Ayn Rand and PCT

[From Bill Powers (940817.0610 MDT)]

I guess it helps to read the materials that other people cite, if you
want to understand where they're coming from. Locke, who has been
criticizing "control theory" (his idea of it), has made reference to Ayn
Rand, as have several others in his camp. I had a direct post from a
person on UseNet after the interchange with Sloman: same reference. And
some people on CSGnet have likewise referred to her work. So Mary got a
couple of books out, one of which is "The virtue of selfishness" (New
York: New Americal Library (Signet), 1961/1964). Listen up.

    Today, as in the past, most philosophers agree that the ultimate
    standard of ethics is _whim_ (they call it "arbitrary postulate" or
    "subjective choice" or "emotional committment") -- and the battle is
    only over the question of _whose_ whim; one's own or society's or
    the dictator's or God's. Whatever else they may disagree about,
    today's moralists agree that ethics is a _subjective_ issue and that
    the three things barred from its field are: reason -- mind --
    reality.

As a hint about what is going on, this and following quotes are from the
first chapter, called "The Objectivist Ethics." But don't scoff too
quickly: what follows gets quite interesting from the PCT point of view.

    "Value" is that which one acts to gain and/or keep. The concept
    "value" is not a primary; it presupposes an answer to the question:
    of value to _whom_ and for _what_? It presupposes an entity capable
    of acting to achieve a goal in the face of an alternative. Where no
    alternative exists, no goals and no values are possible.

Not a bad start. Acting to gain or keep a value is acting to make
something in experience become or remain what is wanted. That is goal-
seeking, which implies a system capable of acting to achieve a goal in
the face of disturbances. If there are no disturbances (ever), goal-
seeking is not required; an open-loop system can work in an absolutely
uniform and predictable world (except for the problem of
hypersensitivity to initial conditions, but that was recognized long
after this was written). To go on:

    I quote from Galt's speech [i.e. from something else Ayn Rand wrote
    -- Galt seems to have achieved the status of an alternate
    personality]: "There is only one fundamental alternative in the
    universe: existence or nonexistence -- and it pertains to a single
    class of entities: to living organisms. The existence of inanimate
    matter is unconditional, the existence of life is not: it depends on
    a specific course of action.... It is only a living organism that
    faces a constant alternative: the issue or life or death. Life is a
    process of self-sustaining and self-generated action. If an organism
    fails in that action, it dies; its chemical elements remain, but its
    life goes out of existence. It is only the concept of "Life" that
    makes the concept of "Value" possible. It is only to a living entity
    that things can be good or evil."

    [Now as Ayn Rand again] To make this point fully clear, try to
    imagine an immortal, indestructible robot which moves and acts, but
    cannot be affected by anything, which cannot be changed in any
    respect, which cannot be damaged, injured, or destroyed. Such an
    entity would not be able to have any values; it would have nothing
    to gain and lose; it could not regard anything as _for_ or _against_
    it, as serving or threatening its welfare, as fulfilling or
    frustrating its interests. It could have no interests and no goals.

    Only a _living_ entity can have goals or originate them. And it is
    only a living organism that has the capacity for self-generated,
    goal-directed action. On the _physical_ level, the functions of all

    living organisms, from the simplest to the most complex ... are
    actions generated by the organism itself and directed to a single
    goal: _life._

    An organism's life depends on two factors: the material or fuel
    which it needs from the outside, from its physical background, and
    the action of its own body, the action of using that fuel
    _properly_. What standard determines what is _proper_ in this
    context? The standard is the organism's life, or: that which is
    required for the organism's survival.

It is not, of course, the ability to produce particular "actions" that
is required, but the ability to produce whatever actions will, in the
environment of the moment, create particular _consequences_ of
importance to the organism. However, that is less important than the
main idea here: behind all the goals of the organism are values that are
built in to the very definition of life. I call them "intrinsic
reference levels." Rand is connecting the hierarchy of reference signals
or goals to the reorganizing system, explaining the particular goals
adopted by an organism in terms of the requirements for staying alive.
In PCT, of course, these requirements are not stated in the abstract,
nor are they stated exhaustively: they exist in the form of specific
reference signals that define the states of specific variables inside
the organism which have been evolved by the species as representing a
viable state. They do not suffice to keep any organism alive
indefinitely, but for us they are good enough for us to last perhaps 80
years, with luck.

Of particular interest here is the statement that only organisms seek
goals. This is precisely my conclusion: the only naturally-occurring
high-gain negative feedback control systems are living organisms, or
were built by living organisms (I presume that building a control system
is a natural occurrance).

Of even more interest is that Rand's comments on goal-seeking and life
are precisely the sentiments that Locke echoes, almost verbatim, as a
reason for rejecting control theory. This is because Locke accepts
Rand's "robot" image, as a machine devoid of sensitivity to its
surrounds, and capable only of blind and unthinking _action_. Rand's
picture of a robot seems almost to be a deliberate definition of an
open-loop system, which acts but is unaffected by the consequences of
its acts. What Rand is describing, of course, is a machine of the old
kind, the kind with neither sensors nor reference signals. But she is
quite right in saying that goal-seeking is the exclusive province of
living systems. What neither she nor Locke understood or understand,
however, is that real robots, closed-loop machines, are imitations of
human goal-seeking processes, built for the purpose of taking over from
human beings the onerous or too-exacting control tasks that human beings
would rather not, or can't, do. In the realm of modeling, control
systems are constructed to help us understand how we ourselves work, by
making explicit things that remain only implicit without such a model.
By studying "robot" designs, we clarify what a goal is and how action
can be directed to make the world match organism-specified goal states
despite interference by and and uncertainty in the environment.

More:

    An _ultimate_ value is that final goal or end to which all lesser
    goals are the means -- and it sets the standard by which all lesser
    goals are _evaluated_. An organism's life is its _standard of
    value_; that which furthers its life is _good_, that which threatens
    it is _evil_.

    Without an ultimate goal or end, there can be no lesser goals or
    means; a series of means going off into an infinite progression

    toward a nonexistent end is a metaphysical and epistemological
    impossibility. It is only an ultimate goal, an _end in itself_, that
    makes the existence of values possible. Metaphysically, _life_ is
    the only phenomenon that is an end in itself; a value gained and
    kept by a constant process of action. ...

    Now in what manner does a human being discover the concept of
    "_value_"? By what means does he first become aware of the issue of
    "good or evil" in its simplest form? By means of the physical
    sensations of _pleasure_ or _pain_. Just as sensations are the first
    step of development of a human consciousness in the realm of
    _cognition_, so they are its first step in the realm of
    _evaluation_.

    The capacity to experience pleasure or pain is innate in a man's
    body; it is a part of his _nature_, part of the kind of entity he
    _is_. He has no choice about it, and he has no choice about the
    standard that determines what will make him experience the physical
    sensation of pleasure or of pain. What is that standard? His _life_.

    The pleasure-pain mechanism in the body of man -- and in the bodies
    of all the living organisms that possess the faculty of
    consciousness -- serves as an automatic guardian of the organism's
    life. The physical sensation of pleasure is a signal indicating that
    the organism is pursuing the _right_ course of action. The physical
    sensation of pain is a warning signal of danger, indicating that the
    organism is pursuing the _wrong_ course of action, that something is
    impairing the proper function of its body, which requires action to
    correct it.

So there you have the embryo of a reorganizing system and a hierarchy of
goal-seeking systems built because of it. I find all this quite
remarkable. I'm sure I read _The Fountainhead_ in my youth (in 1943, its
copyright date, I became 17). I don't recall reading _Atlas Shrugged_
(1957) or anything later by Rand. Over the years I have carried a mostly
negative impression of her writings -- yet it's conceivable that these
passages, or others with similar meanings, stuck in my mind because they
made sense. I would have seen much less wrong with them at the age of 16
or 17 than later. So who knows? Maybe PCT has some roots in Ayn Rand.

Of course it's also possible that the basic idea, if it's right, would
occur to more than one person, both before and after Rand wrote.

What matters here is that we can see in Rand's writing the roots of
Locke's ideas about control theory and his reasons for rejecting it.
They really have nothing to do with the "bad" control theory of other
personality theorists; the idea that only living systems can have goals
goes back much farther than that. The "bad" control theorists present
their ideas as block diagrams, and perhaps that's what reminds Locke of
machinery, but they are such unworkable block diagrams that they can be
rejected on their own merits, as Locke does. And it's clear that neither
their authors nor Locke et. al. are aware that the original control
systems were models of the only natural control systems that exist,
living organisms. If that awareness had existed, the statement that only
living systems can have goals would have been seen as support for
control theory, real control theory, not an argument against it.

···

---------------------------------------------------------------------
Rand was unable to sustain the inspiration behind these passages. When
she got to the level of logic, there she stuck. Logic is not "whim" --
it is reality itself, as she sees it. It seems never to have struck her
that logic is only as good as its premises, and premises have to be
defended on some basis other than logic. The premises on which she
proceeds to develop the rest of her argument get extremely tiresome, and
are mostly imagined. Of course people who imagine the same things about
the world find her logic very impressive. To give due credit, Rand does

struggle with the conflict between selfishness and caring for others,
between "irrational" and "rational" self-interest. But her discourse
rapidly degenerates into prescriptive statements about what we _should_
do, never reaching the point of understanding from which she could see
that people _always do_ act to satisfy their own goals and nobody
else's. There is no need to tell them to do that. Her real arguments
with other people concern the goals they have chosen to pursue, which
are often very different from hers.

The conclusions one reaches using logic are more revealing than most
people would be comfortable knowing. They reveal mainly the conclusions
that one wanted to reach before ever turning on the logic machine. Logic
is part of a control process; we manipulate the premises to make the
logical deductions from them match the deductions we wish to draw. If a
person doesn't want to pay taxes, support evironmentalism, contribute to
domestic or foreign aid, or have his contractual obligations enforced,
it is always possible to find premises that will lead logically to the
conclusion that none of these things need be done. By manipulating
premises, the input to the logic machine, one can get just about any
conclusion, the output of the machine, that one desires, without ever
making a logical error. This is very useful in science, where we specify
desired conclusions -- the hypothesis to be tested -- and see whether we
can find premises -- possible observations -- that will lead logically
to the proposed conclusion. What scientists have to watch out for, of
course, is the danger of imagining the observations that serve as
premises, instead of actually observing that they are true. Under less
consciously-restricted conditions, imagination comes freely into play,
so that people can believe whatever they want to believe, and prove it
is true.

So Ayn Rand believes that there are only three kinds of people: Atilla,
who gets what he wants by brute force, the Witch Doctor, who gets what
he wants by denying reality, and Ayn Rand, who uses objective logic to
see the true state of affairs. In Ayn's world, all issues come down to
opposites. Either you are an altruist, who acts only for the benefits of
others and considers acting for his own benefit as evil, or you are
"selfish", meaning that you act according to your own rational self-
interest and sacrifice nothing for others at all. You are all one thing
or all another thing, and there is nothing between them. If you aren't
for me, you're against me; if you're not my friend, you're my enemy.
It's all very simple, by the time Ayn Rand gets through with it.

To Ayn Rand, who is stuck at the level of logic, there is no system
concept of a world in which it is in one's own self-interest to see to
it that everyone is satisfied, including one's self. But that takes us
in other directions, and off the subject here.
-----------------------------------------------------------------------
Best to all,

Bill P.

"William T. Powers" writes:

(quoting Ayn Rand - )

   Without an ultimate goal or end, there can be no lesser goals or
   means; a series of means going off into an infinite progression
   toward a nonexistent end is a metaphysical and epistemological
   impossibility. It is only an ultimate goal, an _end in itself_, that
   makes the existence of values possible. Metaphysically, _life_ is
   the only phenomenon that is an end in itself; a value gained and
   kept by a constant process of action. ...

When Whitehead pointed out our tendencies to stick upon presumed objects
and immediate sensations he identified the common human _Fallacy of
Misplaced Concreteness_. And in considering the question of ultimate
values, Emmanuel Mesthene * emphasized the need to recognize the more
abstract character and evolving processes which are bound up in any human
appraisal of ultimate values:

  Although particular values may vary with particular times and societies, the
  activity of valuing and social functions of values do not change. This is
  the source of the stability so necessary to human moral experience. It is
  not to be found exclusively in the familiar values of the past. The emphasis
  of value analysis and study will shift from previously fixed values to
  individual and societal processses of assessing and extracting values from
  experience, that is, to the processes of valuing.

(* in *How Technology Will Shape the Future*, by Emmanuel G. Mesthene, in
_Purposive Systems_, the First Annual Symposium of the American Society for
Cybernetics (Spartan Books, 1968))

Consistent with this, it seems to me that PCT helps to elucidate the human
basis for, and the intellectual and social processes through which, along
with life and all its circumstances, more complex (i.e. abstract) values
are derived and can indeed develop.

Perhaps I am only providing a gratuitous reminder to CSG-L regulars of what
they already know well, at least in principle. However I would be
interested to know if there is any substantial disagreement!

Cheers!

Bruce B. (The writer is a recently retired
physician/psychiatrist/administrator who has had a continuing interest in
cybernetics since the 1960's.)

<[Bill Leach 940817.20:32 EST(EDT)]

[Bill Powers (940817.0610 MDT)]

I am delighted to see that you are "taking on" some Rand again. I have
always felt that Rand "was onto something" yet was also at times a little
"uncomfortable". I have never really attempted a "scholarly" examination
of her work (and certainly would not have been qualified to do so,
expecially at the time that I was reading Rand).

I don't think that she was particularly interested in the works of
"behavioural scientists" and had pretty well dissmissed such as she had
seen as irrelevent.

She was, I think, trying to justify on a logical basis alone, the
"American System of Economics" along with a generally accepted system of
ethics and morality.

Though just my own opinion of course, I feel that her efforts were at
least as good as any previous philosopher and probably better. There is
little doubt in my mind that an understanding of PCT would have been "an
awsome" weapon in her hands.

Throughout her work (I think) that you continually see an undefined
understanding that people function best (overall) when not interfered
with or forced. I have not gone back (yet) and re-read Rand, but believe
that PCT tends to indicate (to me) both why she was often right but also
why she was also wrong. I don't pretend to understand either in any
reasonable detail.

If one assumes that the maximum number of "happy" individuals is a
reasonable goal then PCT provides a significant amount of support for
much of what she said. However, when taken to "their logical
conclusions" (which she did do in at least one of her works) her ideas
seem to "be missing something".

For my own purposes I feel that Rand is "ultimately right" only when the
individual is ignored (seems to me to be the same problem for socialism
and other "good ideas").

I do believe that "central management" is doomed to failure. No doubt in
my mind that that "central management" is interested in what is best for
central management as opposed to what is best for "humanity" in general.

Thus, it would seem that some means of allowing maximum freedom (which
necessitates minimum control or attempts at control from external
sources) for individuals would be best. Yet it is Rand herself that
literally proposes anarchy as the best "form of government".

I tend to agree with her that "No individual has the right to initiate
the use of force against another individual." However, as she notes,
government (by consent of the governed) has the right to use force to
maintain "order" (boy is that a loose phrase). She then proceeds to
explain that administration of law should not be a government function
and her work further falls apart from there.

Though she may well have traversed her path over bog at times, I still
think that she at least asked many of the "right" questions and even
headed in the right direction in search of answers.

I doubt greatly that you missinterpret Rand, but the upshot of "The
virture of Selfishness" is that people always act in their own self
interest. What she did not say is that people always act in the own self
interest as THEY perceive such.

I don't know why people "sacrifice" themselves for their spouse,
children, church or whatever. But I do know that they ARE controlling
their own perceptions and that if you somehow "put them on easy street"
they WILL find away to remain the weary self-sacrificing people that they
perceive themselves to be.

Will definately be interested in any other comments and observations that
you have concerning Rand's work.

-bill

[Paul George 940818 15:25]

For those who are interested in testing whether PCT would be accepted by the
followers of Objectivisim there is a newsgroup alt.objectivism where current
advocates hang out. I don't have access to it, but there are occasional cross
posts to sci.econ. They may also be found on various of the libertarian forums.