[From Bill Powers (2010.06.15.0555 MDT)]
A few loose ends:
Martin Lewitt (2010.06.14.0523 MDT) –
BP: What I’m proposing is that
our intentions can have a lot more influence than any supposed impersonal
laws of evolution.
ML: That is naive. If you
are thinking we can decide what human nature shall be, you are going be
as frustrated and disappointed as Stalin or Hitler or the women’s lib
movement were.
BP: It is not naive. It is PCT. Most of the higher-level systems are
formed through reorganization, with the inherited structures defining
only what is possible, not what will be. Reorganization is, as Don
Campbell put it, “blind variation and selective retention.” We
choose what to retain on the basis of our existing organization – we
choose what diminishes error. Therefore our goals play a part in
determining how we will behave and what we will learn. Each of us can try
to persuade others to share our goals, and when that succeeds we change
how we and other people behave and what conditions all of us try to
maintain. Interactions with other people have at least as much influence
on behavior as genetics has, and possibly a great deal more. After all,
wouldn’t the concept of “individual responsibility” be rather
an illusion if genetics determined everything?
ML: You will have to do serious
culling for several generations. Women (or should I say bitches)
are as objectified and dehumanized as ever. People we merely
disagree with are hated, demonized and mocked with spitting vitriol and
it is applauded and celebrated as snarkiness. Evolution hasn’t just
left us with individual natures, but with social group natures that
aren’t very pretty.
BP: It’s not evolution, but interaction with other people and the local
environment that shapes most of our psychological (as opposed to
physiological) characteristics. “Culling” is not called for;
education is. Unfortunately, it’s difficult to find a basis for education
that everyone will agree with (thinking of Kansas boards of education).
It may take us a long time to find out how to allow each person to seek
independent goals with the least conflict, inner and outer. One of the
things I try to persuade others to believe is that we are not merely the
products of genetics, but have a say in what we will do with the hands we
are dealt.
ML: Individually, we are merely
selfish, socially, don’t be fooled by the parochial altruism, the
in-group discipline is punishing and can be viscous and towards the
outgroup we are genocidal. When it comes to abstract communal
social ideals, evolution has left us less than half formed.
Here is a recent revelation, from the journal Science:
BP: I treat “relevations” like that with great skepticism. I am
dead sure that the effects of oxytocin so boldly and confidently reported
in Science are not what I think of as facts at all. They are statistical
facts, a crippled sort of fact, meaning that by comparing many people in
one group with many people in a control group, you can just barely detect
an effect of oxytocin on the variables studied over the population, and
have no way at all of predicting which individuals would show this effect
and which would not. Those who do not show the effect, who are nearly as
numerous as those who do, remain completely unexplained. The explanations
given for the behavior of those who do show it clearly don’t apply to the
rest of the subjects, so the generalization is not “general” at
all. It’s just another grope in the dark. What good is a theory that is
wrong almost half of the time? Other than to actuaries?
ML: I was actually suggesting
that PCT might be a candidate for the “laws” that Roseberg
thought were needed. But even you have to acknowledge that trying
to infer 11 or 12 layers of control hierarchy from just a finite amount
of behavior is a poorly constrained problem.
BP: But there is a methodology for testing guesses about controlled
variables, and it did take me some 30 years of fairly careful and
principled investigation to find 11 levels, so this hierarchy isn’t just
a bunch of random guesses. However, believing in it, or not believing, is
irrelevant; I’ve always hoped there would be some research to settle
questions about the levels. At frequent intervals I advise people not to
memorize those levels.
HY: And as far as divining them
verbally, we aren’t always insightful or honest about ourselves or to
others. The problem is totally unconstrained if you are going to
insist that our intentions are unbounded by human nature. I thought
I was the one that wanted to be free, but at least, I still wanted to be
me.
BP: Who said our intentions are unbounded by human nature? I didn’t. We
can’t successfully intend to fly (without artificial aids), though we can
wish. Some intentions rule out others: if you intend to keep the car on
the road, that rules out most ways of moving the steering wheel, so
you’re not free to intend to hold the wheel any way you like at the same
time you’re choosing to keep the car on the road.
ML: PCT does well at the
demonstration tasks, and those may give insight into how higher layers of
the hierarchy also work, but I don’t think Rosenberg was wrong to compare
the complexity of predicting in intentional psychology to that of the
fitness-environment. We might be able to find something that could
be neurobiological correlates of the 11 or 12 layers of hierarchy.
I wouldn’t be surprised if that were the scale of the complexity of the
human brain. But does PCT yet have the predictive power of complex
human behavior of even something from folk psychology such as the
Myers-Briggs test? That is a pretty low bar.
BP: Yes. That’s why I don’t use that approach. I believe in data,
evidence, tests of hypotheses, challenges to theories. Every theoretical
idea I have laid claim to has been tested in the most direct ways I could
think of. Where I haven’t been able to do that (a rather large
territory), I may make proposals and suggestions but I don’t claim to
know anything.
Actually, the idea that prediction is involved in intentional psychology
comes from a misconception of how control works, probably going back as
far as Brentano. The popular model now has organisms starting with an
analysis of the environment to produce predictions of the effects of
various possible actions. Through techniques like matrix inversion, the
organism then deduces what action will be required to produce some
goal-effect in the environment (dogs, even mice, are apparently very good
at matrix inversion). It uses the same method to deduce what neural
signals will be required in order to produce those actions. When all that
is done, the deduced neural commands are generated, the actions occur,
and the desired consequence appears. These are the principles of control
that left a very expensive robot named Dante upside down in a volcanic
crater in the Aleutian Islands, waiting to be rescued by a
helicopter.
Yes, predicting in a real environment is exceedinly complex and extremely
demanding of physical and computational resources. That is why no
organism controls in that way.
BP earlier: I sense that the fly
in the ointment here is an ideology called Libertarianism. Some good
thinking has come out of that movement, but so have some really
unpleasant excesses. There’s a thread of Me First selfishness running
through it, which appeals to those who just want to get others off their
backs so they can do anything they like (such as, if I remember right,
importing defective cars from India to sell them at a great profit to any
suckers in the US who will buy them).
ML: See you are quick to resort to simple categories and stereo
types.
BP: Well, I’ve had discussions with Libertarians for about 50 years or so
(I don’t recall when that name appeared, but the mind-set has been around
for a long time: Don’t Tread On Me). I guess that could be called quick,
in the grand sweep of history.
ML: Those cars from India would
be much safer than a motorcycle and carry 4 US adults (5 in India) at
one-third the price. They’d be a bit under powerd for US tastes but
only exceeded by the 4 or 5 current US models in fuel economy at less
than an eighth of what those models cost. In my search for fuel
economy, I found myself really concerned about the safety and all-weather
comfort of the motor scooters or cycles, I sure would like the option of
one of those cars from India, even if it doesn’t meet US safety
regulations.
If you’re going to sell the cars, why do you worry about safety, fuel
economy, or comfort? Wouldn’t that cut into your profit? Apparently you
don’t think that US safety regulations mean anything, so why bother with
that at all? Most people are pretty ignorant about such details, and you
don’t have to tell them the truth, anyway. It’s up to them to make
responsible evaluations before they buy anything, so it’s not your fault
if a wheel falls off and somebody gets killed, is it? The buyer had ample
opportunity to determine the safety of the vehicle, so you can say he
chose to be killed that way.
I never said
“no” regulations. In fact, I happen to favor anti-trust
regulations and think unions should no longer be exempt. The cars
from India should be street legal, in the sense of having a
headlight, brakes, windshield wipers and the proper bumper height,
actually, I don’t think the US has done a good job on that bumper height
thing yet.
Wait a minute. You want some regulations? How do you determine which ones
to have? Is it just your personal preference, or do other people have any
say as to which ones to keep and which to discard? Is retention of a
regulation put to a vote, or do you just tell everyone which ones to
keep? If a vote determines retention (or acceptance of new regulations),
what if the vote goes against you? For example, what if the voters reject
safety regulations? You said you wanted safety regulations: would you not
sell the cars from India because you couldn’t guarantee their safety?
Would you buy one of those cars? What if a lot of other people did buy
them out of ignorance, misunderstanding, or misrepresentation – is that
their problem, or would you feel an urge to do something about
it?
Then there’s the problem of what to do when somebody simply ignores
whatever regulations are retained including those you favor. Do you just
let the market take care of that in the long run? And what business is it
of yours anyway, if that person chooses to go his own way? Would you
include any provisions for enforcing, say, an anti-trust law or a law
about automobile safety? Or would compliance be voluntary? If it’s not
voluntary, just what would “enforcement” consist of?
The test of any system concept including Libertarianism is to let natural
disturbances occur, or generate some, and see whether the system manages
to survive intact. All the Libertarian principles I have heard about fall
apart when tested in that way. There is too much assumption of a benign
human nature – actually, quite different from the one you seem to see
around you. We have laws and regulations because people are not all
benign, or smart, or well-informed, or truthful, or altruistic. Even if
we didn’t want regulations to start with, we had to adopt them because
there are just too many people who insist on going against the current
and despoiling the commons. And we had to invent and implement means of
enforcement because mere admonishment or shunning had too little effect.
Utopias are excellent reference signals, but useless if you can’t get
there from here.
As you can see from reading chapter 17/18 in BCP 1/2, I am all in favor
of personal freedom, individual responsibility, and respect for the will
of others. But I also recognize that to get such such a salubrious state
we will have to deal with real disturbances including those from people
who don’t much want such a state to exist for anyone but themselves.
Societies are the way they are for reasons, and the reasons have to do
with unresolved conflicts both inside and between individuals. We need to
learn effective ways to deal with such conflicts, but more immediately,
we need ways that work at least a little bit, which is what we have now.
The role of the life sciences is to find feasible routes from where we
are now to where we would rather be.
Best,
Bill P.