responsibility; social conflict; purposive light bulbs

[From Bill Powers (950208.1640 MST)]

[Martin Taylor 950208 10:30]

     Sounds great. Is Utah snowy in winter?

For the past 10 days it's been beautiful, clear and warm. Today, after
we had breakfast 40 miles down the road, we turned around and came back.
Yes, Utah is snowy in winter. We waited a bit too long.

···

----------------------------------

But all of this means nothing
with respect to what a person's responsibilities actually, factually,
ARE.

     Since we are dealing ONLY with an agreement about the usage of a
     word, the notion of "what a person's responsibilities actually,
     factually, ARE" is kind of vacuous, don't you think?

Put the shoe on the other foot. Isn't the point that I'm arguing against
that there are, in fact, responsibilities that people "have?" For
example, you say

     But a perception or set of related perceptions exists in many of
     us, that we label "responsibility." That perception has a value.

To say it "has a value" is clear enough if we all understand that it has
a value for the person who has adopted the responsibility. The value is
the reference state of the variable for which the person has decided to
be responsible.

     Yes, you can say that the car driver "is" or "should be held"
     responsible for these things. Or you can say otherwise, that ONLY
     the controlled CEV is the driver's responsibility. It's a question
     of where you apply your value threshold to the perception in saying
     that the perception "exists."

What you or I say about the driver's responsibility is irrelevant,
because that is OUR perception of consequences of the DRIVER's action.
We have no control over those consequences, unless we decide to erect
safety barriers, take away drivers' licences of people who drive
dangerously or suicidally, and so on.

The point I am trying to make is that when we say another person "has" a
responsibility, we are either trying to describe a fact, or to prescribe
the adoption of a goal. If the former, the only meaning that makes sense
is that we are saying that the driver does in fact have control of the
event for which we say he or she has responsibility. If the driver does
not have control of the event, then saying the driver has a
responsibility for it is simply false unless we really mean to be
prescribing, saying the driver _should_ take responsibility, _should_
take control of the event.

This leaves out one possible meaning of saying that a person "has" a
responsibility: the meaning that says that responsibilities objectively
exist, independently of the person who has or should have or wants
someone to have them. This is the position of many moralists.

This is a subtle way of proposing moral rules as if they are natural
laws, imposed from outside any mere human agency. The god in the volcano
says you are responsible for providing sacrifices of virgins once each
year, whether you want to be or not. This purported imposition of
reponsibility by a superhuman agent is a ploy to keep the person
informing you of this responsibility from being blamed personally for
telling you that you must accept the goal of sacrificing virgins every
year.

Even this attempt to shift blame is revealed as an attempt to control a
person's intentions when the question arises, "What if I don't take
responsibility for sacrificing virgins?" Ah, says the High Priest, then
you will be horribly punished. Your crops will fail, you will suffer
disease, you will be characterized as a witch, and the volcano god will
command the other villagers to dismember you and burn you. And don't
look at me like that, I'm just telling you what the volcano god told me.

So my argument is that there is no reasonable way in which we can talk
about responsibility that makes it into something objective, independent
of people's intentions.
----------------------------
One last point: the perception of being a responsible person is a
different matter entirely. That is a self-concept, and people can
consider themselves to be responsible people without actually taking
responsibility for any specific thing. I am am talking here about being
responsible for specific outcomes of behavior, not about ways we
characterize ourselves in general.
-----------------------------------------------------------------------
Joel Judd (950208.0800CST)--

     With the recent discussion of respect for the nature of individual
     control systems, I get the impression that in fact perceptual
     control systems (i.e., human beings) are NOT particularly social
     beings; in fact, we would be better off having LESS contact with
     other in many cases, especially after we mature--after puberty.

An interesting point you bring up. I think we all know the advantages of
forming social systems -- every word I write in the manner and for the
purpose I write here is an example -- but nobody in this discussion has
mentioned the disadvantages before.

Living with other people is a problem of collision avoidance in
hyperspace. You not only have to avoid physically bumping into others,
but you have to avoid using means of control that create serious
disturbances of many controlled variables maintained by all others in
your vicinity. There's a conflict involved: you need the other people to
make your own control easier, yet the other people start to snap and
growl when you start encroaching on their sets of controlled variables
(and you snap and growl right back when they do it to you). The more
people you have to interact with, the greater the conflict gets.

I can see a kernel here for an extremely interesting sociological
analysis of human interactions. To resurrect an old term, it's an
approach-avoidance conflict in multiple dimensions. One way we try to
minimize it is to propose rules: here, you hold my package for a minute
and I'll go through first, then you can hand all the packages to me and
come through after me -- OK? We set up and agree on little temporary
rules like that all the time, to help each other do things while
avoiding collisions. Laws and customs are just larger and somewhat more
permanent versions of the same thing.
----------------------------------------------------------------------
Bruce Buchanan (950208.1300 EST) --

The citation from Pierre de Latil is science fiction (maybe that's why
the forward to his book was written by Isaac Asimov). Latil's great
discovery about classes of effectors is pure sophistry. I don't know why
it has greatly influenced you for 40 years, unless perhaps it triggered
off thoughts of your own that were considerably more competent than
Latil's.

I agree completely with Martin Taylor's reply.
----------------------------------------------------------------------
Best to all,

Bill P.

[Martin Taylor 950209 11:00]

Bill Powers (950208.1640 MST)

Too bad about Utah.

On responsibility, I don't think you are arguing in opposition to me at all,
but are using my posting to get at what you see as a common viewpoint. Right?

I was dealing with the point that "responsibility" is a kind of perception
one can have about someone else's behaviour, and that perception can have
any value. I pointed out that trying to get someone else to accept YOUR
value for their perception of "responsibility" for THEIR OWN behaviour
was control of a quite different perception. In your reply, you did not
even touch on either of these points, but treated at some length matters
with which I am in fairly close agreement (I hesitate to say "total" without
a closer read). Like you, I don't believe that there is any responsibility
that people "have."

However--and I suspect that this leads into quite a different thread for
future discussion--in a control system of multiple Elementary Control Units
(ECUs), "side effects" of control by one ECU are not unobservable, and can
themselves be perceived and can be the objects of control by other ECUs.
It is this possibility for potential control that leads to the assignment
(as a perception) of "responsibility" for consequences that were not the
intention of the actor.

Incidentally, without applying The Test, which is hard to do in respect of
a one-shot event, how does an external observer guess what consequence is
intended and what is a side effect? In other words, how can an external
observer correctly assign responsibility, in the sense that to be correct
is to agree with the intention of the actor?

Martin

<[Bill Leach 950210.04:39 EST(EDT)]

[Martin Taylor 950209 11:00]

However--and I suspect that this leads into quite a different thread for
future discussion--in a control system of multiple Elementary Control
Units (ECUs), "side effects" of control by one ECU are not unobservable,
and can themselves be perceived and can be the objects of control by
other ECUs. It is this possibility for potential control that leads to
the assignment (as a perception) of "responsibility" for consequences
that were not the intention of the actor.

Oh NO! I must be in trouble now!

You bring up a point that I meant to mention before. While we may not
notice "side effects" for various reasons, they are not necessarily
outside our ability to perceive them. I doubt that there is the human
alive that has not done something (with the best of intentions) that
"hurt" someone else AND realized that the unintentional "hurt" occurred.
That nebulous term "responsibility" clicks into the picture and one
"feels" a "responsibility" for the unintended harm that was caused.

Of course this "feeling" is a result of both the perception AND ones' own
internal standard or systems concepts. Thus, it is NOT as a result of a
"natural law" or any such thing. Indeed, there are certainly people that
just plain do not seem care how much hurt or damage they might do to
others but IF they have a "sense of fair play" in a fashion that most of
us in a mainstream society seem to have then yes, randomly hurting others
will disturb a controlled variable.

Incidentally, without applying The Test, which is hard to do in respect
of a one-shot event, how does an external observer guess what
consequence is intended and what is a side effect? In other words, how
can an external observer correctly assign responsibility, in the sense
that to be correct is to agree with the intention of the actor?

Obviously, the answer to that question is that they can not... at least
not with a very high degree of certainty.

If one settles for less than a high degree of certainty then upon hitting
the lamp pole, if the victims were completely unknown to the driver and
there are items of physical evidence to indicate that loss of control of
the vehicle was accidental, there is a high probability the victims, lamp
pole and vehicle were not destroyed with intent.

OTOH, if one of the victims turns up to be a "blackmailer" of the driver
then it is likely that the lamp pole was an unintended victim... etc.

-bill