🚀 go-pugleaf

[From Bill Powers (990506.1311 MDT)]

Kenny Kitzke (990505.2045 EDT)--

><If A can make B do anything that A wants, what difference in the outcome is
>produced by what B wants?>
>
>If A can, but does not, would you call it coercion?

No. If B is doing something A doesn't like, but A takes no effective action
to force B to desist, then A is not coercing B.

>  I was burning yard
>trimmings today in an open fire.  There is a zoning law and a law enforcement
>person in the Township against such burning who can make me stop (coerce me).
> Obviously, he did not from my dual definition.  I did what I wanted.

Right. This zoning person is not coercing you. Even a threat to bring the
law down on you is not coercion, because it is not a _credible_ threat. It
has never been enforced against you, so you're justified in predicting it
will not be enforced this time, either.

><B, inside, may feel either controlled or uncontrolled, but what does
>A care about that?>
>
>A does not have to care about that, but A does have to care whether B
>perceive A's coercive power.  If the zoning officer rides by, sees me
>burning, stops his car, and sees me put out the fire, he may conclude I was
>coerced.  But, what if I did not see his car stop and just decided I wanted
>to go in for lunch and did not want to leave the fire unattended.  Do we want
>to define that as coercion of B's behavior in PCT terminology?  I say no.  It
>would make PCT not only look like a weird science to others, they would think
>we are lunatics with a crazy set of defined terms.

Coercion as I define it involves the coercer being set up to apply force or
credibly threaten to apply it _whenever your behavior departs from what the
coercer wants to perceive_. If you happen to do exactly what the coercer
wants for your own reasons, and not because of the implied threat, then the
coercer sees no deviation and takes no action. But he's always ready to do
so as soon as you depart from exactly satisfying his reference conditions.
I suspect that after you've put out the fire and started eating lunch, your
doorbell will ring, and the zoning officer will be telling you that you
failed to put the fire all the way out, or that you left ashes in the
street, or something like that. The chances of your accidentally satisfying
exactly the reference condition the zoning officer wants to maintain are
very small, especially over any protracted length of time.

>Isn't Bruce Nevin, the language expert?  Can't we listen to him to gain
>counsel?  As I recall he disagreed with calling the above type example
>coercion.

We are the language experts. Coercion is whatever most people who use the
term want to mean by it. Look in the dictionary to get some idea of what
that is. The people who write dictionaries do a lot of research on how
people actually use words.
>
><And what difference does it make in any visible outcome?>
>
>I would like to propose that we agree that a third party observer can't tell
>for sure if coercion is taking place.

Oh, I think we could at least when a direct conflict was involved. Maybe
not just on causal observation, but if we measure the forces being applied,
we would discover that the coercee's output forces are being totally
cancelled out by just a fraction of the coercer's possible output forces.

In the case of threats, it would be harder, but that just means you'd have
to observe the interaction over a longer time, and perhaps interview the
particip-ants. We ask the (potential) coercee, "Do you want to be doing
that?" and the coercee answers "no". We ask "then why are you doing it?"
and the answer is "Because HE will clobber me if I quit before his shoes
are shined." And we would say, "Well, why not keep him from clobbering you,
so you don't have to do it?" and he would say, "Are you kidding? Look at
him, and look at me!" And we would ask, "Yeah, but has he ever actually
clobbered you?"  and the answer is "Why do you think I'm shining his shoes?"


>We could take an educated guess when
>we see the cookie jar broken on the floor and Freddie is getting a paddling
>from dad, coercion is taking place.  But, to know for sure, the coercer A
>would have to say, 1) if B has not done as I want, I'll force B against
>his/her will to do what I want AND coercee B would have to say, 2) because of
>the threat of force of A, I did what A wanted even though I wanted to do
>something else.  Positive answers here would confirm a coercive interaction
>between A and B and it would not matter whether an observer thought what they
>observed was or was not coercion.

Right, that's essentially what I would do.

>This would be double insanity to me.  Then, the PCT description for me
>picking up my pencil is that I am coercing it.

Maybe. But since the pencil does nothing at all unless you make it do
something, I would just say that the pencil is just a means by which your
actions can control your perceptions. I don't think we use the word
coercion unless there is at least the possibility that the coercee could do
something other than what you want it to do.

>What I thought that would be
>described as in PCT science is I was acting to control my perception of where
>the pencil should be: in my hand instead of on my desk.

That's what I would think. But when you put another live person in the
place of the pencil, you bring in the possibility that the person may want
to do something other than behave as you wish, and may resist your actions.
Then you have to choose between simply applying overwhelming physical force
to get your way (if you can), and negotiating, persuading, or just asking a
favor.

Best,

Bill P.

[From Bruce Nevin (990507.0858 EDT)]

I've stayed out of this because I have nothing to say that is grounded in a
model.

This discussion is mired in the fallacy of (unintended) equivocation.
Equivocation is when a discussion slithers among different meanings of a
word, so that conclusions that are warranted for one meaning are assumed to
apply to other meanings.

Here's one technical PCT definition of "coercion": a special case of
conflict between two autonomous control systems, such that only one of them
is unsuccessful at controlling the physical variable that is the focus of
conflict, due to the superior physical strength (output capacity) of the
other being applied to that variable. (We'll get to threats in a moment.)

We have a spreadsheet simulation of this. The superior output capacity of
the coercer determines the state of the victim's input quantity qi' or the
victim's output quantity qo'. The simulation is oversimplified even for
this most basic form of coercion because when the victim loop reaches its
maximum output it sustains it indefinitely, which a living control system
does not do. There is also a demo in which the human user of the demo
experiences the frustration of being unable to control a variable. Here,
there is the opportunity to observe what the victim does and to introspect
what is going on when one plays this role. Data for the actions (cursor
movements) of the software and the user are recorded. However, no model has
been attempted that replicates these data, and no consideration has been
given to a model that behaves as the human victim does when interacting
with the demo.

This definition does bear some relation to usage as represented in
dictionaries, but it is safe to say that it is not the way most speakers of
English define it. Dictionary definitions do not reflect understandings of
PCT.

More important, usage and the dictionary definition of "coercion" are more
inclusive than this. They include the rapist, the extortionist, the
torturer in the Spanish Inquisition, the Nazi executioner, and so on. The
core disagreements in this discussion hinge on whether scenarios such as
these are or are not covered by the PCT definition of coercion. It is
suggested that they are covered if we add the words "or the threat of
force" to the definition. But just adding words is not science. If we add
these words it is no longer a PCT definition, because we haven't modelled a
threat or the response to a threat. So far as I know we have no data to
model.

Just because it is the same word, "coercion," is no guarantee that it is
the same phenomenon.

Coercion in the narrow sense technically defined above and (sort of)
modelled in PCT almost always has merely the purpose of making the threats
credible. The bully who forces open your hand to take the money is largely
an experience of childhood; the robber usually holds a gun, etc. The
physical variable that is the focus of conflict typically has not much if
anything to do with the perceptions that the coercer is demanding the
victim to control on his behalf: give me the money or I'll break your
thumb. The coercer does something that causes pain or distress, and offers
to make the pain (or the threat of it) go away in exchange for what he
wants. Thus, coercion in all but its most rudimentary sense is identical in
form to bribery, where the coercer offers to make the lack of some desired
thing go away. First, the briber evokes imaginings of the desired state of
affairs, until the victim feels the pain of lacking that; then the briber
makes the offer. Same process.

Just because they are different words, "coercion" and "bribery," is no
guarantee that they are different phenomena.

In both kinds of coercion, the bully unfolding your fist and taking the
money, or saying "give me the money in that safe or I'll break your thumb,"
or for that matter saying "you can have sex with this person if you open
your employer's safe for me," emotional reactiveness is surely involved
(anger, fear, desire), and that's something we know squat about modelling.
Though there's plenty of evidence to observe on this net.

Absolutely nothing can be firmly concluded about coercion in general on the
basis of the very narrow, specific form of coercion that has actually been
modelled. Until we have more adequate models, any generalizations about
threats and so on are so much hot air. Exchanging hot air can be fun, and
we can even learn useful things, but it's not science, and it is demeaning
to ourselves and to PCT to pretend that it is.

        Bruce Nevin