right now in present
time.
When you speak of a contrast between words, you're not talking
about that kind of situation. The words aren't being perceived
simultaneously, while some higher system judges the difference
between them. Only one word is being perceived; its contrast with
a rather large set of other words is hypothetical, not actual.
The other words are not actually being perceived; there is no
basis for detecting a contrast.
I think that this usage of "contrast" is erroneous. The same
effect can be obtained from a model in which separate systems
emit signals when inputs satisfy their input criteria for
recogizing a word. You seem to think that such a perceptual
function would require an exact match, so that one and only one
set of phomeme signals would result in a perception of a given
word. There's no such requirement on an input function. Input
functions can be as sharply or broadly tuned as is required to
fit the facts.
Your comments on the grape and the elephant seem to assume that
the grape and the elephant are distinguished because they REALLY
ARE different, whereas phonemes are distinguished because they
only SEEM different. This is a spurious argument, because in
either case, the perceptions are generated out of lower-order
perceptions of sensations. The only difference between the two
situations is that for the grape and the elephant, the two
responding perceptual functions would probably NEVER respond at
the same time except under very unusual circumstances (a
distorted grape with a bit of stem hanging from it that makes the
grape look a little like an elephant). In the case of the
phonemes, more than one detector can respond to a given set of
sensations, and there can be interactions among the detectors
that cause the recent presence of one detector's signals to
affect the operation of other detectors. The detectors may accept
a rather broad range of input sensations as sufficient excuse to
produce a perceptual output signal. In no case is it necessary to
propose that the detectors are responding to something called
"contrast." The idea of contrast in this connection is a
dormitive principle: we detect the difference between words
because they contrast; a contrast is that which enables us to
detect a difference between the words.
In fact, we don't detect differences between words, we detect the
words. "Differences" exist only for a difference-detector. In
order for a difference or contrast to be detected, the things
that differ or contrast must already have been separately
discriminated. So difference or contrast has nothing to do with
the fact that one system responds maximally to one condition,
while another responds maximally to another condition. Both
conditions may actually be present at the same time, so we get
both signals. Then it's up to higher-level systems to sort them
out by context.
Am I getting across at all?
-------------------------------------------------------------
Best to all,
Bill P.