[From Bill Powers (950216.0715 MST)]
Lash LaRue (1995.02.15 / 1010 EST)--
... the law always requires some sort of -- let me use lawyer's
Latin here -- *mens rea* and *actus reus* -- which is the "guilty"
mind and the "guilty" act; built in to the definition of an "act"
is some sort of voluntariness, so as to factor out sleepwalking,
muscle spasms, etc. And then the type of "bad mind" that is
required may vary from purposefulness, through recklessness, and on
down to negligence.
So the spectrum covers the range from "you shouldn't have wanted to do
that" to "you should have cared about the side-effects that you should
have known would occur."
I think the law is least successful where it tries to make people care
what the law says they should do (if they don't already care). For
settling disputes the law works pretty well: the litigants may not like
the ruling, but it's more or less taken for granted even by the
litigants themselves that they will comply with it. However, when it
comes to punishing people for what they have already done to keep them
from doing it again or to deter others (rather than simply trying to put
matters right) the law is next to helpless.
The whole concept of "deterrence" is flawed. Not even John Staddon can
demonstrate that punishing one rat has any striking effect on another
rat's behavior. Of course we, as human beings, are supposed to be able
to learn vicariously where a rat is limited in that ability. But at the
same time that we believe in deterrence, we praise people for their
ability to press ahead in the face of adversity and triumph over
opponents in business, sports, and war. If punishment really had a
strong deterrent effect, nobody would ever start up a small business,
there would be no such thing as professional football, and England would
be ruled from Berlin.
Also, there is a saying in the business community and elsewhere that the
only crime is getting caught. So punishment merely trains criminals to
take more pains not to be detected, if they learn anything from it. It
doesn't make them internalize the values that the punishers want them to
adopt. Punishment is carried out mainly for the satisfaction of the
punisher who wishes to express the extent to which he or she considers
another person's behavior absolutely unacceptable -- provided there is
the power to apply punishments without much fear of successful reprisal.
The most positive case one can make for punishment is that if the crime
and the punishment are clearly described in advance, and if the link is
made to seem inevitable, a rational person will probably avoid either
doing the crime or doing it in such a way as to be blamed for it.
But the only real deterrent to crime is belief in a social system that
requires a different sort of interpersonal interaction in order to work.
If a criminal comes to understand and agree with a new system concept
under which crime goes against his or her own beliefs, then of course
the criminal would be a criminal no longer.
Now all we have to do is figure out how to get people to change their
system concepts. Right after I have finished getting conventional
psychologists to change theirs, I will turn my attention to that little
problem.
···
-------------------------
The basic problem with determining intent is that you need more than one
instance of it. Basically what you have to determine is that the person
will bring about the same end again and again even though disturbances
would tend to make something different happen. If I knock over your
lemonade glass while reaching for the telephone, I can claim that it was
an accident. But if you move the lemonade glass to the other side of the
room, and the next time the phone rings I get up, cross the room, knock
over the glass, and then return to answer the phone, you may begin to
suspect that I am doing it on purpose.
There is no way to determine another person's intention without applying
disturbances and establishing a pattern of opposition to the effects of
the disturbances. In other words, the Test for the Controlled Variable
is the only valid and objective way (that I know of) to find out whether
any given effect of a person's actions is under control by that person.
Is anything like this method ever used in court to determine intent?
---------------------------------------------------------------------
John Staddon (950215.0230 GMT via Gary Cziko) --
Hello, John, it's been a while since we corresponded. Why not bite the
bullet and sign on to our discussion group? If you're overwhelmed by the
volume, you can always drop out of some other lists that are less
worthwhile!
I did not mention criminal "purpose" because the definition of
"responsibility" I discuss does not use it.
I think you should try out the PCT concept of purpose (without quotes).
A reference signal does everything a purpose is supposed to do: it
defines a state of affairs that does not yet exist, and (through the
associated control system) produces whatever behavior is required to
bring about that state of affairs, even in a variable environment.
Purpose in its common-sense meaning was rejected long ago by
behaviorists, but that was only because they couldn't think of any
plausible or scientific way to explain it. All those decisions were
taken before control system analysis became a discipline; all the old
guys thought that purpose had to entail magic or metaphysics, or some
effect of the future on the present. Now we know that this isn't true;
purpose is a perfectly physical phenomenon, and it works just as common
sense always said it did. And now we have ways of detecting it
experimentally, and making working simulations that match human
purposive behavior (in simple cases). There is no longer any reason to
reject purposiveness as a scientific concept of behavior.
Actually, my PCT-based definition of responsibility is not so far from
yours. I think we would agree that a person is probably responsible for
an outcome if that person's behavior caused the outcome. But PCT goes a
little further, in requiring that the person continue to produce the
same outcome _even if different behavior is required to do so_. In other
words, to show responsibility, we have to show that the person is
_controlling_ the outcome in question, not merely producing it as a
side-effect. For every change in the environment tending to change the
outcome, the person must alter the actions in just such a way as to
maintain the same outcome. That's an informal statement of the Test for
the Controlled Variable.
But if I roll the marble under the feet of the old lady and the
same thing happens [she slips and falls under a bus], I am held
responsible, because I am presumed to understand the causal link
between my action and the injury.
As a judgement which only has to be plausible, that is OK. But to show
that the act was in fact purposive, we would have to show that when the
old lady avoided the marble, the person followed her and rolled another
marble under her feet, and kept doing that until she fell, and then
stopped rolling marbles. If her falling under a bus is to be considered
an intended effect, we would have to show that this person always waited
until a bus was pulling up to the curb before going back to rolling
marbles. We would then have much stronger scientific evidence that the
entire outcome was purposive, intentional.
We can "hold" people responsible even if they are not in fact
responsible. As Lash Larue said above, "ignorance of the law is no
excuse," but in fact it can be a damned good _reason_ for a person's
violating a statute on page 11,779 of the Federal Record; another good
reason is that a law is, basically, whatever the last judge says it is.
It's impossible to determine objectively whether a person knew about a
law or how it would be interpreted, so Larue's comment is a purely
practical precept; without it we couldn't hold anyone responsible for
breaking any law. Since we desire to hold people responsible for the
effects of their actions, we adopt whatever precepts will permit us to
do so. Another example of purposive behavior.
By my way of conceiving the situation, the perpetrator's intention
is irrelevant; all that matters is that individuals who find
themselves in the proximity of old ladies, with marbles ready, will
not be tempted to act as the perpetrator acted, because they will
be aware of the punishment meted out to him.
How can you say that "Intention is irrelevant" and at the end of the
paragraph say that awareness is relevant? I think intention is a lot
easier to determine than awareness is. Also, just how much effect of
punishing one person is there on another person? I thought that theories
of punishment were concerned primarily with the effect on the person
being punished, relative to future behavior of that same person. A rat
who steps on a certain spot in a cage and gets shocked will avoid
stepping on that spot, but what about another rat who saw this happen?
As I recall, evidence about vicarious learning from punishment is pretty
spotty and certainly doesn't apply to every organism.
One of the main points of the exercise is to get away from the idea
that the state has any interest whatever in the defendant's mental
state. The legal issue is deterrence, not psychoanalysis.
"Psychoanalysis" is a behaviorist's cussword. In fact we can learn some
things about people's "mental" states, but I'll admit that we can't do
it by the methods currently used in court. Actually, the state has
considerable interest in whether a person will be inclined to repeat an
offense; the purported reason for sending people to jail for limited
terms is to change them into law-abiding citizens. But the recidivism
rate is probably the best available evidence that this theory does not
work when applied to most criminals. In practical terms, you are right,
but in scientific terms I think it's a mistake to reject all attempts to
understand the internal organization of a behaving system. After all,
what do you think causes behavior? The brain and muscles must have
_something_ to do with it.
----------------------------------------------------------------------
Brian Frew (950215.1946) --
Welcome to CSG-L, Brian!
I'm interested in what led Bill Powers to devlop PCT. Did he start
off by working in one of the associated systems fields and branch
out?
I began as a physicist and electronics engineer interested in human
behavior but disappointed by psychology. I thought cybernetics was the
answer, back in the early 1950s, and started learning about control
systems in order (I thought) to catch up with the cyberneticists. As it
turned out, cybernetics gave up on control theory almost immediately, so
in collaboration with R. K. Clark and later R. L. MacFarland, I started
trying to work out a theory of behavior that took feedback effects into
account. I never joined up with any of the other approaches or even knew
much about them. The nearest I ever got to advanced academia was a year
of graduate school in psychology at Northwestern, and then being the
systems engineer for Northwestern University's astronomy department in
the 1960s and early 1970s during which I gave a few one-quarter seminars
on what we now call PCT (just before my book came out in 1973). I came
into this field more from engineering than from psychology or social
science.
----------------------------------------------------------------------
Best to all,
Bill P.