With Ayn Rand
Part 2
by John Hospers
(Originally published in Liberty magazine, 1987)

In our last issue, John Hospers related what it was like to talk philosophy with Ayn Rand. Now, in the conclusion to his memoir, he details some of their philosophical differences and relates the inevitable falling out between the philosopher and the visionary. . .
Ayn occasionally expressed some disquiet (perhaps resentment) that she
was not recognized as a philosopher by the contemporary philosophical
community. In spite of long philosophical passages in Atlas Shrugged,
philosophers had never taken note of her views, and her philosophizing
in Atlas had largely fallen on deaf ears in the academic community.
I told her that philosophical discussion goes on almost entirely in
philosophical journals. What about philosophical books? she asked.
"Yours is a philosophical book," I said, "but it is a novel. It's not
that philosophers don't read novels—though a lot of them don't—but
they don't consider it their professional duty to do so." Besides, I
added, she had acquired a right-wing image in the popular press, and
that is a position that most academicians are strongly opposed to.
There were a few well-placed curses from Ayn about the prejudices of
the "liberal establishment."

I told her that if she wanted to become known in philosophical
circles, she should write a piece or two and submit it to the Journal
of Philosophy or the Philosophical Review or the Review of
Metaphysics. After its publication, I said, it would be studied,
commented on, and probably criticized. She would then respond to these
criticisms, which again would evoke more from others, and at that
point, I said, "I guarantee that you will be known as a philosopher."
But she never did this. She did not want to enter the arena of public
give-and-take with them. She wanted them to come to her. What she
wanted of philosophers, other than recognition, is not easy to say. I
am sure she would have cursed them soundly if they offered criticisms.
Even a mild criticism would often send her to the stratosphere in

At the same time, I must add, she would often tolerate criticism, even
revel in responding to it, if (1) it was given "in the right spirit"
(the vibes had to be non-hostile) and (2) it was sort of "on the right
track"—the sort of thing that could be said by someone who was "on his
way to the truth" but hadn't yet arrived there; then she would
"correct him" painstakingly and in detail.

I sometimes pondered how people could approach so differently the
enterprise of philosophy. I thought of the composers Igor Stravinsky
and Richard Strauss; each occupies a high place in contemporary music,
but neither could tolerate the other's musical idiom. Similarly, was
it just a difference of style among philosophers? Surely not. Each
comes to philosophy as a satisfaction for a felt need. I had been
"burned" early on by over-eager philosophic generalizations, and I was
weary of systems in which different philosophers said opposed things,
with no apparent way of resolving the issues in favor of the one or
the other. I had come to the conceptual-analysis route as a way of
resolving (or sometimes dissolving) problems that had long haunted me.
Ayn had aimed instead at a "final philosophical synthesis," and
regardless of its strengths or weaknesses, that is what she had to
present to the world.
Human beings are distinguished from all other creatures by the power
of choice. I agreed with Ayn about this—we know that the dog scratches
at the door but we don't know that he chose to do it (nor do we know
that he didn't). But I tended to disagree with Ayn about some of the
things that (according to her) we choose. Do we really choose "to
think, or not to think"? I for one (I said) don't remember making such
a choice. I would often think about things, perhaps because I am a
questioning sort of person and don't usually take things on faith.
Yes, often when confronted by a specific problem, I have said "I'll
think about it." But when my first acts of thinking occurred I no more
chose "to think or not to think" than I chose "to be or not to be."
But more than that I considered the scope of human choice to be much
more limited than she did. Some limitations we would both agree on: a
dunce can't choose to be a genius, and a crippled person can't choose
to walk (he can only choose to try, unsuccessfully). Without practice
a person can't choose to do shorthand or typing at 60 words a minute.
Neither can a person, just by choosing (or even by choosing and
trying), extricate himself from situations that have been years
abuilding. An obsessive-compulsive cannot just stop doing whatever he
obsessively has been doing for years, such as putting the key in the
lock three times and then tapping the floor three times (or whatever
his ritual is). And if a teenager ran away from home to escape
alcoholic parents and now has lived on the city streets for two years,
she can't just suddenly "straighten out" and become a normal
citizen—the gutter-instincts (survival by any means) are just too
strong by now. And so on for thousands of cases in which we may
unthinkingly believe people could have chosen to do what we want them
to do.

At this point in my diatribe Ayn reminded me that people do escape
from the slums, that with determination they overcome seemingly
impossible odds and sometimes become leaders in society. Prepared for
this observation, I granted that it was true; but the fact that one
person, A, can do this, doesn't show that other persons, B, C, and D,
can also do it. Each of them acts under somewhat different conditions
from A. They have one common denominator, slum upbringing; but some
had the love and trust of their parents, and the wherewithal to
prepare them to surmount adversities, and others did not; some had
father-figures with whom they could identify; and so on. (If a person
tries hard enough, he will succeed; but what is meant by "hard
enough"? Would you call it "hard enough" if he did not succeed?
Doesn't the statement come to the tautology "If you try till you
succeed, you'll succeed"?)

Anyway, all this preparatory conversation was so much chaff in the
wind, for Ayn hit me with the charge that I was sure she would come up
with sooner or later. "You don't believe in freedom at all, you are a

I knew what dense philosophical thicket lay in waiting here, with
vague and overlapping meanings of crucial terms like "free,"
"determined," and "caused." I hesitated even to embark on it. One must
come at the issue from so many different aspects, breaking one stone
and then another along the way—and most people lack the tenacity to go
through it all, they want quick and easy solutions, so that they can
repeat certain verbal formulas and convince themselves that they have
the problem mastered. So I began simply: "Determinism is just
universal causation. Everything that happens has some cause or other,
that's the core meaning of 'determinism' (to which other meanings have
sometimes become attached). The causes may be matter or mind, spirits
or God—all that determinism says is that everything has a cause, even
if we never find out what all the causes are." This was determinism in
its most neutral, vanilla-flavored sense, without the punch it was
supposed to pack, for there was nothing in my formulation that made it
incompatible with freedom, yet that was the main feature which led
many people to oppose it.

Of course, I continued, if everything is caused, events in human life
are caused too. Every decision you or I make is caused. But so what? I
decide to rake the leaves because I think the lawn looks unsightly. So
what's so hostile to freedom in that? Would it be better if I
causelessly raked the lawn?

But of course, no matter how many actions are caused by decisions (or
other things going on in the mind), ultimately these events in the
mind are caused by things that take place in the world outside the
mind. They may be hereditary factors or factors in the environment,
all very complex indeed, but if my decisions are caused, so are the
factors that caused them, and so on back. And over the hereditary and
early environmental factors I had no control at all. So am I really

Once the term "free" is raised, more clarification is called for. (I
discussed this with Ayn at much greater length than I have indicated
here.) The word "free," I began, does have a use; it does describe
something. Ordinarily we say that I am free when I am not coerced,
when no one has forced me to act as I do; I act as a result of my own
choice, unforced and unconstrained by others. If she marries him
because she wants to, she does so freely, but if she is dragged to the
altar she is forced. This is a rough-and-ready distinction that
everyone understands and uses. Does determinism (I said) really deny
this? Determinism says "My act is caused"; freedom says "I caused my
act." The difference is between the active and the passive voice.
Ayn started to object, but I went on. Sure, you can find causal
antecedents of human actions in the brain, in the environment, in
parental influences—in such complex causation as this there are
antecedents to be found all over the place. Most of the factors,
however, we don't know at all, such as what makes one person make this
decision and another person in the same circumstances make a different
decision. In the human realm we are very far from having established
determinism as we have done in physics and astronomy, where we can
predict an eclipse to the split-second a hundred years ahead.
Determinism asserts the universality of causes in the human realm,
without having gone much of the distance toward proving it that has
been accomplished in the natural sciences.

Ayn expressed the belief that in the area of human choices, there are
indeed causes, but that a person in so acting is self-caused (causa
sui). I expressed doubt as to what this could mean. If something is
caused, isn't it caused by something else, something other than
itself? How could my decision cause itself? Cause has to do with
origination, and how could the origin of choice X be choice X itself?
We can say, truly, that I caused my choices—that I, a complex set of
actual and dispositional characteristics, caused this act of choosing
to occur—but that is not the same as saying that X caused X. I was not
able to see causa sui as anything but a desperate attempt to escape
"the dilemma of determinism."

At any rate, what I wanted to make crystal clear to Ayn was that the
"principle of determinism" (or Causal Principle), that everything that
occurs has a cause, is not merely a statement (true or false) about
nature's workings; I tried to give her a sense that it had a much more
complex and ambivalent epistemological status than that, which
rendered labels like "true" and "false" extremely dubious.
I tried to make the epistemological point very simply. Suppose a
chemistry student gets some quite unexpected results when he repeats a
laboratory experiment. He then reports to his teacher that the same
effects don't always arise from the same cause: he set up the
experiment exactly the same both times, yet got different results (an
orange precipitate in the first case, none in the second). Conditions
C produced result E-l the first time and E-2 the second time—different
effects from the same cause! Yet his teacher wouldn't tolerate this
for a moment. Maybe he had some evidence that the C's weren't the
same—he might find an impurity in the liquid the second time that
wasn't there the first. But more usually he had no evidence at
all—there was a difference in the E's, he reasoned, so there had to be
a difference in the C's. And we would say this whether we know it or
not, whether we ever discover it or not.

And so on in general, I said. If after repeated trials we discover the
cause of something, we say that confirms the Causal Principle even
more; but if after repeated trials we fail to discover the cause, we
don't say it had no cause, but only (and always) that it's there but
we haven't discovered it yet. Isn't this a remarkable asymmetry? Isn't
this very peculiar—a principle that discoveries confirm but no
discoveries can disconfirm? A principle that parades as a truth about
the world, yet is apparently immune to refutation by discoveries about
the world? What does this show? Isn't there "something funny going on"
here? Aren't we trying to run with the hare and hunt with the hounds?
Isn't this asymmetry a ground for suspicion?

I was not sure whether Ayn followed the direction in which I was
pointing, but I went on. I suggested that the much-vaunted Causal
Principle was not a statement about the world at all—not like "All
birds fly," which can be disconfirmed by finding a few ostriches. That
which can be confirmed by experience but not disconfirmed by
experience is not a statement about the world. It might be an a priori
truth, like the Law of Identity, not subject to, and not requiring,
confirmation by experience. But I could not think it a priori because
it made claims about nature which, I suggested, could only be
confirmed by observing nature—which can't be done from one's armchair.
Instead, I suggested that it was a kind of scientific rule-of-the-game
("heuristic maxim") that has stood us in good stead because when we
used it in the past we have found lots of causes, but one which we
don't permit to be disconfirmed, for there's nothing that we could do
that we need to count as disconfirming it. It's a rule, the following
of which has pragmatic value—it helps us to find more causes; but
since it isn't falsifiable it doesn't count as an empirical rule,
which is what it would be if it were like "All birds fly" or "All
bodies gravitate."

Something may look like a plain and simple statement about the world,
the only question about it being "Is it true or false?" But what looks
like a statement needn't be a statement, and perhaps this one
isn't—instead maybe it's a rule that we use to guide our future
scientific activities, or express a faith in some ultimate uniformity
of nature. And if it has that status, then our talk about the
Principle of Determinism being true or false is mistaken from the
outset. We have been misled into thinking it has this simple
true-false status at all.

I could not expect Ayn or anyone else to grasp the import of this at
once: to someone who has spent most of a lifetime asking "Is it true
or is it false?" it is disorienting and mind-blowing to be told that
this distinction may not be applicable to the question at hand. One
has to see how this approach can be applied to other philosophical
problems (not just determinism), and how it clarifies or dissolves
those problems rather than leaving them forever intractable. But to
appreciate all this requires much more one-on-one philosophizing than
I had done with Ayn. I had high hopes that we might yet do it. But
whether it was the defects of my presentation or her disinclination to
think outside the traditional categories with which she had operated
for many years, I was never able to get far with her on this—it
remained terra incognita to her, and her responses seldom indicated
that she had grasped the true import of what I had said.
It seemed to me that she failed to appreciate the subtle shifts of
meaning of crucial terms that often occur midway in a discussion, and
result in total confusion unless the shifts are pointed out when they
arise. She seemed to have a number of ideas packaged together under
the heading she called "determinism" and assumed that the term
retained the same meaning in its various contexts of use (a common
enough error). One example that I particularly remember is that she
would say that according to determinism a person never could  do other
than he did; and that if exactly the same circumstances were to arise
again (according to determinism), the same result would occur.
"And if the same thing didn't recur," I said, "then you'd conclude,
without further evidence, that some factor in the circumstances
leading up to it were different this time. And you would say it," I
insisted, "as an a priori assumption, without any independent evidence
that any of the conditions were different." I remember using this
analogy: A says "All swans are white," and B replies that there are
black swans in Australia; to which A replies, "If they're not white,
they're not swans."

I tried to open up to her the logic of the word "could." I said that
"could" is an ability word: when someone says "You couldn't have done
otherwise," this charge invites the retort, "Not even if I wanted to?"
And of course if I had wanted to I would have done something
different—I would have continued reading the paper instead of going to
the kitchen. My wanting to do X instead of Y could well be the
deciding factor that caused me to do X instead of Y. So, I said, it
isn't true that I couldn't have done Y; I would have done Y if I had
wanted to.

But the next step, of course, was "According to determinism, you
couldn't have wanted anything other than you did." But what, I said,
does "couldn't" mean in this sentence? That I wouldn't have wanted
anything else even if I had wanted to? No? If not, then what does
"could" mean in this sentence? I suggested that it would be preferable
to say that if exactly the same conditions were repeated the same
event would have happened—and then show the unprovability of that
statement because of the impossibility of tracking down all the

Ayn was impatient with such subtleties. When we recapitulated, she
would always return to the position that if you are a determinist you
believe that nothing could have happened except what did happen. And
once again I would inquire what "could" might mean in that
sentence—and we would start on the merry-go-round once again.
Of course, I went on, there are (as usual) other senses of "could" as
well, not specifically applying to human action. We may say that when
you let go this pencil from your hand it could not fly upwards, that
it could not do anything but go downwards in accordance with the law
of gravity. But that is only to say that the downward motion of the
pencil is the one that accords with laws of nature. That is, if you
assume certain laws of physics, then the pencil could not (logically
could not) have moved in any other way. The "could" here is a logical
"could" (not an empirical one) expressing the logical connection
between statements—statements of the laws of nature, statements about
the mass and volume of the pencil, and the third statements (the
conclusion) about the behavior of the pencil. We can say that granted
certain premises, this behavior could not have been other than it was.
(But, I added, saying that the pencil could not have behaved otherwise
is already a departure from the central meaning of "could," which has
to do with ability.)
I never made much progress with her on determinism, but when we talked
one evening about a specific kind of causation—extra-sensory
perception—I evoked in her an unexpectedly vigorous response.
I do not remember how the subject arose, and I didn't even consider it
a philosophical area of discussion, but I was describing to her Soal
and Bateman's book Experiments in Parapsychology. I explained that out
of thousands of tries, a few people made very good subjects; they were
able to state with considerable accuracy truths that (as far as we
knew) were discoverable only by sense-perception, but which they could
not have known through sense-perception.

A man was sealed into a room evening after evening, and there was no
possible communication between this room and another room three doors
away—there were scientists who averred that there was no way a person
in Room 1 could convey information to someone in Room 4. In each of
these two sealed-off rooms, cards were being pulled from a deck one
per minute. Every minute a bell would ring, at which moment a card
would be pulled from a deck in one room and the subject in the other
room would write on a piece of paper which card he thought it was.
There were five different kinds of cards (apple, elephant etc.) and
thus one chance out of five of guessing correctly. Getting the correct
result slightly above chance (20 percent) for a time wasn't
particularly noteworthy, but getting results like 40 percent correct
over 100,000 attempts was quite remarkable, the chances against this
being some trillions to one. Yet several subjects were reported to
have done just that, and no one knew how. Ayn looked skeptical but
allowed me to proceed.

Moreover, I went on, the subjects had improved with practice. From a
fifth they had gone gradually to a quarter and even to a third. No one
could figure out how they got the ability to do this. They themselves
didn't know: they weren't aware at the time that they were guessing
correctly, they just "put down the first thing that popped into their
heads." And then the rules of the game were changed—"You will now
write down the card that was being pulled last night at this point in
the sequence"—and their achievements vanished (went down to chance),
but came up again with practice to the previous fraction.

And then, most curious of all, the rules were changed once more: "You
will write down the card that is going to be pulled at this point in
the sequence tomorrow evening." Again the results went down to chance,
but again with practice the record gradually improved. But the
implications of it shocked me: How could they possibly know the
future? What if between tonight and tomorrow night the entire building
burned down? And so on.

Ayn was now taken quite aback, and thought I should give no credence
to any of this. It implied reverse causality, she said, and that was
impossible—something at a later time causing something at an earlier
time. I agreed that reverse causality was impossible—such as the rain
tomorrow helping the crops grow today. But I didn't think the example
involved reverse causality but only precognition. We all predict the
future, I said, usually with some evidence; what made this case
peculiar was the ability of the person to make a correct prediction
again and again without apparently having any evidence whatever. (At
least there was nothing known to science that we would call evidence.)
That was what I found different about this kind of case, and I
couldn't think of any explanation.

Ayn was quite shocked that I would take any of this "mystery-
mongering" seriously. (It was hard to convey briefly the
import of entire books on the subject, and the extraordinary lengths
to which people had gone to make sure there was no sensory route by
which A could have known B.) Didn't I know that reality does not work
in that way? Perhaps so, I said—and I added I didn't much care whether
reality does work in that way or not—but whether it does or doesn't is
not something we can know by just pontificating about it from our
armchairs: we have to go the difficult route of empirical
investigation to find out whether people can know truths about the
universe that are not mediated through sense-organs. One cannot know
this a priori, I claimed; one has to go the more difficult route of
checking it all out in detail. But I gathered that she considered this
all a matter of necessity—that it was necessarily the case that nature
doesn't work in this way. She was more disturbed about my
permissiveness on this subject than I had thought she would be.
Instead of saying that nature can't work in this way, the question for
me was whether in fact it does; if it does, then it won't do to say
that it can't.

For me, the question of what caused what is entirely a contingent
matter, on which we can make judgments only in the light of
observation of the world. But it dawned on me that Ayn didn't accept
the distinction between necessary and contingent at all. For her, it
seemed (though I never got it in just these words) every statement
that is true is necessarily true. "Doesn't everything that happens
have to happen?" she once asked me.

I replied that one would first have to inquire about the meaning of
the phrase "have to." In most locutions, "have to" involves a command
or order—"I have to be in by midnight." When one says that events in
nature, such as a comet entering the earth's atmosphere, have to
happen, it sounds first off as if this event is being commanded,
perhaps by God. But this is surely not what most people mean when they
say it. Perhaps we mean that if one accepts certain laws of nature
(concerning gravitation, mass, velocity), and if one grants certain
initial conditions (Comet X is in such-and-such a position at
such-and-such a time), then Comet X must be another place at a
specific other time. (Not that the comet must—but that the
statement—the conclusion—logically must be true if the premises are
true. The "must" is about the relation between statements, not about
phenomena in nature.) When I say that if I let go of this pencil it
must fall, doubtless I am saying that the statement that it does (or
will) follows from certain laws of nature plus initial conditions. But
it would be clearer if I just said that the pencil will fall.

There are many uses of "must" and "have to" (I took her through
several more) and I told Ayn that I thought she was telescoping
several disparate uses of the term "must" into one, without
distinguishing among them, and that this might be why she was led to
make such a statement as "whatever happens must (has to) happen." (If
you take it quite literally, I said, it seems like a more extreme
fatalism than any view I have ever countenanced.)
Ayn usually let me take the initiative in deciding what subjects we
should discuss. The conversations described in this paper reflect
largely my choice of topics—these were the things about which I was
interested in sounding her out. I reflected later that in this respect
I had probably made a mistake. Only occasionally did we get around to
discussing topics that were central to her philosophy. That is why
some topics central to her are largely absent from these pages. Her
papers on these subjects had yet to be written.
"A is A" is, I insisted, a tautology, but an important one: every time
a person is guilty of a logical inconsistency he is saying A and then
in the next breath not-A. Thus "A is A" is something of which we need
to remind ourselves constantly. But it is not, I said, an empirical
statement: we don't have to go around examining cats to discover
whether they are cats. (We might have to examine this creature to
discover whether it is a cat.)

But, I said, statements of what causes what, such as "Friction causes
heat," are empirical statements; we can only know by perceiving the
world whether they are true. How, I wondered, can the Law of Causality
be merely an application of the Law of Identity? You could manipulate
the Law of Identity forever and never squeeze out anything as specific
as a single causal statement.

But (I went on) I could see how such a confusion might be generated. A
tautology can easily look like something else. "A thing acts in
accordance with its nature" might be one example. This might be taken
as an instance of the Law of Identity: if a creature of type X acts in
accordance with laws A, B, C, and this creature doesn't do that, then
it isn't an X. If dogs bark and growl and this creature hisses and
meows, it isn't a dog; that is, we wouldn't call anything a dog that
did this. So we can plausibly classify the statement about what we
call "a thing's nature" as special cases of the Law of Identity. But
this, I insisted, tells us nothing about the world, but only about how
we are using words like "dog" and "cat."

What is a thing's "nature" supposed to be anyway? I went on. Is a
thing's nature its definition? Some might say yes: it's the nature of
water to be two parts hydrogen and one part oxygen. But one might also
answer no: it's the nature of water, one might say, to flow downwards,
and this is no part of any (usual) definition of "water." It wouldn't
even be true if atmospheric pressure were ever so much less than on
earth (it might evaporate and not flow). So to answer the question, we
have to know what the person means by talking about a thing's nature.
Often, I suggested, when we talk about a thing's nature we are talking
about a set of dispositional traits: thus, "It is the nature of cats
to prowl"—yet so far as I know the tendency to prowl is not listed in
the definition of "cat." Or, when we say "I used to think his lying
was just a quirk, but now I think it's his nature," we are saying that
his tendency to lie is a more fundamental trait than we had previously

I could see that Ayn was getting bored, so I summarized the moral of
the tale: that statements about "X's nature" sound simple and easy,
but that under this linguistic simplicity lies a morass of vagueness,
which comes out only gradually as we explicate one case after another
in which we actually use the expression. I seemed unable to convey to
Ayn any sense of this; and yet, it seemed to me, what was wrong with
the usual philosophic formulations, including hers, couldn't be
appreciated without going through the detailed "digging" required to
turn up these disparate meanings, and their confusion with one another
from which the errors flow. Philosophic formulas, I said, merely give
us "philosophy on the cheap."
It was inevitable that sooner or later we would get to the subject of
definition. I never had an opportunity to present my views on this
systematically, from the ground up. I had done this in some detail in
my book Introduction to Philosophical Analysis, in the long 100-page
introductory chapter entitled "Words and the World." I gave her a copy
of the book and encouraged her to read the relevant chapter. But she
never did; I was disappointed by this, for I had thought we could use
this material at least as a starting place for discussion, but in time
I realized that she read almost no philosophy at all. And I was amazed
how much philosophy she could generate "on her own steam," without
consulting any sources.

She began by insisting that one should search for true definitions,
and I responded that definitions were neither true nor false. But it
shortly turned out that I was talking about definitions of words and
phrases, and she was talking about definitions of things (entities in
the world) or, sometimes, concepts of those things. But I expressed
ignorance as to what the phrase "the definition of a thing" meant. (We
also discussed "definition of concepts," examining the differences
between words and concepts.)

I suggested that there were no true or false definitions. "The word
'symphony' once referred to any orchestral composition, without voice,
in four movements," I said. "Then, as in Beethoven's 9th, voices would
sometimes be introduced and the work would still be called a symphony,
so that was no longer a defining feature. Then in the 20th century
came one-movement symphonies, such as Sibelius' 7th, so the
four-movement requirement fell out. What happened was that the word
'symphony' was no longer used to describe what it had described
before. But there is no true or false definition of 'symphony.'"

A simple case to the contrary, Ayn said, was that H2O is a true
definition of water; if someone said water was HO or H2SO4, he would
be mistaken.

I responded that I saw nothing but confusion in this. "It depends on
what you mean in the first place by the word 'water.' If by 'water'
you mean H2O, then course 'Water is H2O' is true because you've
already defined water to mean that. All you get that way is 'H2O is
H2O,' a simple tautology. But of course you might not already mean
that by the word 'water'—early man surely did not. He meant the liquid
that flows in streams and rivers. In that meaning, it is true that
water is H2O—that is, the liquid in streams and rivers has the
chemical formula H2O. That is a true statement about water—an
empirically true statement, not a definition. Once you are clear what
you mean by the word, the issue is resolved."

Ayn alleged that man is a rational animal, and that this is a true
definition. It is true, in other words, that that's what man is. I
replied that it all depends what you mean by "man" in that sentence.
As a rule we employ a biological definition of man—man is a creature
with two legs, two arms, walks upright, etc.; that's how we identify
creatures as human without knowing anything more about them than our
senses present to us. Now, the creature that fulfills that biological
requirement is also a rational animal (that is, has rational
potentialities, even if unfulfilled)—that is a true statement: not a
definition, but a statement about the creatures identified by the
first (biological) definition. (Of course, again, if by "man" you
already mean "rational animal," then it's a sheer tautology.)
We could say, I suggested, that man is a laughing animal, or an
aesthetic animal (the only creature that enjoys works of art), a
volitional animal (the only creature capable of choice), and perhaps
several others. But, as Ayn aptly pointed out, these features are less
fundamental. If we were not rational animals we would not be able to
comprehend works of art or see the point of jokes; the rationality
explains the other characteristics, not vice versa. I assented to
this; but I insisted that my point still held, that if "man" is
already defined as a rational animal, the statement that man is a
rational animal is a tautology (merely an example of A is A); whereas
if "man" is defined biologically, as we ordinarily do, then the
statement that man is a rational animal is true, but not a definition.
A stipulative definition, I said, merely tells others how we're going
to use a word ("I'll use this noise to mean so-and-so"), and a
stipulation isn't a true statement, just a proposal to use a noise in
a certain way. A reportive definition is a report of what a word is
used to mean in a language-group. Thus, "A father is a male parent" is
a report (in this case a true one) of what the word "father" is used
to mean in the English language. And finally, if you already mean by
"father" a male parent, the definition of "father" as male parent is
presupposed, and the statement "A father is a male parent" comes to "A
male parent is a male parent," another instance of "A is A." Confusion
comes only if we get these scrambled together.

Is "Steel is an alloy of iron" a true definition of steel? No, I said,
it is a definition of "steel" if that is what you choose to mean by
the word "steel." It is also a true report about how users of the
English language use the word "steel," and as such it is a true
reportive definition. And if you already mean "alloy of iron" by the
word "steel," then again you have a tautology, Steel is steel, A is A.
It seemed to me that these distinctions clear up the question. In
every case we define words and phrases, and we describe things (using
the words or phrases).

Whales were once thought to be fish. When it was discovered that they
were mammals, wasn't this a discovery of the true definition of
whales? The discovery (an empirical one), I said, was that those
creatures which we called "whales" (on the basis of their shape, size,
and general appearance) also had the feature of being mammalian. We
then changed (or biologists did) the definition of the word so as to
include being mammalian as a defining feature; biological
classification on the basis of mammal, reptile, etc., had already long
been in place; so after the discovery nothing that looked like a whale
but was a fish would have been called a whale. The re-definition of
the term was simply an adaptation to existing methods of biological
classification. But the discovery, that these creatures were mammals,
was an empirical one, like the discovery that some nebulae are
actually galaxies.

This is one of the issues that seemed so obvious to me that I did not
see how anyone could think otherwise. That is why I tended not even to
remember opposing remarks as long as they were not clear to me. Rather
than misreport what Ayn said, I have chosen not to say anything about
her remarks: what I said is very clear to me, what she said is not.
At the time being described, Rand's non-fiction works, including
Introduction to Objectivist Epistemology, had not yet been written. I
would like to think that our discussions helped motivate her to write
some of these non-fiction works. At the time of our discussions she
was writing very little. Time was on her hands, and perhaps that was
one reason for inviting me back.
She vehemently denied the validity of certain distinctions, like
analytic vs. synthetic and a priori vs. a posteriori. Both were
Kantian distinctions, and her hatred of Kant may have played a part in
the rejection; but more likely her rejection of the distinctions
played a part in her hatred of Kant.

Already at the time of our discussions there was critical talk in
philosophic circles about the analytic-synthetic distinction. Is it
analytic to say that all green things are extended? Quine had asked,
and concluded that the failure to provide a satisfactory answer was
due to the unclarity of the term "analytic," not to any defects in
"green" or "extended." But the examples I used were of the very
simplest sort: "All A is A" is analytic, I said (it's another
formulation of the Law of Identity), and "All A is B" is not. "Lions
are lions" is analytic and "Lions are fierce" is not—to determine that
you have to observe lions. And the same for a priori: you don't have
to go to the next room to discover whether the cat is a cat, but you
do have to in order to find out whether the cat is lying on the bed

Why did Ayn deny a distinction that seemed to me so obvious—perhaps
not for far-out cases like colors being extended, but for ordinary "A
is A" type cases? She seemed to think, as Leibniz had done for
different reasons, that the distinctions do not apply because all the
statements are really in the same bag. All the features of lions,
whether now known or not, are really a part of their definition. All
statements about X follow from X's definition—that seemed to be the

But I did not see how this could be so. That this table is a solid
object does follow from (or is contained in) the definition of a
table. But that we are now sitting at this table does not. Nothing in
any definition of a table known to me could possibly tell us whether
it is true that we are now sitting at the table.

Perhaps the issue has a different focus: This would not be the egg
that it is if it had not been laid by this hen, and I would not be the
person I am if I had not been born to the specific parents I had.
True—but would I also have to have the characteristic of having been
born at the moment that I was? If I had been born a day earlier (to
the same parents etc.), wouldn't it still have been me? True, it
wouldn't have been me if the birth had taken place in ancient
Greece—the parents wouldn't have been the same, etc. But would one
really be prepared to say that all features of me are defining,
including the mole on my cheek and the fact that a bee had just stung
me? I saw nothing but endless confusion in that way of trying to deny
the difference between necessary and contingent statements.
I tried using some examples, of the kind that made my students catch
on to the distinction most quickly. That this flower is red, that
there are six of them on this plant, that such plants exist at
all—these are contingent statements, they depend on the way the world
is, which can't be known a priori; that 2 + 2 = 4, that the angles of
a triangle equal 180 degrees, that if A is larger than B then B is
smaller than A—these are necessary truths, I tried to explain, even if
one doesn't accept the analytic-synthetic distinction.

Or again, with regard to possibility and impossibility: I can't jump
20 feet high, but I (logically) might, and if I claimed to do so my
statement would be false, but there would be no contradiction in it.
But if I claimed to have gone backward in time, and disappeared from
1961 to 2500 B.C. (and what could that mean?), and actually helped the
Egyptians build the pyramids—this, I said, was a logical
impossibility, because contradictions would be involved in asserting
it: I would be saying that (for example) the pyramid-building occurred
without me (I wasn't born yet) and also that I participated in it (by
"going back" in 1961 to 2500 B. C.); and that there were, let's say,
5,368 persons building the pyramids and (with the new addition of
myself) there were 5,369—but there (logically) couldn't have been both
5,368 and other than 5,368. And so on. She granted the impossibility
in the second case, but perhaps not for the reason I mentioned. To her
all impossibility was of one stripe, and she did not admit the
distinction between logical and empirical possibility.
I stated a problem (or pseudo-problem) which seemed to fascinate my
students: "How do you know that you and I are seeing the same color?
True, we both pass the color-blindness tests, and you say you see
green when you look at the tree, just as I do, but how do I know you
aren't the victim of a "reversed spectrum," for example that you
regularly see red where I see green and vice versa, but of course you
call it green like everyone else, since that's the word you've been
taught to use in describing the color of trees? But perhaps if I could
see what you're seeing, I'd call it red, or something else. After all,
how do I know?" Maybe the outcome has no practical import, but it's a
nice theoretical question anyway—the sort of thing that science seems
unable to answer.

I cannot say that Ayn was fascinated by this question. She regarded it
as rather trivial. But she heard me out. I suggested that you can
(usually, perhaps always) get to what a questioner means by his
question, if he can tell you what sort of thing would satisfy him as
an answer—what precisely does he want to know? Now consider these
possibilities (I said): (1) Suppose it were technically possible, as
one day it may be, to connect one person's eyes and optic nerve with
another person's brain. You could, then, quite literally see through
the other person's eyes; and then you would know whether the leaves
looked the same color to you as they did when you looked through your
own eyes. You'd be able to compare what you saw with your former eyes
with what you saw through your new eyes. Perhaps when you did this you
would say, "They still look the same to me," and that would settle the
question; or you might say "They don't look as they used to at all,"
and that too would settle the question.

But of course (I pursued) one may object that this won't do. (2)
Exchanging eyes isn't enough, runs the objection. The interpretation
of these visual data takes place in the brain. To settle the issue, I
would not only have to have your eyes, I'd have to have your brain (or
at least a part of it). But now we run into what's called the problem
of personal identity. If my brain were put into your body and vice
versa (assuming this to be as technically possible as exchanging eyes)
would it still be me? Would it still be me, with all my brain's
memory-traces now inside your head? Here we run into a problem that's
more than a technical problem; what is it that constitutes one's self,
if not one's perceptions, dispositions, and memories? How can I
exchange brains with you and still be me? Thus, if this second
alternative is the one demanded to resolve the problem, then unlike
the first alternative, it can't be solved: the conditions demanded for
the solution are self-contradictory.

Ayn wasn't very impressed with all this. She didn't consider the
issue to be of any importance in the first place. She was
temperamentally unsympathetic to this way of doing philosophy. And she
had no patience with the distinctions I used in order to arrive at a
solution. For her it was a non-solution to a non-problem.
In spite of her lack of concern for shifts of meaning in a word or
phrase, I had to be very careful what terms I used in her presence;
for some terms, if I used them, would trigger in her an instant
conclusion that was quite foreign to anything I meant. When I
mentioned that a theory in science can be accepted or rejected on
pragmatic grounds—as a device for explaining the most by means of the
least—she would hear the term "pragmatic" and accuse me of being a
pragmatist. And then I would explain at some length that I was not a
pragmatist in any sense that she probably had in mind—for example, I
did not hold that the truth of a statement had anything to do with its
utility. I only used the term within a definite context, with a
meaning defined within that context—and one should not jump to the
conclusion "You're a pragmatist," for I wouldn't even know what she
meant by the term in that sentence.

For a person who was always insisting on "iron-clad definitions," I
found her linguistic habits quite sloppy. I was aware that Rome wasn't
built in a day and that she had not grown up in a tradition in which
sensitivity to these matters was considered important—one just strode
over the issues in seven-league boots (my characterization, not hers).
Still, philosophic outcomes depend so much on just such subtleties
that I became discouraged when after many hours of discussion she
showed no more awareness of where I was really coming from than she
had when we started.

I had no problems with her ignorance of modern logic or physics (such
as Heisenberg's principle), but when the very issues she raised
required a finely honed instrument to grapple with them insightfully,
and she seemed quite unaware of what that instrument could do, and
remained so after time, I gradually became as discouraged with her as
she was impatient with me.
Somewhere she had picked up the idea that philosophers in the
twentieth century were skeptical about the existence of an "external
world" (tables, trees, stars, etc.). I told her that skeptical
arguments in this area were still extensively examined, in the
tradition of Hume, but that no one so far as I knew had any actual
doubts about the existence of the chair they were sitting on, and so
on. But that, she said, was the mistake: they don't doubt it in
practice but they do in theory—they don't practice what they preach. I
explained that when skeptical arguments occur, as in Hume, they have
to be met, in an attempt to make theory accord with practice; one
can't just assume that "common sense" is always right. I explained a
similar situation in Zeno's paradoxes, and Parmenides' attempt to deny
the reality of motion. I said there were lots of problems about the
relation of the world to the senses by means of which we perceive it.
I did mention, almost incidentally, an attempt to prove that we know
the existence of the external world for certain, namely by Prof.
Norman Malcolm in his essay "The Verification Argument" (in Max
Black's anthology, Philosophical Analysis). Instantly she picked up on
this, inquiring about Malcolm as a possible ally. She wanted to know
more about him and even to invite him to New York for a personal
meeting. She did not read his article, or anything else by him, but I
outlined the rather complex argument of the article for her in two
typed pages, trying to state his premises accurately and show how they
yielded his conclusions. She expressed gratitude to me for doing this.
But, she wondered, why should a person go to such lengths to defend a
thesis that was so obvious? I realized that to Ayn the existence of
the physical world was axiomatic and didn't require defense, and told
her that she would probably find no particular ally in Malcolm, who
was most interested (in the essay) in exploring the implications of
terms like "verification" and "certainty." At any rate, there the
matter dropped. She took my word as to what his arguments were, and as
far as I know she  never read anything to enlighten her further on the
We discussed many other philosophical issues, often in a brief and
fragmentary way, before concentrating on something else. I omit here
those issues of which I could not now give an accurate account from
memory. In many cases I remember more clearly what I said than what
she said. Her non-fiction works had yet to be written, and what I
endeavor to record here is what she and I said then, not what we might
have said later. Moreover, most of my readers will probably be
acquainted with her position on various issues, but unacquainted with
mine; and I want to provide some conception, however brief and
unsystematic, of where I was coming from on the issues we discussed.
When we discussed metaphysical and epistemological issues, a certain
tension between us would very gradually and almost imperceptibly
arise. I could usually avoid an unpleasant scene by attributing
(correctly) the view being discussed to some actual philosopher,
living or dead, and then she could curse the philosopher in question
and take the heat off me. It's not that I wanted to avoid
responsibility for the view, but I wanted to avoid unpleasant scenes,
which only impeded the progress of our discussions, and achieved no
worthwhile end that I could think of. But it was clear that I was not
"giving in" to her brand of metaphysics, and equally clear that my
methods of what I liked to call philosophical clarification were
falling on arid ground in the present case. I became somewhat
discouraged, especially since she seldom acknowledged an error and
seemed less interested in learning than in defending prepared
positions. Moreover, what seemed like a blinding philosophical light
to me would be a total dud to her, and her highly abstract
philosophical pronouncements often seemed to me confused, unclear, or
false, effective though they might be as banners for enlisting the
philosophically un-washed.
Meanwhile, several incidents occurred that distressed me. There was a
professor at a midwestern university who had been denied tenure some
months earlier, for saying that he wouldn't mind too much if his
daughter slept around a bit before she decided on whom to mate with
for life. The faculty was up in arms against the university
administration for terminating him, and started a nation-wide petition
on his behalf. I had also signed a petition requesting that he not be

When I showed Ayn the letter to which I had responded on his behalf,
Ayn saw my name on the letterhead and urged me strongly to dissociate
myself from any attempt to defend him. He should not have referred to
his daughter publicly in that way, she said. I asked her whether she
really thought he should be denied tenure just on account of having
said what he did. And Ayn's reply stunned me: he should have been
terminated from his job, she said, even if he'd had tenure. Knowing
all that tenure means to someone who has worked for years to earn it,
I found her reply shocking and astonishing.
Newsweek wrote a terribly unfair piece about Ayn. I responded to it by
letter, trying to answer their charges point by point. I gave Ayn a
copy of my letter. Newsweek never published it, but that, said Ayn,
made no difference; what mattered was that I had come to her defense
by writing it and responding to the false charges.

Not long after, New York University's philosopher Sidney Hook attacked
her in print, and she wanted me to take him on as well. Knowing
Sidney, I was disinclined to do this. He already knew about my
acquaintance with Ayn, but we had never discussed it further (I hardly
ever saw him). Should I now condemn him publicly and destroy a
long-standing friendship? I knew that this friendship would be at an
end if I condemned him.

Ayn was sure that nothing less than a public condemnation was required
to prove to him how much I was devoted to "intellectual objectivity."
But she had very little conception of the manners and morals of
professional academicians—they can get along well and even be friends,
while disagreeing strongly with one another on rather fundamental
issues. The philosophic arena was one for the friendly exchange of
diverse ideas. But for her, it was a battlefield in which one must
endlessly put one's life on the line. I was not willing to risk years
of occasional friendly communion with Sidney by condemning him
publicly, even if I thought he was mistaken in some of his

But for Ayn this was a betrayal. It almost cost us our friendship. In
the end she attributed my attitude to the misfortune of having been
brainwashed by the academic establishment, at least with regard to
their code of etiquette.

I once mentioned to her my friendship with Isabel Hungerland, a
distinguished aesthetician from Berkeley with whom I would discuss
issues at philosophical conventions. Ayn inquired what her politics
were. "As far as I know, she's a liberal," I said. "What!" exclaimed
Ayn, "a friend of yours—a liberal?"

I realized then that I was expected, once I knew Ayn, to sacrifice
the friendship of all persons with political (and other) views opposed
to hers. Not that I would have to—I was supposed to want to. It was
immoral of me to continue to deal with such people. With many of them,
as with Isabel, I had a kind of relaxed, laid-back relationship, never
talking politics at all from one year to the next, and often not
knowing what their political views were. But now I was supposed to
excommunicate them all. "If thine hand offend thee, cut it off." I was
not willing to plant a flag on a new terrain and thereby disavow my
allegiance to all other views, and I deeply resented Ayn's attempt to
steer me in that direction—or should I say, her assumption that I
would "of course" do such a thing.

It wasn't that I would have been unwilling to declare where I stood,
if I had been totally convinced and was prepared to defend it. I try
not to back off of commitments. But my whole way of coming at
philosophy was quite different from hers, and in spite of various
attempts I don't think she ever understood mine. With her, it was as
if she were developing a Euclidean geometry from a set of axioms; I,
on the contrary, was the gadfly who kept puncturing the axioms or
finding their meaning (in some cases) to be vague or confused. As a
result of this I was convinced that "the high priori road" was not the
way to go in philosophy; I was sure that a careful, step-by-step,
case-by-case approach, frustrating though it might be in the work
required and the time needed to get anywhere with it, was the only
road to progress. This wearied her, bored her, and ultimately repelled
The more time elapsed, the more the vise tightened. I could see it
happening; I hated and dreaded it; but knowing her personality, I saw
no way to stop it. I was sure that something unpleasant would happen
sooner or later. The more time she expended on you, the more
dedication and devotion she demanded. After she had (in her view)
dispelled objections to her views, she would tolerate no more of them.
Any hint of thinking as one formerly had, any suggestion that one had
backtracked or still believed some of the things one had assented to
previously, was greeted with indignation, impatience, and anger. She
did not espouse a religious faith, but it was surely the emotional
equivalent of one.

When I was authorized by the American Society for Aesthetics to ask
Ayn to give a twenty-minute talk at their annual meeting, which would
take place this time in Boston the last weekend of October 1962, I
passed on the offer to her at once. She accepted, with the provision
that I be her commentator (all papers were required to be followed by
a response from a commentator). She thought that I would understand
her views better than those who had no previous acquaintance with
them. I consented.

And so it was that on the last Friday night of October 1962, she gave
her newly-written paper "Art and Sense of Life" (now included in The
Romantic Manifesto). In general I agreed with it; but a commentator
cannot simply say "That was a fine paper" and then sit down. He must
say things, if not openly critical, at least challengingly exegetical.
I did this—I spoke from brief notes and have only a limited
recollection of the points I made. (Perhaps I repressed it because of
what happened shortly thereafter.) I was trying to bring out certain
implications of her talk. I did not intend to be nasty. My fellow
professors at the conference thought I had been very gentle with her.
But when Ayn responded in great anger, I could see that she thought I
had betrayed her. She lashed out savagely, something I had seen her do
before but never with me as the target. Her savagery sowed the seeds
of her own destruction with that audience.

When her colleague Nathaniel Branden and I had a walk in the hall
immediately following this exchange, there was no hint of the
excommunication to come. But after the evening's events were
concluded, and by previous invitation I went to Ayn and her husband
Frank's suite in the hotel, I saw that I was being snubbed by everyone
from Ayn on down. The word had gone out that I was to be (in Amish
terminology) "shunned." Frank smiled at me, as if in pain, but he was
the only one. When I sensed this, I went back to my room. I was now
officially excommunicated. I had not so much as been informed in
advance. It was all over. In the wink of an eye.
So now a two-and-a-half-year friendship was at an end. It had come
with such suddenness, I couldn't quite handle it at first. The long
evenings with Ayn were now a thing of the past. I was now the one to
feel a sense of betrayal.

But my pain was not entirely unmixed with relief. The pressure had
been mounting, and certain tensions between us had been increasing
steadily. Being forced to choose between friendship and truth as I saw
it (even if I saw it mistakenly), was not my way of conducting
intellectual life. I would sooner or later have had to escape from the
vise, I reflected. Perhaps it was better this way, with an outside
event precipitating the break. Sooner or later, probably sooner, I
would have been too explicitly frank or honest, and she would have had
an angry showdown with me, and that would have been that. Or so I told
myself. At any rate, along with the pain and the desolation, I felt a
sense of release from an increasing oppressiveness, which had been
inexorably tightening.

At dinner earlier that evening, when the radio announcer said that
Kennedy would not call off his blockade of Cuba even at the risk of
nuclear war, Ayn had said, "Good!" Privately I wondered whether she
had also said "Good" in connection with the break in our relations.
Perhaps she merely reflected with regret that the years of her efforts
on my behalf had been largely wasted.

At any rate, that night was the last time I ever saw her.
But I heard her once after that. In the late summer of 1968, not long
before the Big Break, Nathan phoned me in California and said "I want
to put you on the line to someone." The conversation with Ayn was very
brief. "I understand that you are presenting my philosophy to your
classes," she said. I replied that I was—I considered Ayn's views in
several of my courses, without thereby implying that I did so with
total agreement. She seemed gratified, and wondered how I was, and
then turned the telephone back to Nathan.
I thought of her endlessly during the years. Her enthusiasm for ideas,
her intensity, her unfailing bluntness and those piercing eyes—the
image of these things was never far away from me, especially when I
assigned some of her essays in my classes and discussed them with
students point by point. But I never regretted that I had not been
enveloped further in the web of intellectually stifling allegiances
and entanglements, the route I had seen so many of her disciples go.
In the next few years, as her non-fiction essays appeared, I read them
avidly and made many notes and comments in the margins—points to raise
with her, questions to ask her. But of course I never got to ask them.
And then, almost fifteen years after my expulsion, I heard on the
radio that she had died. I felt, even after all these years, a
devastating sense of loss. It was hard to stay in control during my
talk at the memorial service for her in Barnsdall Park in Los Angeles.
How often, on visiting New York, I had almost stopped at her apartment
building. No, I thought, her friendships are broken but her enmities
last. It wouldn't be any good. And surely she had treated me pretty
shabbily. But I thought of her, up there in that apartment, without
Frank now, and I wanted to be mesmerized by those piercing eyes once
again, and have another all-night discussion as in the old days.
I never got up the courage to take that step. It would probably have
been useless.  The occasion is past, and the past is gone forever.
That, I thought to myself with a certain grim irony, is at least one
necessary proposition to which she would have given her assent.