Determinism and freedom in the age of modern science

BS37

$1.f.

Determinism
and Freedom J

In the Age of Modern Science
Edited by Sidney Hook

William Barrett, iVIax Black, Brand Blanshard,
Ernest Nagel, F. S. C. Morthrop and many others

Is man responsible for his

actions? Does present

scientific knowledge modify

concepts of law and chance?

%-^:

>*> .; ^

W

UNIVERSITY
OF FLORIDA
LIBRARIES

COLLEGE LIBRARY

Digitized by the Internet Archive

in 2010 with funding from

Lyrasis IVIembers and Sloan Foundation

http://www.archive.org/details/determinismfreed01c2newy

Determinism and Freedom in the
Age of Modern Science

SIDNEY HOOK, Editor

c-;^

o)rvinn

ill

//,■ :

COLLIER BOOKS
NEW YORK, N.Ya

This Collier Books edition is published by arrangement with New
York University Press.

Collier Books is a division of The Crowell-Collier Publishing
Company.

First Collier Books Edition 1961

© 1958 by New York University

Hecho en Ids E.E.U.U.

Printed in the United States of America

The contents of this volume comprise the Proceedings
of the first annual New York University Institute of
Philosophy, held at Washington Square, New York,
February 9 and 10, 1957.

Preface

The new york University Institute of Philosophy is an experi-
ment designed to further fruitful discussion in philosophy.
The annual regional meetings of the American Philosophical
Association perform an excellent and indispensable profes-
sional function. But because of the large number of persons
involved, the broad scope of the programs, and the adherence
to fixed time schedules — natural limitations of all gatherings
of this kind — it seems desirable to multiply opportunities for
free and sustained interchange of views. Toward this end we
therefore resolved to try something new: to select one philo-
sophical topic or theme for intensive discussion by a small
group of philosophers and other scholars and thinkers deeply
concerned with it, and willing to explore it together around
a long seminar table. We hope our action will inspire similar
efforts in other regions of the country.

Our great problem was to keep the number of participants
within reasonable compass without making invidious distinc-
tions. No inference is to be made as to a philosopher's emi-
nence or professional qualification to analyze the theme under
discussion on the basis of his absence from this Institute. For
obvious reasons geographical considerations played a large
role. Nonetheless we hope to establish a revolving member-
ship for our Institute so that before long all philosophers with
a strong interest in the themes under discussion and within
easy traveling distance from New York City will on one or
another occasion join us as participants in the Institute.

On the other hand, the choice of a theme for our first Insti-
tute was not very much a problem. "Determinism and Free-
dom" is not only a perennial philosophical issue but seems
today to be moving once more into the forefront of intellec-
tual concern. For example, almost contemporary with our
Proceedings, the January 1957 issue of Mind contains two
long articles on the subject, and the issue before that includes
a piece in which it is argued that the entire notion of "moral
responsibility" is moribund and should be extruded from
vocabulary of intelligible expressions. ^ Our theme was selected
long before these and similar articles appeared.

Not only professional philosophers but also our colleagues

^ W. I. Matson, "On the Irrelevance of Free Will to Moral Responsibility,
and the Vacuity of the Latter," Mind (October 1956).

7

8 / Preface

in other fields are astir over the issue. Three converging tend-
encies of thought have contributed to a revival of interest in
the question of determinism and freedom.

The first is political and social. Whereas in the past the ex-
tension of the deterministic philosophy in the natural sciences
was hailed as a support of human freedom because it in-
creased man's power of control over nature, today belief in
determinism in the social sciences and social affairs is feared
by many because it increases the power of men to control
other men. Some individuals confess that it is not only Or-
well's scientifically controlled 1984 that causes their hackles
to rise in fear but even the picture of comparatively sunny
communities like that drawn in Skinner's Walden II. In some
quarters this reaction has gone so far that even the term
"planning" and the notion of a "planned society," which a
short generation ago was the hallmark of a rational social
philosophy, are viewed with suspicion as suggesting, if not
evidencing, a conspiracy against human freedom.

The second reason for a revival of interest in our theme is
that in the very stronghold of traditional determinism, the
natural sciences, the field in which determinism celebrated its
greatest triumph, the belief in the doctrine or postulate of
universal determinism seems to have been surrendered in an
effort to understand the world of subatomic behavior. The
actual relevance of these glad tidings to the momentous issues
of human freedom and responsibility remains to be estab-
lished. But in the writings of several eminent scientists it has
been eloquently proclaimed.

The third reason for the actualite of our theme derives from
growth of interest in modern psychology, psychiatry, and
psychoanalysis in all their scientific and mythological forms.
The apparent upshot of the acceptance of determinism in ex-
plaining the human psyche is the belief that the more we learn
about a man's past history, the less he seems responsible for
his present behavior. This conclusion has affected thinking
and practice in law, pedagogy, and social work, and has pro-
duced something of a revolution in penology. Sometimes its
proponents rather inconsistently blame us for blaming Hitler
and Stalin for the crimes they voluntarily committed, on the
ground that since Hitler and Stalin were once babies — hard as
it is to imagine — they must have inherited or acquired the
comr'exes and obsessional drives that caused them to do what
they couldn't help doing. Even when pruned of inconsistency,
this argument threatens to produce a revolution in moral

Preface / 9

theory by asserting that the concept of moral responsibility is
completely vacuous.

The strategy of our discussion it to begin with an analysis
of the general concept of determinism, to proceed to the
notion of determinism in physics, and to conclude with a dis-
cussion of freedom and responsibility in law and ethics.

The papers of the readers and selected commentators are
here reproduced substantially in the form in which they were
presented to the Institute. Most of the other comments by
participants in the Institute are elaborations of comments and
criticisms made after the papers were presented. Some of the
comments are a result of post-Institute reflection on the
spirited give and take among the participants, which technical
difficulties prevented us from reproducing in their entirety.

Sidney Hook

New York University

INTRODUCTION 15

by President Carroll V. Newsom,
New York University

Part I Determinism in Philosophy

> 1 THE .CASE FOR DETERMINISM 19

by Brand Blanshard, Yale University

2 MAKING SOMETHING HAPPEN 31

by Max Black, Cornell University

3 DETERMINISM AND NOVELTY 46

by William Barrett, New York University
Part II Determinism in Modem Science

1 DETERMINISM IN MODERN SCIENCE 57

by Percy W. Bridgman, Harvard University

2 THE RELATIVITY OF DETERMINISM 76

by Milton K. Munitz, New York University

3 THE CASE FOR INDETERMINISM 83

by Alfred Lande, Ohio State University

4 DETERMINISM AND THE COSMOS 90

by Dennis W. Sciama, Trinity College,
Cambridge University

Part III Determinism and Responsibility
in Law and Ethics

1 LEGAL RESPONSIBILITY AND EXCUSES 95

by H. L. A. Hart, Oxford University

V 2 HARD AND SOFT DETERMINISM 117

by Paul Edwards, New York University

N 3 WHAT MEANS THIS FREEDOM? 126

vy^ by John Hospers, Brooklyn College

12 / Contents
Part IV Discussion

1 "excusing conditions" and moral responsibility 145

by Elizabeth Lane Beardsley, Lincoln
University

2 DETERMINISM AND THE JUSTIFIABILTY OF MORAL 149
BLAME

by Richard Brandt, Swarthmore College

V 3 DETERMINISM AND PUNISHMENT 155

by Percy W. Bridgman, Harvard University

4 RESPONSIBILITY AND AVOIDABILrTY 157

by Roderick W. Chisholm, Brown University

N5 DETERMINISM, FREEDOM, AND RESPONSIBILITY 160

by C. J. Ducasse, Brown University

6 SOME REFLECTIONS ON "THE CASE FOR 170
DETERMINISM"

By Carl G. Hempel, Princeton University

7 SOME FURTHER REFLECTIONS ON MORAL 176
RESPONSIBILITY

by Howard W, Hintz, Brooklyn College

8 NECESSITY, INDETERMINISM, AND SENTIMENTALISM 180

by Sidney Hook, New York University

9 PUNISHMENT AS JUSTICE AND AS PRICE; ON 193

RANDOMNESS

by Abba Lemer, Roosevelt College

10 SOME NOTES ON DETERMINISM 196

by Ernest Nagel, Columbia University

11 CAUSATION, DETERMINISM, AND THE "GOOD" 201

by F. S. C. Northrop, School of Law, Yale
University

12 DETERMINISM, FREEDOM, MORAL RESPONSIBILITY, 212
AND CAUSAL TALK

by Arthur Pap, Yale University

13 SOME EQUIVOCATIONS OF THE NOTION OF 219
RESPONSIBILITY

by Alfred Schultz, New School for
Social Research

Contents / 13

14 OBSERVATIONS 222

by Dennis W. Sciama, Trinity College,
Cambridge University

15 DETERMINISM AND THE THEORY OF AGENCY 224

by Richard Taylor, Brown University

16 COMMON SENSE AND BEYOND 231

by Paul Weiss, Yale University

17 ON CAUSATION 237

by H. Van R. Wilson, Brooklyn College

INDEX 245

{From the welcoming address to participants in the first New York University
Institute of Philisophy)

In behalf of New York University it gives me great pleasure
to welcome the participants to this the first meeting of the
New York University Institute of Philosophy. A university
exists to achieve many purposes. But however varied its pur-
poses, the spirit of the university is found at its best only
where men are found thinking together. Thinking together is
something more than explicit teaching and learning, although
it may be that, too. It is a process of mutual stimulus and
response in which our minds become clearer, even when they
are not altered by the give and take of intellectual exchange.
It is a process whose outcome determines whether our ideas
— most of them held with equal initial conviction — are firm
principles or only familiar prejudices. Sometimes thinking
together is not only a dialogue that contributes to conceptual
clarification and self-understanding but a common voyage of
intellectual discovery. Even when discoveries are made by a
solitary thinker, they are not seldom influenced by the funded
results of previous thinking together.

The New York University Institute of Philosophy has in-
vited not only distinguished philosophers from other leading
universities to these deliberations but representatives of other
disciplines. This expresses a unity or convergence of interest
on common problems, not a unity of doctrine or point of
view. In a world where specialization tends to make every
scholar in the university an expert and none able to under-
stand or talk to his colleague, opportunities for intellectual
co-operation should be encouraged.

I therefore hope that this series of meetings will be the first
of a long line of similar proceedings contributing to the en-
richment of intellectual life in America.

Carroll V. Newsom

President, New York University

15

PART I

Determinism in Philosophy

Chapter 1

The Case for Determiiissm

Brand Blanshard, Yale University

I AM A DETERMiNisT. NoDc of the arguments oflfered on the
other side seem of much weight except one form of the moral
argument, and that itself is far from decisive. Perhaps the
most useful thing I can do in this paper is explain why the
cormnoner arguments for indeterminism do not, to my mind,
carry conviction. In the course of this explanation the brand
of determinism to which I am inclined should become gradu-
ally apparent.

But first a definition or two. Determinism is easier to de-
fine than indeterminism, and at first glance there seems to be
no difficulty in saying what one means by it. It is the view that ^
all events are caused. But unless one also says what one means
by "event" and "caused," there is Ukely to be trouble later.
Do I include among events not only changes but the lack of
change, not only the fall of the water over the cataract's edge,
but the persistence of ice in the frozen river? The answer is
"Yes." By an event I mean any change or persistence of state
or position. And what is meant by saying that an event is
caused? The natural answer is that the event is so connected
with some preceding event that unless the latter had occurred
the former would not have occurred. Indeterminism means
the denial of this. And the denial of this is the statement that
there is at least one event to which no preceding event is
necessary. But that gets us into trouble at once, for it is doubt-
ful if any indeterminist would want to make such an assertion.
What he wants to say is that his decision to tell the truth is
undetermined, not that there is no preceding event necessary
to it. He would not contend, for example, that he could tell
the truth if he had never been bom. No, the casual statement
to which the indeterminist takes exception is a different one.
He is not saying that there is any event to which some nam-
able antecedents are not necessary; he is saying that there are
some events whose antecedents do not make them necessary.
I He is not denying that all consequents have necessary ante-
cedents; he is denying that all antecedents have necessary
consequents.' He is saying that the state of things just before

19

20 / Determinism and Freedom

he decided to tell the truth might have been exactly what it
was and yet he might have decided to tell a lie.

By determinism, then, I mean the view that every event A
is so C9nnected with a later event B that, given A, B must
occur. ByCindeterminism I mean the view that there is some
event B that is not so connected with any previous event A
that, given A, it must occur. Now, what is meant here by
"must"? We cannot in the end evade that question, but I hope
you will not take it as an evasion if at this point I am content
to let you fill in the blank in any way you wish. Make it a
logical "must," if you care to, or a physical or metaphysical
"must," or even the watered-down "must" that means "A is
always in fact followed by B." We can discuss the issue use-
fully though we leave ourselves some latitude on this point.

With these definitions in mind, let us ask what are the most
important grounds for indeterminism. This is not the same as
asking what commonly moves people to be indeterminists; the
answer to that seems to me all too easy. Everyone vaguely
knows that to be undetermined is to be free, and everyone
wants to be free. My question is rather. When reflective
people accept the indeterminist view nowadays, what con-
siderations seem most cogent to them? It seems to me that
there are three: first, the stubborn feeling of freedom, which
seems to resist all dialectical solvents; second, the conviction
that natural science itself has now gone over to the indeter-
minist side; and, third, that determinism would make nonsense
of moral responsibility. The third of these seems to me the
most important, but I must try to explain why none of them
seem to me conclusive.

One of the clearest heads that ever devoted itself to this old
issue was Henry Sidgwick. Sidgwick noted that, if at any given
moment we stop to think about it, we always feel as if more
than one course were open to us, that we could speak or be
silent, lift our hand or not lift it. If the determinist is right,
this must be an illusion, of course, for whatever we might
have done, there must have been a cause, given which we had
to do what we did. Now, a mere intuitive assurance about our-
selves may be a very weak ground for belief; Freud has shown
us that we may be profoundly deceived about how we really
feel or why we act as we do. But the curious point is that,
though a man who hates his father without knowing it can
usually be shown that he does and can often be cured of his
feeling, no amount of dialectic seems to shake our feeling of
being free to perform either of two proposed acts. By this

The Case for Determinism / 21

feeling of being free I do not mean merely the freedom to do
what we choose. No one on either side questions that we have
that sort of freedom, but it is obviously not the sort of free-
dom that the indeterminist wants, since it is consistent with
determinism of the most rigid sort. The real issue, so far as
the will is concerned, is not whether we can do what we
choose to do, but whether we can choose our own choice,
whether the choice itself issues in accordance with law from
some antecedent. And the feeling of freedom that is relevant
as evidence is the feeling of an open future as regards the
choice itself. After the noise of argument has died down, a
sort of intuition stubbornly remains that we can not only lift
our hand if we choose, but that the choice itself is open to us.
It this not an impressive fact?

No, I do not think it is. The first reason is that when we
are making a choice our faces are always turned toward the
future, toward the consequences that one act or the other
will bring us, never toward the past with its possible sources
of constraint. Hence these sources are not noticed. Hence we
remain unaware that we are under constraint at all. Hence we
feel free from such constraint. The case is almost as simple as
that. When you consider buying a new typewriter your thought
is fixed on the pleasure and advantage you would gain from it,
or the drain it would make on your budget. You are not delv-
ing into the causes that led to your taking pleasure in the
prospect of owning a typewriter or to your having a complex
about expenditure. You are too much preoccupied with the
ends to which the choice would be a means to give any atten-
tion to the causes of which your choice may be an effect. But
that is no reason for thinking that if you did preoccupy your-
self with these causes you would not find them at work. You
may remember that Sir Francis Galton was so much im-
pressed with this possibility that for some time he kept account
in a notebook of the occasions on which he made important
choices with a full measure of this feeling of freedom; then
shortly after each choice he turned his eye backward in search
of constraints that might have been acting on him stealthily.
He found it so easy to bring such constraining factors to light
that he surrendered to the determinist view.

But this, you may say, is not enough. Our preoccupation
with the future may show why we are not aware of the con-
straints acting on us, and hence why we do not feel bound by
them; it does not explain why our sense of freedom persists
after the constraints are disclosed to us. By disclosing the

22 / Determinism and Freedom

causes of some fear, for example, psychoanalytic therapy can
remove the fear, and when these causes are brought to light,
the fear commonly does go. How is it, then, that when the
causes of our volition are brought to light volition continues
to feel as free as before? Does this not show that it is really
independent of those causes?

No again. The two cases are not parallel. The man with the
panic fear of dogs is investing all dogs with the qualities —
remembered, though in disguised form — of the monster that
frightened him as a child. When this monster and his relation
to it are brought to light, so that they can be dissociated from
the Fidos and Towsers around him, the fear goes, because its
appropriate object has gone. It is quite different with our feel-
ing of freedom. We feel free, it was suggested, because we are
not aware of the forces acting on us. Now, in spite of the
determinist's conviction that when a choice is made there are
always causal influences at work, he does not pretend to reveal
the influences at work in our present choice. The chooser's
face is always turned forward; his present choice is always
unique; and no matter how much he knows about the will and
the laws, his present choice always emerges out of deep
shadow. The determinist who buys a typewriter is as little
interested at the moment in the strings that may be pulling at
him from his physiological or subconscious cellars as his
indeterminist colleague, and hence feels just as free. Thus,
whereas the new knowledge gained through psychoanalysis
does remove the grounds of fear, the knowledge gained by the
determinist is not at ail of the sort that would remove the
grounds for the feeling of freedom. To make the persistence
of this feeling in the determinist an argument against his case
is therefore a confusion.

The second reason, I suggested, why so many thoughtful
persons remain indeterminists is that they are convinced that
science has gone indeterminist. Well, has it? If you follow
Heisenberg, Eddington, and Born, it has. If you follow Rus-
sell, Planck, and Einstein, it has not. When such experts dis-
agree it is no doubt folly for the layman to rush in. But since
I am discussing the main reasons why people stick to indeter-
minism, and have admitted that the new physics is one of
them, I cannot afford to be quite prudent. Let me say, then,
with much hesitation that, as far as I can follow the argu-
ment, it provides no good evidence for indeterminism even in
the physical world, and that, if it did, it would provide no
good evidence for indeterminism in the realm of will.

The Case for Determinism / 23

First as to physical indeterminism. Physicists now tell us
that descriptive statements about the behavior of bodies are
really statistical statements. It was known long ago that the
pressure that makes a football hard is not the simple quality
one feels in pushing something: it is the beating on the inner
surface of the football of millions of molecular bullets. We now
know that each of these bulelts is a swarm of atoms, them-
selves normally swarms of still minuter somethings, of which
the proton and the electron are typical. The physicist admits
that the behavior of an enormous mass of these particles, such
as a billiard ball, is so stable that we may safely consider it as
governed by causal law. But that is no reason, he adds, for
assigning a like stability to the ultimate particles themselves.
Indeed, there is good reason, namely the principle of indeter-
minacy, for saying that they sometimes act by mere chance.
That principle tells us that whereas, when we are talking about
a billiard ball, we can say that it has a certain momentum and
direction at point S as a result of having a certain momen-
tum and direction at point A, we can never say that sort of
thing about an electron. Why? Because the conditions of ob-
servation are such that, when they allow us to fix the position
exactly, they make it impossible to fix the momentum exactly.
Suppose that we can determine the position of a moving par-
ticle with more accuracy the shorter the wave length of light
we use. But suppose that the shorter the wave length, the more
it interferes with the momentum of the particle, making it leap
unpredictably about. And suppose there is no way of deter-
mining the position without in this way leaving the momentum
vague, or of determining the momentum without leaving the
position vague. It will then be impossible to state any precise
law that governs the particle's movement. We can never say
that such-and-such a momentum at point A was necessarily
followed by such-and-such a momentum at point B, because
these statements can have no precise meaning, and can be
given none, for either antecedent or consequent. Hence to
speak any longer of nature as governed ultimately by causal
laws — i.e., statements of precise connection between ante-
cedent and consequent — is simply out of the question.

This argument, as Sir David Ross has pointed out, may be
interpreted in different ways. It may be taken to say that,
though the particle does have a certain position and momen-
tum, we can never tell, definitely and for both at the same
time, what they are. Many interpreters thus understand the
theory. But so taken, there is of course nothing in it to throw

24 / Determinism and Freedom

the slightest doubt on the reign of causality. It is merely a
statement that in a certain region our knowledge of causal
law has limits. Secondly, the theory might be taken to mean
that electrons are not the sort of things that have position and
momentum at all in the ordinary sense, but are fields, per-
haps, or widespreading waves. This, too, has no suggestion of
indeterminism. It would not mean that general statements
about the nature and behavior of electrons could not be made,
but only that such statements would not contain references to
position and momentum. Thirdly, the theory might mean that,
though these particles do have a position and a momentum,
the position or momentum is not definitely this rather than
that. Even laymen must rise at this point and protest, with all
respect, that this is meaningless. Vagueness in our thought of
a position makes sense; vagueness of actual position makes
none. Or, finally, the argument may mean that, though the
particle does have a definite position and momentum, these
cannot even in theory be correlated with anything that went
before. But how could we possibly know this? The only
ground for accepting it is that we do not know of any such
correlates. And that is no reason for denying that any exist.
Indeed, to deny this is to abandon the established assumption
and practice of science. Science has advanced in the past pre-
cisely because, when things happened whose causes were un-
known, it was assumed that they had causes nevertheless. To
assume that a frustration of present knowledge, even one that
looks permanent, is a sign of chance in nature is both practic-
ally uncourageous and theoretically a non sequitur.

But let us supfKDse that the Eddingtonians are right and that
what has been called "free will among the electrons" is the
fact. Would that imply indeterminism in the realm that most
nearly concerns us, the realm of choice? I cannot see that it
would. The argument supposed to show that it would is as
follows: Psychical processes depend on physical processes.
But physical processes are themselves at bottom unpredict-
able. Hence the psychical processes dependent on them must
share this unpredictability. Stated in the abstract, the argu-
ment sounds impressive. But what does it actually come to?
We are told that, even if there is inconstancy in the behavior
of single particles, there is no observable inconstancy in the
behavior of masses of them; the particles of a billiard ball are
never able to get together and go on a spree simultaneously.

The Case for Determinism / 25

Eddington admitted that they might, just as he admitted that
an army of monkeys with a million typewriters might produce
all the books in the British Museum, but he admitted also that
the chance of a billiard ball's behaving in this way were so
astronomically remote that he would not believe it if he saw it.

The question of importance for us, then, is whether, if acts
of choice are dependent on physical processes at all, they de-
pend on the behavior of particles singly or on that of masses of
particles. To this there can be but one answer. They depend
on mass behavior. An act of choice is an extremely complex
process. It involves the idea of one or more ends, the associa-
tion of that idea with more or less numerous other ideas, the
presence of desires and repulsions, and the operation of habits
and impulses; indeed, in those choices for which freedom is
most demanded, the whole personality seems to be at work.
The cortical basis for so complex a process must be extremely
broad. But if it is, the great mass of cells involved must, by the
physicist's admission, act with a high stability, and the corre-
lated psychical processes must show a similar stability. But
that is what we mean by action in accordance with causal
law. So, even if the physicists are right about the unstable be-
havior of single particles, there is no reason whatever for
translating this theory into a doctrine of indeterminism for
human choice.

We come now to the third of the reasons commonly ad-
vanced in support of indeterminism. This is that determinism
makes a mess of morality. The charge has taken many forms.
We are told that determinism makes praise and blame mean-
ingless, punishment brutal, remorse pointless, amendment
hopeless, duty a deceit. All these allegations have been effec-
tively answered except the one about duty, where I admit I
am not quite satisfied. But none of them are in the form in
which determinism most troubles the plain man. What most
affronts him, I think, is the suggestion that he is only a ma-
chine, a big foolish clock that seems to itself to be acting
freely, but whose movements are controlled completely by
the wheels and weights inside, a Punch-and-Judy show whose
appearance of doing things because they are right or reason-
able is a sham because everything is mechanically regulated
by wires from below. He has no objections to determinism as
applied by physicists to atoms, by himself to machines, or by
his doctor to his body. He has an emphatic objection to deter-
minism as applied by anyone to his reflection and his will, for

26 / Determinism and Freedom

this seems to make him a gigantic mechanical toy, or worse,
a sort of Frankenstein monster.

In this objection I think we must agree wtih the plain man.
If anyone were to show me that determinism involved either
materialism or mechanism, I would renounce it at once, for
that would be equivalent, in my opinion, to reducing it to
absurdity. The "physicalism" once proposed by Neurath and
Carnap as a basis for the scientific study of behavior I could
not accept for a moment, because it is so dogmatically anti-
empirical. To use empirical methods means, for me, not to ap-
proach nature with a preconceived notion as to what facts
must be like, but to be ready to consider all kinds of alleged
facts on their merits. Among these the introspectively observ-
able fact or reflective choice, and the inference to its existence
in others, are particularly plain, however different from any-
thing that occurs in the realm of the material or the publicly
observable or the mechanically controlled.

Now, what can be meant by saying that such choice, though
not determined mechanically, is still determined? Are you
suggesting, it will be asked, that in the realm of reflection and
choice there operates a different kind of causality from any we
know in the realm of bodies? My answer is: Yes, just that. To
put it more particularly, I am suggesting ( 1 ) that even within
the psychical realm there are different causal levels, (2) that
a causality of higher level may supervene on one of lower
level, and (3) that when causality of the highest level is at
work, we have precisely what the indeterminists, without
knowing it, want.

1. First, then, as to causal levels. I am assuming that even
the indeterminist would admit that most mental events are
causally governed. No one would want to deny that his
stepping on a tack had something to do with his feeling pain,
or that his touching a flame had something to do with his get-
ting burned, or that his later thought of the flame had some-
thing to do with his experience of its hotness. A law of associ-
ation is a causal law of mental events. In one respect it is like
a law of physical events: in neither case have we any light as
to why the consequent follows on the antecedent. Hume was
right about the billiard balls. He was right about the flame
and the heat; we do not see why something bright and yellow
should also be hot. He was right about association; we do not
understand how one idea calls up another; we only know that
it does. Causality in all such cases means to us little if any-
thing more than a routine of regular sequence.

The Case for Determinism / 27

Is all mental causation like that? Surely not. Consider a
musician composing a piece or a logician making a deduction.
Let us make our musician a philosopher also, who after add-
ing a bar pauses to ask himself, "Why did I add just that?"
Can we believe he would answer, "Because whenever in the
past I have had the preceding bars in mind, they have always
been followed by this bar"? What makes this suggestion so
inept is partly that he may never have thought of the preced-
ing bars before, partly that, if he had, the repetition of an
old sequence would be precisely what he would avoid. No, his
answer, I think, would be something like this: "I wrote what
I did because it seemed the right thing to do. I developed my
theme in the manner demanded to carry it through in an
aesthetically satisfactory way." In other words, the constraint
that was really at work in him was not that of association; it
was something that worked distinctly against association; it
was the constraint of an aesthetic ideal. And, if so, there is a
causality of a different level. It is idle to say that the musician
is wholly in the dark about it. He can see not only that B
succeeded A; as he looks back, he can see in large measure
why it did.

It is the same with logical inference, only more clearly so.
The thinker starts, let us say, with the idea of a regular solid
whose faces are squares, and proceeds to develop in thought
the further characteristics that such a solid must possess. He
constructs it in imagination and then sees that it must have six
faces, eight vertices, and twelve edges. Is this association
merely? It may be. It is, for example, if he merely does in
imagination what a child does when it counts the edges on a
lump of sugar. This is not inference and does not feel like it.
When a person, starting with the thought of a solid with
square faces, deduces that it must have eight vertices, and
then asks why he should have thought of that, the natural
answer is, Because the first property entails the second. Of
course this is not the only condition, but it seems to me con-
trary to introspectively plain fact to say that it had nothing to
do with the movement of thought. It is easy to put this in such
a way as to invite attack. If we say that the condition of our
thinking of B is the observed necessity between A and B, we
are assuming that B is already thought of as a means of ex-
plaining how it comes to be thought of. But that is not what I
am saying. I am saying that in thinking at its best thought
comes under the constraint of necessities in its object, so that
the objective fact that A necessitates B partially determines

28 / Determinism and Freedom

our passing in thought from A to B. Even when the explana-
tion is put in this form, the objection has been raised that
necessity is a timeless link between concepts, while causality
is a temporal bond between events, and that the two must be
kept sharply apart. To which the answer is: Distinct, yes; but
always apart, no. A timeless relation may serve perfectly weU
as the condition of a temporal passage. I hold that in the
course of our thinking we can easily verify this fact, and, be-
cause I do, I am not put off by pronouncements about what
we should and should not be able to see.

2. My second point about the causal levels is that our
mental processes seldom move on one level alone. The higher
is always supervening on the lower and taking over partial
control. Though brokenly and imperfectly rational, rational
creatures we still are. It must be admitted that most of our so-
called thinking moves by association, and is hardly thinking
at all. But even in the dullest of us "bright shoots of everlast-
ingness," strands of necessity, aesthetic or logical, from time
to time appear. "The quarto and folio editions of mankind"
can follow the argument with fewer lapses than most of us;
in the texts of the greatest of all dramas, we are told, there
was seldom a blot or erasure; but Ben Jonson added, and no
doubt rightly, that there ought to have been a thousand. The
effort of both thought and art is to escape the arbitrary, the
merely personal, everything that, causal and capricious, is ir-
relevant, and to keep to lines appointed by the whole that one
is constructing. I do not suggest that logical and aesthetic
necessity are the same. I do say that they are both to be dis-
tinguished from association or habit as representing a different
level of control. That control is never complete; all creation
in thought or art is successful in degree only. It is successful
in the degree to which it ceases to be an expression of merely
personal impulses and becomes the instrument of a necessity
lying in its own subject matter.

3. This brings us to our last point. Since moral choice, like
thought and art, moves on different causal levels, it achieves
freedom, just as they do, only when it is determined by its
own appropriate necessity. Most of our so-called choices are
so clearly brought about by association, impulse, and feeling
that the judicious indeterminist wiU raise no issue about them.
When we decide to get a drink of water, to take another nibble
of chocolate, to go to bed at the usual hour, the forces at
work are too plain to be denied. It is not acts like these on
which the indeterminist takes his stand. It is rather on those

The Case for Determinism / 29

where, with habit, impulse, and association prompting us
powerfully to do X, we see that we ought to do Y and there-
fore do it. To suppose that in such cases we are still the pup-
pets of habit and impulse seems to the indeterminist palpably
false.

So it does to us. Surely about this the indeterminist is right.
Action impelled by the sense of duty, as Kant perceived, is ac-
tion on a different level from anything mechanical or associ-
ative. But Kant was mistaken in supposing that when we were
determined by reason we were not determined at all. This
supposition seems to me wholly unwarranted. The determina-
tion is still there, but, since it is a determination by the moral
necessities of the case, it is just what the moral man wants
and thus is the equivalent of freedom. For the moral man,
like the logician and the artist, is really seeking self-surrender.
Through him as through the others an impersonal ideal is
working, and to the extent that this ideal takes possession of
him and molds him according to its pattern, he feels free and
is free.

The logician is most fully himself when the wind gets into
his sails and carries him effortlessly along the line of his calcu-
lations. Many an artist and musician have left it on record that
their best work was done when the whole they were creating
took the brush or pen away from them and completed the
work itself. It determined them, but they were free, because
to be determined by this whole was at once the secret of their
craft and the end of their desire. This is the condition of the
moral man also. He has caught a vision, dimmer perhaps than
that of the logician or the artist, but equally objective and
compelling. It is a vision of the good. This good necessitates
certain things, not as means to ends merely, for that is not
usually a necessary link, but as integral parts of itself. It re-
quires that he should put love above hate, that he should re-
gard his neighbor's good as of like value with his own, that he
should repair injuries, and express gratitude, and respect prom-
ises, and revere truth. Of course it does not guide him infall-
ibly. On the values of a particular case he may easily be mis-
taken. But that no more shows that there are no values present
to be estimated, and no ideal demanding a special mode of
action, than the fact that we make a mistake in adding figures
shows that there are no figures to be added, or a right way of
adding them. In both instances what we want is control by
the objective requirements of the case. The saint, like the
thinker and the artist, has often said this in so many words.

30 / Determinism and Freedom

I feel most free, said St. Paul, precisely when I am most a
slave.

We have now dealt, as best we can in a restricted space,
with the three commonest objections to determinism. They all
seem to admit of answers. To the objection that we always feel
free, we answer that it is natural to feel so, even if we are
determined, since our faces are set toward results and not to-
ward causes, and the causes of present action always elude us.
To the objection that science has gone indeterminist, we an-
swer that that is only one interpretation of recent discoveries,
and not the most plausible one, and that, even if it were true,
it would not carry with it indeterminism for human choice.
To the objection that determinism would reduce us to the level
of mechanical puppets, we answer that though we are puppets
in part we hve, as Aristotle said, on various levels. And so far
as causality in reflection, art, and moral choice involves con-
trol by immanent ideal, mechanism has passed over into that
[rational determinism that is the best kind of freedom.

Chapter 2

Max Black, Cornell University

I

You ARE TmRSTY, but there is a glass of beer within easy
reach; you stretch out your hand, bring the glass to your lips,
and drink. Here is what I call a perfectly clear case of making
something happen. When you brought the glass nearer, that
was a perfect instance of what all of us call "making some-
thing happen."^ But of course many other simple actions
would serve just as well: closing a window, opening a drawer,
turning a doorknob, sharpening a pencil. Any number of per-
fectly clear cases can be found of making something happen.

The following is not a clear case of making something hap-
pen. On hearing the opening of this paper, a member of the
audience leaves the room, to be found later in the nearest
saloon. To establish that my remarks made him leave the
room would require a specific investigation. Evidence could
be obtained for or against the view that talk about drinking
had made the hearer leave; until such evidence had been pro-
vided, the final verdict would remain in doubt.

In the case of the thirsty man reaching for the glass, an
investigation to determine whether or not he really did move
the glass would be out of place. There would be an absurdity
in saying that evidence could be provided for or against the
view that he had moved the glass; or in saying that whether
he had made anything happen was a hypothesis. It would be
absurd to say that there was a question whether he had moved
the glass, and that the answer would be undecided until
further evidence had been weighed.

For what could be the goal of the supposed investigation?

* Or rather a clear case of "moving a glass." The expression "making
something happen" is introduced for brevity in referring to similar cases.
The first part of the paper investigates a class of transitive verbs, like "mov-
ing," "breaking," "opening," "upsetting," etc., indicated by the blanket ex-
pression "making something happen." When the expression "making some-
thing happen" occurs, the reader may usually imagine the more specific
expression "moving a glass" substituted — with the understanding, however,
that the discussion is intended to apply indifferently to an entire class of
similar expressions.

31

32 / Determinism and Freedom

If somebody is not already satisfied that the familiar episode
is a case of what we ordinarily call "making something hap-
pen," it is inconceivable that further empirical evidence would
satisfy him. The supposed investigation would have no termi-
nus; criteria would be lacking by which to judge the relevance
and strength of testimony.

I am trying to affirm something noncontroversial and hence
acceptable in advance of any philosophical analysis or com-
mitment. I am contending that we do all treat simple episodes
like the one I described as perfectly clear cases of making
something happen.

We do in fact recognize the absurdity of a supposed attempt
to find out whether the drinker had made the glass move.
Suppose I were to say to somebody, "You saw that man reach
out for that glass of beer just now — well, find out whether
he moved the glass." I have no doubt that a layman would be
dumbfounded and quite at a loss to know what could be
meant. A sufficient reply would be : "Surely, we saw him move
the glass." If I insisted that I wanted evidence that the drinker
had moved the glass, my interlocutor might begin to suspect
the situation was abnormal — for this would be one way of
making sense of my demand. Suppose we were suddenly to
see the glass of beer levitate and fly like a homing pigeon
straight to the drinker's mouth! Then we might rub our eyes
and begin to wonder whether the man in the armchair had
really moved the glass the first time. We might then plausibly
suspect ourselves in some magician's establishment, well
stocked with trick devices for making objects move in extraor-
dinary ways. This would be a fantastically abnormal situ-
ation. In describing the case of the thirsty drinker I wished to
present a situation that was normal, one whose description was
intended to exclude monstrosities and miracle.

I was therefore taking for granted that the person con-
cerned was neither hypnotized nor walking in his sleep nor
obeying a neurotic compulsion to reach for the glass nor act-
ing in response to threats. And I was assuming that the glass
of beer was an ordinary vessel, having no concealed magnet
or other special devices and subject to no remote physical or
mental controls. In short, I was taking for granted that the
exemplary situation was a perfectly familiar, ordinary case. If
a situation were not of this familiar sort, it might be necessary
to investigate and find out whether the man concerned really
had made something happen.

So far, I have been contrasting a perfectly clear case of

Making Something Happen / 33

making something happen with cases in which an investiga-
tion would be in order. The latter we might call problematic
cases. A second kind of contrast could be made between a
perfectly clear case and a borderline case.

Suppose you jogged my hand, so that my elbow knocked
against the glass and spilled its contents. Did / spill the glass,
or did you in fact do it? Both anwsers are plausible. We are
inclined to say something like, "I spilled the beer all right, but
you made me do it, so really you spilled it." Here the presence
of the qualification "really" in "really you spilled it" is a sign
that criteria for the use of the expression "making something
happen" are no longer precise and determinate. We would not
teach a child what "making something happen" means by cit-
ing this kind of case or a case in which somebody's involun-
tary gesture displaced an object. Similarly, we would not teach
somebody the meaning of "orange" by showing color patches
that most of us should hesitate to label either orange or
yellow.

The uncertainty here is not due to lack of information and
could not be removed by any empirical investigation. Uncer-
tainty of application is a feature of our uses of "orange" and
can be removed only by stipulation. Our use of the expression
"making something happen" is infected by similar uncertainty.
Only I should wish to deny that any such uncertainty of ap-
plication is to be found in the clear case that I began by
describing.

Indeed, if anybody were to show genuine hesitation about
using the expression "making something happen" in the situa-
tion described, that would be evidence that he did not really
understand that expression. If I were teaching a foreigner how
to use that English expression, a test of my success would be
his unhesitating identification of the exemplary situation as a
case of making something happen. (Of course, he must also
hesitate in borderhne cases.) Should the pupil waver in the
clear case, we might try to find out whether he suspected some
hidden mechanism or trick device, i.e., whether he mistakenly
took the situation to be abnormal. But if he convinced us that
he fully understood our description of the familiar case, yet
still did not know whether to say that something had been
made to happen, we could be sure that efforts to teach him
the uses of the English expression had not yet succeeded.

I have said that the case of the thirsty drinker is a perfectly
clear one of making something happen, leaving no room for
further empirical investigation — a case neither "problematic"

34 / Detenninism and Freedom

nor "borderline." I want to add now that the episode is also a
paradigm for application of the phrase "making something
happen." That it is a paradigm is closely connected with its
being a perfectly clear case; yet to call it a "paradigm" is to
say something new.

Suppose we are faced with something that is not a clear
case of making something happen, and wish to decide whether
or not to apply the expression. A natural recourse would be
to compare the doubtful case with some perfectly clear case,
with a view to finding sufficient similarity or dissimilarity to
arrive at a correct decision. By treating the clear case as a
standard we can base our decision to use or withhold the ex-
pression upon reasons: we appeal to the clear case to resolve
doubt. It follows, therefore, that no reason can be given why
the clear case itself should bear the identifying label in ques-
tion. There is nothing besides itself to which the clear case
can be compared — nothing else to serve as a standard. The
absurdity of asking that reasons be given for using the clear
case as a standard would be just like that of trying to give
reasons why the standard meter rod is counted as one meter
in length, or a standard color sample is accepted as "red."
Should someone demand reasons in defense of calling my
exemplary instance a case of making something happen, the
best I could do by way of reply would be to say: "That's what
I call 'making something happen.' " Now here I am not offer-
ing a genuine reason, but repudiating the demand for a rea-
son. The retort "That's what I call 'making something hap-
pen' " is a way of showing how I use the expression. In mak-
ing that retort I show that I treat the instance as a paradigm.
But showing is not arguing, and brandishing a paradigm is not
offering a reason.

The case of the thirsty drinker differs from that of the
meter rod in one important respect. The expression "one meter
long" is formally defined in terms of a standard measure,
so that those who understand how the expression is used know
that a dispute about the correctness of any attribution of met-
ric length would ultimately have to be resolved by appeal to
a known and identifiable standard of comparison. But there is
no formal definition of "making something happen," and of
course no permanent and identifiable situations to serve as
standards of comparison. We have a wide range of choice in
exhibiting "perfectly clear cases," and they are not preserved
in official bureaus of standards. Nevertheless, we do appeal to
them in case of doubt— our choice of just those situations as

Making Something Happen / 35

acceptable standards being a feature of our use of the expres-
sion in question. Instead of the unique arbiter, we have, as it
were, a reserve of available judges, any of which indifferently
can serve to remind us of our linguistic conventions. Pressed
to give reasons, we eventually stop at situations of which we
can say no more than, "That's what I call such-and-such." Our
choice of halting places shows which instances we in fact treat
as paradigms. In calling my exemplary situation a paradigm
for the use of the expression "making something happen,"
therefore, I am claiming that it is useful as a standard for the
correctness of application of the expression in question.

Paradigm cases also serve as standards of reference when
we pass from primitive uses of an expression to other uses
derived by resemblance, analogy, and metaphorical extension.
Uses of "making something happen" and cognate expressions
are strikingly various; yet the paradigm helps to illuminate all
such uses. The exemplary instance, or sufficiently similar alter-
natives, functions as a prototype for the derivative uses of
"making something happen." We refer to it in testing the
plausibility of analogy and metaphor.

As we pass from the homespun language of "making some-
thing happen" to the more sophisticated language of "cause"
and "effect," the influence of the paradigm remains powerful.
We continue to model descriptions of cases remote from the
prototypes on the simpler primitive cases, often by using
metaphors literally applicable only to those clear cases. In
order to understand clearly what we mean by "cause" and
"effect" we must labor to understand what we mean by the
precausal language in which the more sophisticated vocabulary
is embedded.

If my exemplary situation is a clear case (a paradigm, a
prototype) of making something happen, it follows that it
would be nonsensical to speak of there being any possible
doubt whether something was made to happen. This remains
true, no matter how much the original description of the situ-
ation might be augmented or elaborated by scientific explana-
tion, provided only that the additional information did not
conflict with the original assumption of "normality." A scien-
tist might explain why the pressure of the fingers required the
glass to change position without slipping through the hand;
another scientist mighit offer elaborate explanations of the
physiology of thirst; a psychologist might connect your pres-
ent thirst with childhood deprivation. But such accounts, in-
formative as they might be, would have no tendency to dis-

36 / Determinism and Freedom

credit the correctness of the use of the expression "making
something happen." They could not do so, because the de-
scription of the paradigm case is complete. If the description
left gaps to be filled by scientific data as yet unknown, none of
us would be able to use the expression "making something
happen" correctly. The expression would be a blank check
drawn on an uncertain future.

My chief contention, so far, is the commonplace one that
it is perfectly certain that persons do sometimes make some-
thing happen. It might be unnecessary to insist on anything
so obvious, had not philosophers sometimes claimed to have
arguments to show that it is logically impossible for anything
to be a cause, since the notion of a cause is self-contradictory.
Now, to make something happen is to cause something to
happen. It is certain therefore that the notion of a cause is
not self-contradictory.

n

Once we are satisfied that we have identified and sufficiently
described a paradigm for the use of a given expression, we can
proceed to look for features and criteria of application. That
is to say, we can ask, "What is it about this clear case that we
treat as relevant in using it as a standard of comparison?"
Sometimes the search for criteria leads nowhere: to the ques-
tion, "What is it about this clear case of red that makes us call
it 'red'?" there is no answer. But sometimes a demand for cri-
teria can be met. If the question is raised about our paradigm
case of making something happen, it can elicit a set of rele-
vant features, some trite and uninteresting, but others surpris-
ingly at variance with accepted analyses of causation. I shall
list some of these features and comment on a few of them. In
order to save time, I shall refer to the person who moved the
glass as "P." I shall call the object moved "O," its motion
"M," and the action performed by the agent "A."

The following assertions about the episode seem to me
plainly true:

1 What happened was made to happen by P.

2 What he made happen was a motion of O (i.e., M).

3 P made this happen by doing something (moved his
hand to O, clasped it, and brought it back to him).

4 In doing A, P was acting freely (was not in any way
being forced or constrained to do A).

Making Something Happen / 37

5 A occurred throughout the time that M was occur-
ring.

6 M (the motion of O) would not have occurred unless
A had occurred.

7 When A occurred, M had to occur.

If we used the accepted terminology of discussions of cau-
sation, we could roughly summarize the foregoing seven
points by saying that the cause was a free act of a person, the
effect was a motion of an inanimate object, the cause and
effect were cotemporal (operative through the same time
interval), and the effect was a necessary consequence. (We
might add that the cause and effect were spatially contiguous,
in a way too obvious to detail.)

I shall now comment on some of these points.

The agent acted freely. The contrast to be made here is
with forced or constrained action; and, again, with action as
an intermediary. If P had been compelled by physical coer-
cion or by threats, we should confidently say he had not acted
freely; there might be more hesitation about saying the same
if he had acted because asked to do so, or if he had expected
to receive some reward, or had some other ulterior motive.
But neither coercion nor inducement was present in our para-
digm case: P took the glass because he "just wanted to." If
he had any motive at all, it may have been that he was thirsty,
but he might equally well have had no antecedent and sepa-
rable thirst. There would be no harm in saying the act was un-
motivated— which is not to say it was irrational or unintel-
ligible. On the other hand, the presence in other cases of a
distinct and separable motive prior to the act would not dis-
qualify an episode as a clear case of making something hap-
pen. Neither presence nor absence of a separable motive func-
tions as a clear-cut criterion.

It may provoke surprise and an accusation of anthropomor-
phism that the presence of a person is insisted on as a fea-
ture of the paradigm. But the insistence is necessary. A candid
examination of causal language will show that our prototypes
involve persons. Certainly the word "make" strongly suggests
a maker; and we find it not at all unnatural to substitute
"make" for "cause." If this be anthropomorphism, we must
make the best of it.

To return to our illustration. Not only is it true that the
agent acted freely: we are entitled, I think, to add that the
very same situation is a clear case and a paradigm for acting

38 / Determinism and Freedom

freely. This means, as I previously explained, that it would
be logically absurd to demand an investigation as to whether
the agent acted freely; it also means that there could be no
doubt that he had acted freely, nor any further reason to show
that he had so acted.

Now, if this is so, it follows that so far from there being a
radical conflict between the notion of causation and freedom,
as many philosophers have insisted, the two notions, or their
informal progenitors, are logically inseparable. Our para-
digmatic conception of causing something to happen is a con-
ception of somebody freely making something happen. So
anything having a tendency to show that the agent was not act-
ing freely, but responding to contraint, duress, or ulterior in-
ducement, would immediately have a tendency to show that he
was not the cause but merely an instrument or an intermediary
between the true cause and its effects.

It also follows that no scientific elaboration of the anteced-
ents of the paradigmatic episodes could destroy their char-
acter as paradigms of acting freely, and so causing something
to happen. No physiological or psychoanalytical explanation
of the unconscious cause, if any, of P's moving O can have the
least tendency to discredit our calling his act a case of freely
making something happen. Of course the case would be
altered if such scientific elaboration led us to view his act as
pathological, but this outcome was excluded by our descrip-
tion of the paradigm.

The effective action lasted throughout the motion it pro-
duced. It has been a truism for writers on causation that the
cause must precede the effect. And certainly there would be
a logical absurdity in supposing that the cause might succeed
its effect. But our paradigm has cause and effect occurring
together. It might be objected that the initiating action, A,
began before the motion, M, it produced. But it would be
easy to define the action as lasting for exactly the same period
of time as the motion generated; we must therefore allow that
sometimes cause and effect can be simultaneous or cotem-
poral. TTiis will not render the causal relation symmetrical, as
might be feared; the desired asymmetry is here ensured by
the fact that the cause is a free action while the effect is not
an action at all but the motion of an inanimate body. And
when one person acts on another, so that one action is con-
tiguous and cotemporal with another action, we can still im-
mediately identify the cause as that action of the two that was
free. If John pushes James, John acs freely, but James doesn't;

Making Something Happen / 39

and conversely, if James was moving of his own free will,
John didn't push him.

Now cotemporality of cause and effect is not a mere peculi-
arity of "prescientific" thinking; it is a commonplace of causal
description at scientific levels, as philosophers have occasion-
ally noticed. The moon's gravitational pull lasts as long as the
tide it produces; difference of temperature registers through-
out the period that thermometric expansion occurs; a catalyst
continues to act during the chemical reaction it is influencing;
and so on, for any number of similar cases. There is some
reason to regard the principle of strict priority of the cause as
a metaphysical prejudice. And, like other metaphysical preju-
dices, it can be opposed by equally powerful metaphysical
prejudices of opposite tendency.

The induced motion would not have occurred but for the
action that produced it. As it stands, this formula is incorrect.
It is untrue to say the glass would not have moved as it did
unless P had made it do so, for if P had not moved it some
other person might have done so. What we mean, of course, is
that the glass would not have moved by itself: that is, if P had
not performed action A, or some other action resulting in O's
moving, the glass would have remained stationary. In short,
had A not occurred, the glass would not have moved, though
all other features of the setting remained unchanged. One
might perhaps say that A is conditionally necessary for the
occurrence of M. Or again: the occurrence of A included
some of the necessary conditions for M to occur.

In speaking of this feature of the paradigm I have used a
"counterf actual." I said that, in the presence of certain con-
textual factors, M would not have occurred but for A 's occur-
rence. Alternatively: if A had not occurred (other things re-
maining unchanged), M would not have occurred. Now such
a statement is a so-called "counterfactual." Some contempo-
rary philosophers have found this notion troublesome — pos-
sibly because they have failed to explain satisfactorily how a
counterfactual conditional could be verified. Yet the notion
"would not have happened unless" is as primitive and un-
problematic as the notion "making something happen." Both
are applicable in the same circumstances, long before any
question o!" scientific terminology arises. There are accordingly
relatively direct and unsophisticated ways of establishing such
a claim as: The glass would not have moved unless somebody
had moved i:. In making such an assertion we do, in fact,
simply rely on our commonplace knowledge that when no-

40 / Determinism and Freedom

body is "doing anything" the glass stays put, and that when
somebody does "do something" of a certain sort the glass
moves. It would, however, be a mistake to say that the state-
ment "The glass would not have moved by itself" had the
same meaning as "Glasses do not move when left alone"; for
the two statements are made in different contexts and have
different uses, even though the procedure of verifying them
may sometimes be identical.

When pushed, the glass had to move. Certainly it is natural
to say this, and there must be some sense in which it is true.
No doubt m3^hology plays a part: there is a discernible in-
clination to think of the moving object as animate — a mani-
kin, helpless in our grasp, "having no choice" but to move. But
good sense remains when mythology has been discarded. We
need only remind ourselves of the circumstances in which we
say that an object acted on by an external force does not have
to move. We say so when the given force is insufficient to pro-
duce the desired motion. If I push my cat gently, Hodge may
or may not move, though if he does I shall say he did because
I pushed him; but if I push hard enough, Hodge has to move.
Again, a penny tossed into the air has to come down again,
but it does not have to come down tails. Here and elsewhere,
the relevant contrast is between what sometimes happens and
what invariably happens. To generalize: we say that M had
to happen when A happened, only if M would always ensue,
given an unchanged setting and the same concomitant. Using
a phrase parallel to one introduced earlier, we might call A
conditionally sufficient for M. Alternatively, we might say that
A is a part of a certain sufficient condition for the occurrence
of M.

In this cursory examination of some features of a paradigm
of making something happen, I have had little occasion to
refer to any "constant conjunction" between producing action
and induced motion. The omission has been deliberate. The
assertion "P made M happen by doing A" does n^t mean the
same as "If P were to repeat actions sufficiently like A, then,
other things being equal, motions sufficiently like M would in-
variably ensue." If the analysis were correct, the original
causal statement would include as part of its meaning ^
generalization whose verification would need repeated ob-
servation and an induction upon an indefinite number of
situations resembling the original situation. The orginal state-
ment ("P made something happen, etc.") is so far from being
verifiable by inspection that a lengthy inquiry would be needed

-i

Making Something Happen / 41

to establish its truth. (It is as if we had to perform a long
inductive inquiry into the behavior of meter rods before we
could use a given meter rod to measure a given object.) But
I think the truth of the matter is much simpler: in order to be
sure that P made O move, we need only look. The verifying
situation is right before our eyes. To establish conclusively
that P did make O move, we need only be sure that P did do
such-and-such, and that O was moving thus-and-thus mean-
while.

I do not say we should be right in maintaining that A made
M happen whenever an action and a cotemporal motion are
contiguous. In using the language of "making something hap-
pen" we take for granted that the episode in view has a special
and appropriate character. Should we be challenged to specify
these conditions in full detail, we should eventually have to
talk about constant conjunctions; and in deciding in unusual,
unfamiliar, or abnormal settings whether the use of causal
language is appropriate, prolonged inductive investigations
might be needed. But such investigation would establish the
presuppositions for the proper use of causal language, not the
meaning of the assertions made by means of such language.

Consider the following analogy: If I say, "Jones just made
the move 'pawn to king four,' " a full and sufficient verifica-
tion of my assertion is that Jones shifted a characteristically
shaped piece of wood from a certain place on a chessboard
to another place. Yet this is only part of the story. I would
not say that Jones had made the move if he knew nothing of
the game and was merely moving the piece at random; nor
would I say so if he knew how to play chess but was amusing
himself by replaying some master's game — or was composing
a chess problem. In using the language of chess I take for
granted the institution of chess-playing and a host of related
facts. Before I can teach anybody how to use the language of
chess, I must acquaint him with this background of presup-
positions. But once the background has been established, I do
not refer to it each time I announce somebody's move.

There is a general background also for talking about "mak-
ing things happen." The subject could not be dealt with
properly by anybody ignorant of a host of familiar facts
about motions of human bodies, obstructions and resistances
offered by other bodies, the dependable behavior of relatively
permanent solid objects, and so on. But when we say, "Jones
moved the glass," we do not refer to these uniformities or to
the remainder of the background of presuppositions. When we

42 / Determinism and Freedom

say, "Jones moved the glass," we draw a line around an epi-
sode whose relevant features are directly observable. An in-
formal causal statement is a straightforward report. Stripped
of its background of presupposition, it would have the simple
form, "While P did this, that happened."

There is a sense, therefore, in which "P made M happen by
doing A" can be said to mean the same as "While P did A, M
happened" — the sense in which both statements would be veri-
fied by the same state of affairs. But there is also an important
sense in which the two are strikingly different — because they
imply different presuppositions and are connected with di-
verse linguistic practices.

A full account of the linguistic practices connected with the
vocabulary of "making something happen" would be lengthy
and complicated. One obvious connection is with the language
of imperatives. When we order somebody to do something, we
envisage his making something happen. If our language con-
tained no provision for isolating causal episodes, we could
issue no orders, give no commands. And the same could be
said for recipes, plans of operation, and other features of
linguistic transaction. All that part of our life concerned with
getting things done, or with anticipating and controlling the
consequences of our actions, uses the language of "making
things happen" and is inconceivable without it.

Another connection is with moral language. To say that
somebody made M happen is to hold him responsible for it; it
can be a prelude to the assignment of praise or blame, punish-
ment or reward. And the further connections with ethical
practices are equally obvious. To state the point negatively: a
language containing no provision for linking persons with
events for whose occurrence they were held responsible would
be one in which moral judgments as we now know them
would be impossible.

Ill

So far I have been considering primitive cases of making
something happen. But we also talk about "making something
happen" in an enormous variety of derivative situations. Some
of the ways in which these related uses are connected with the
paradigm are fairly obvious. I have been confining myself to
cases where some person causes a motion. But it is natural to
extend the language to cases where the agent produces a ces-
sation of motion, i.e., where the motion would not have ceased

Making Something Happen / 43

but for the person's intervention. Or, again, it is equally
natural to talk of "making something happen" when what is
produced is a qualitative change. And so we pass, by easy
transitions, from the material realm to that of the affections
and sentiments. We talk of making somebody laugh, of mak-
ing somebody reconsider, or making somebody happy — with-
out always realizing how far we have strayed from the
prototypes.

Criteria for the use of causal language can also shift in
other ways. For instance, we commonly speak of intermedi-
aries as causes. If I make a billiard ball move in such a way
as to set another in motion, I can think of the impinging bil-
liard ball as the causal agent. Here we discard the criterion of
the human agent, allowing the motion of an inanimate object
to count as that which "makes something happen." It is easiest
to make this type of transition when the new field of applica-
tion most plainly resembles the original paradigm. We freely
attribute causal efficiency when some motion can be made to
look like the motion of a human body. So we find no difficulty
in conceiving of "forces" that push or pull, bend, or squeeze,
but experience extreme discomfort in trying to imagine "ac-
tion at a distance." The idea of a body "impressing" an ex-
ternal force is altogether natural, but we cannot understand
how one body can "attract" another without being joined to it
by an unbroken chain of physical intermediaries.

Anybody with a logician's desire for clear-cut distinctions
may well be exasperated by the lack of systematic principle
in these patterns of analogical and metaphorical extensions
of causal language. A search for a common denominator in
this kaleidoscope of applications leads at best to "universal
conjunction" or the even vaguer notion of "predictability." But
such abstract and simplified formulas fail to do justice to the
actual uses of "cause" and its cognates. It would be more to
the point to ask what role the language of causation plays — to
inquire into the purposes served by passing from the home-
spun language of "making things happen" to the more abstract
language of causation.

A partial answer might be that the language of causation
seems most fitting when we are concerned with the effective
production, prevention, or modification of events. Roughly
speaking, an event X is most plainly eligible as a possible
cause of another event Y, if we can manipulate X in such a
way as to modify Y. A cause is something that we can or
might be able to control. But we invoke causes also when we

44 / Determinisni and Freedom

are interested in explaining something rather than controlling
it. And as our accepted patterns of explanation become more
complex, our notion of a cause becomes correspondingly more
elusive, until it threatens to vanish altogether into the abstract
conception of a law, a parameter, a boundary condition, or
some combination of all of these. As scientific modes of in-
vestigation develop, the language of cause tends to its own
supersession.

But this is not a special quirk or weakness of the language
of causation. It happens regularly and characteristically in the
transition from ordinary language to scientific terminology.
Dominating my discussion throughout has been the notion
that the vocabulary of "cause" and its informal progenitors is
indigenous to ordinary language — the language of practical
affairs and common-sense observation or understanding. The
vocabulary of causation can be adapted to a scientific context,
jbut the sophistication it suffers proves ultimately fatal. Scien-
jtific insight is the death of causal conceptions. But this does
not mean there is anything amiss with the language of causa-
tion when employed in its proper settings. To say the opposite
would be as implausible as to hold that the supersession of
words like "hot" and "cold" in favor of the scientific termi-
nology of thermometry shows that there is something wrong,
or in need of correction, in the prescientific uses of thermal
words.

IV

I have been arguing that "cause" is an essentially schematic
word, tied to certain more or less stable criteria of application,
but permitting wide variation of specific determination accord-
ing to context and the purposes of investigation. Now, if this
is so, any attempt to state a "universal law of causation" must
prove futile. To anybody who insists that "nothing happens
without a suflBcient cause" we are entitled to retort with the
question, "What do you mean by 'cause'?" It is safe to predict
that the only answer forthcoming will contain such schematic
words as "event," "law," and "prediction." These, too, are
words capable of indefinite further determination according
to circumstances — and they are none the worse for that. But
universal statements containing schematic words have no
place in rational argument. The fatal defect of determinism is
its protean capacity to elude refutation — by the same token,
its informative content is negligible. Whatever virtues it may
have in encouraging scientists to search for comprehensive

Making Something Happen / 45

laws and theories, there can be no rational dispute about its
truth value. Many of the traditional problems of causation
disappear when we become sufficiently clear about what we
mean by "cause" and remind ourselves once more of what a
peculiar, unsystematic, and erratic notion it is.

Chapter 3

Petermiriism and Noveify

William Barrett, New York University

Most of what I shall have to say in this comment will be
directed at Professor Blanshard's paper. This is not to be taken
as a judgment on the relative merits of the two foregoing
papers. Professor Black has suggested a very interesting para-
digm by which to explore the question of determinism, and it
would be fruitful to examine the question wholly within the
framework he sets. But since in most matters I am, I think, in
agreement with him, and since the function of the commen-
tator seems to be not so much to applaud in agreement as to
join battle in debate, I shall be concerned principally with
Professor Blanshard's contribution.

Professor Blanshard has presented an admirably clear, can-
did, and persuasive paper in defense of the determinist posi-
tion. On a subject like this, which has been so long discussed
that nothing new is likely to be said, it is good to get a paper
that performs a real act of simplification, that clears the
ground of clutter, and so suggests the possibility of taking up
the old beat-up problem with some degree of freshness.

Let us begin on the prephilosophical level of everyday life.
Determinism is a position repugnant to most people. Why is
this? The reasons. Professor Blanshard says, are our stub-
born feeling of freedom and the belief that "determinism
would make nonsense of moral responsibility." Here he omits,
I think, one of the main motives in the rebellion against de-
terminism, not only on the part of ordinary people but also
of those modern philosophers who have been most vigorously
opposed to the determinist position: namely, the desire for
freshness, novelty, genuine creation — in short, an open rather
than a closed universe. Such is the main impulse in the criti-
cism of determinism by philosophers hke Peirce, James,
Bergson, Whitehead, and Dewey.

It is worth while in this connection to recall the story by
the Italian poet Leopardi about the Almanac Vendor. The
vendor appears under a window hawking his wares, crying

46

Determinism and Novelty / 47

out that he has good predictions to sell for the year. A man
leans out of the window and engages him in conversation:
How long has the vendor been selling almanacs and making
this same cry? "Twenty years, Excellency." And if he had the
chance of living over any of those years? "No, Excellency,
certainly not; not for all the money in the world." Here the
man who sells predictions would not care, as a human being,
to have those predictions true in detail. The point of the story
is that even a good year — had we to live it over again with
every detail fixed beforehand — would stifle us with boredom:
our food would taste dull to our palate, our most spontaneous
talk sound as uninspired as the playback of a tape-recorded
conversation, and our words of love would sound hollow be-
cause we should know beforehand the precise moment of
fatigue when they would expire.

For my own part, I find the traditional concepts of free
will extremely tenuous and rarefied in the face of my experi-
ence of people and the concrete situations of life, and from
that point of view I could swallow determinism without too
much trouble. Indeed, on the prephilosophical level my con-
viction tends rather toward the old primitive notion of fate,
Moira, of Homer and the tragic poets; but Homeric Moira is
a very different thing from Newtonian or neo-Newtonian de-
terminism, and my recoil from the latter is due to its dreary
prospect of a stale and routine world from which surprise and
genuine novelty may ultimately be banished.

But to come now to the philosophical level. Here, too, de-
terminism seems to me to make very good sense, for it isolates
the central issue: namely, the question of the predictability
of phenomena, and moreover predictability in detail. For
determinism predictability in general is not enough: it has to
assert (and prove, if it can) predictability down to the last
detail — ^lock, stock, and barrel, and even down to the last ,
scratch on the barrel. Anything less than this, and the thesis \
I of determinism must cnmible. Determinism cannot afford
to leave any loose ends lying around. Small and great are
inextricably linked in the happenings of nature and history;
and unpredictable detail might trigger an enormous explosion,
and empires and battles do sometimes hang on a straw.

Professor Blanshard presents a very much simplified
schema for examining the determinist assertion: "Given A, B
must occur." Such simplification allows us to cut to the core
of the matter. Professor Blanshard generously waives the
[question of what "must" means here, and proposes that we

48 / Determinism and Freedom

can go on with the discussion without settling it. I am sorry
I cannot quite accept his generosity. So far as I can see, the
empirical evidence for determinism lies in the fact that some
phenomena can be predicted, and the "must" in Professor
Blanshard's schema therefore has to be understood in con-
nection with predictability. Determinism is surely not a priori;
not only is its denial conceivable by us, but it has actually
been proposed as sober theory by scientists. Such scientific
proposals may, in the end, turn out to be unwarranted by the
facts, but the mere fact that they are seriously made and
utilized for purposes of explanation should suffice to show
that determinism itself is not founded on any a priori neces-
sity. My discussion will therefore present an altogether differ-
1 ent intellectual perspective from Professor Blanshard's, for it
j will turn centrally on the notion of predictability.

With the problem thus shifted, its nature changes consid-
erably. We should be engaged in a painstaking examination
of those regions of experience where we have established pre-
dictability, of the regions where we have not, and of the likeli-
hood of our extending predictability to these latter regions. A
good deal of very antiquated dialectic would promptly fall by
the wayside. This may be laying very rough hands on deter-
minism, but it has been dealing so long in the I.O.U.'s and
promissory notes of hypotheticals — // all conditions are given
(when they so obviously can't be); // we knew everything
(when we so obviously never shall), etc., etc. — that at long
last we may be pardoned for demanding cash or cutting off
trade. Recasting the problem thus, we must also attempt to
make the deterministic schema more specific by applying it to
some definite models (in the mathematical sense of this last
term). When Professor Blanshard lays down as the schema of
determinism "Given A, B must occur" I am unclear about his
meaning, not only because of the ambiguity of "must" but
because I am imsure of the range of values to which A and B
must apply. Suppose we substitute for A and B the following:
Given the nineteenth century, the twentieth century must oc-
cur— that is, exactly in the form it does, down to the last
detail; or: Given the first millennium of the Christian period,
the second must follow exactly as it did. Such theses are, to
my mind, too staggeringly vast to prove or disprove. Clearly
some contraction of the range of variables is called for.

Accordingly, I should like to make some speculations about
the question of predictability in connection with four specific
fields: (1) mathematics; (2) physics; (3) intellectual and

Detenninism and Novelty / 49

artistic creation; (4) history, in the sense of individual and
relatively circumscribable events in time. Obviously within the
scope of these brief remarks I can hope only to indicate, at
most, certain issues that have a different weight for me and
for Professor Blanshard, or at least lead my mind in a differ-
ent direction from his.

1, Mathematics. Godel, as is well known, has established
the incompleteness of mathematics. What this means, in more
human and concrete terms, is that mathematicians, as long as
they remain creative, will always be exposed to the possibility
of unpredictable surprises. Of Godel's result itself, a mathe-
matician of my acquaintance remarked: "Mathematicians
didn't foresee any result like that. In fact, if you had asked
them beforehand, they would have said that just the opposite
would turn out to be the case." Of course, there is the distinc-
tion between the psychological and the logical, and the sur-
prises, after we have understood them, turn out to be logically
inevitable. To be sure, to be sure; but it may be questioned
whether from a human and practical point of view this will
make much difference, since mathematicians, as a conse-
quence of Godel's discovery, will have to face 2,500 years
from now a possibility of shock and surprise like that experi-
enced by the Pythagorean brotherhood at the discovery of
irrationals — which, according to one tradition, cost the dis-
coverer his life at the hands of the outraged brotherhood.

The point becomes stronger if we consider Godel's result a
little more specifically. A Godel formula is a perfectly con-
structible formula in elementary arithmetic; syntactically in-
terpreted, this formula says that it itself is not provable. But
it is not impossible — and this, thanks to Godel himself — that
a worker in number theory, operating with minimal formali-
zation and maximal use of the devices of analysis, as is usually
done, and not at all following the syntactical model, might
come up with a demonstration, a long chain of formulas, at
the end of which would be a Godel formula. Not only is it
unpredictable that this will not occur; it is unpredictable how,
when, and where it will occur, if it does — for if all this were
predictable, we should have produced the proof before it was
produced. What would have happened here would be merely
that a contradiction would have turned up in classical mathe-
matics. Such a contradiction, of course, would not be catas-
trophic; mathematics, like life itself, thrives on contradiction
and is fructified thereby; mathematicians would simply get
busy plugging the leak so that the vessel could continue its

50 / Determinism and Freedom

voyage — a voyage that, as a result of Godel's work, now
shows itself to be over an ocean that is shoreless.

The determinist tends to think that the advance of a science
always means an advance in predictability. Sometimes, how-
ever, an advance brings about a situation of disorder rather
than order. Mathematics after Godel is, so to speak, in a state
of greater disorder than before him. We have to think of it
less as a tightly organized monarchy dependent on some cen-
tral court and more as a federation of states, sometimes with
rather fluid and overlapping borders — borders at which con-
flict and contradiction may occur.

In bringing up mathematics for consideration I am of
course deliberately raiding the classical citadel of necessity.
After centuries of Platonism, we seem compelled at last — and
by the results of the mathematicians themselves — to construe
mathematics as a human enterprise, one that has, therefore, a
thoroughly human future with the unpredictable possibilities
that such a future involves.

2. Physics. Here Professor Blanshard has delightfully
cleared the air by suggesting that it is all a question of which
physicist one reads. I shall not debate with him the relative
consensus among physicists, especially of the younger gener-
ation, though I think it favors my side; rather, I should like
to accept his point and push it one step further. The reports
that come to the layman these days from the physical labora-
tories are so rich and teasing to the imagination — with their
talk of left-handed and right-handed neutrons, antimatter,
and the rest — that perhaps we had better let physics alone for
the present and not try to draw any philosophical conclusions
from it. The change might be welcome to the physicists them-
selves: physics will be just physics and not something else.
Certainly it has ruled the roost long enough on this philo-
sophic question of determinism, so that it may now be time
for a change. Physics now seems in such a state of upheaval
that the whole science provides a good example of unpre-
dictability; the new theoretical synthesis of all recent findings,
if and when it comes, is unpredictable now. But these are
matters for the physicists at tomorrow's session, when they
will perhaps tell us something very different.

But Professor Blanshard does not seem to me to assess
quite fairly the damage done to his position by the indeter-
ministic leanings of modern physics, for these do at least
make questionable what has hitherto been the main prop of
determinism. The physical model (as in the case of Laplace's

Determinism and Novelty / 51

demon) has hitherto been the chief model of all determin-
istic thinking. Despite Heisenberg, Professor Blanshard tells
us, nature in itself may be deterministic. To which we can
only answer: then again, it may not be. "May be" sounds
very curious issuing from the lips of a determinist; and when
he is forced to the minimal assertion that determinism, after
all, may be true, he is already traflScking in contingency. Be-
hind the screen of appearances God may be in his heaven and
every detail in nature in its place in accordance with unalter-
able law; but if we, by the natural laws themselves, are to live
always on the hither side of the screen, we shall have to leave
God's prospect to God and not hanker for something beyond
our own human limitations. Professor Blanshard will find my
view here, as throughout, incurably anthropocentric — to
which charge I plead guilty — but I see no other center when
we are talking about the possibilities of human knowledge.

3. Artistic and intellectual creation. Here wc pass from
hypothetical considerations of unpredictibilities that lie at
the fringes of knowledge to actual cases of unpredictability.

There is the famous case of Poincar^'s perceiving the solu-
tion of a mathematical problem that had agitated him for
months at the moment he set foot on the bus at Coutances.
The mathematician himself could not have foreseen this event:
he would already have had to possess the solution to this
problem, together with the psychological knowledge when,
where, and how this solution would come to him — an obvious
contradiction. But if Poincare could not have foreknown the
event, neither could anyone else; for no one else had carried
that problem as far as he. We cannot imagine a super-
psychologist or superpsychoanalyst, armed with an impossibly
complete knowledge of Poincar6's unconscious, predicting
this event, for this psychologist, too, would already have to
know the solution to the mathematical problem. Of course,
there remains the possibility of foreknowledge by the Divine
Mind, or by his secular surrogate, the demon of Laplace, but
I think we have to renounce such figments once and for all
and recognize that if we are talking about prediction we have I
to have in mind some possible human being, representative r
of some body of human knowledge, who is going to make the
prediction. To be sure, one must distinguish between unpre-
dictable in fact and unpredictable in principle; but predictions
do not float in the air by themselves; they are made by hu-
man beings, and if there is no conceivable human being who
can make the prediction, then we have to say (as in the pres-

52 / Determinism and Freedom

ent case) that we are dealing with something unpredictable in
principle.

Even if we consider the fictitious picture of the mind of
the creative genius dismembered into all its mental atoms (if
there be such), does the possibility of prediction present it-
self? The best empirical study on this subject that I know
happens to be the work of a literary scholar: it is The Road
to Xanadu, by J. L. Lowes, a dissection of Coleridge's "An-
cient Mariner" and, principally, "Kubla Khan." The latter
poem, "Kubla Khan," is a rare thing in the history of litera-
ture, since it comes closer than any other work to being a
purely spontaneous and unconscious creation. Coleridge tells
us that he woke from a nap with the whole poem in his mind
and immediately began to write it down just as it had come
to him in his sleep; there came a knock at the door, announc-
ing that mysterious and forever unknown visitor, a neighbor
from Porlock; Coleridge talked with him for some minutes,
then returned to his desk to finish transcribing the dream; but
it had completely vanished, and we are left with the teasing
fragment of "Kubla Khan" as it is — in its own way, however,
perfectly complete as a poem. Around 1900 some early note-
books of Coleridge's turned up on the literary market and
were edited by German scholars. Armed with a hindsight
knowledge of the two poems, Lowes read through all the
books that had fed Coleridge's imagination.

Now, Lowes did find that a great number of Coleridge's
images and even some of his phrases could be traced back to
his earlier reading, particularly of travel books. Suppose, for
argument's sake, that we could trace every image and phrase
back to such antecedent reading. There would still remain
the selection, fusion, and transformation of these in the poem.
The poem is as unmistakably by Coleridge as any of his rela-
tively more conscious creations; only he could have written it.
Our hypothetical superpsychoanalyst could not have predicted
the dream without writing the poem; but this, of course, no
one could do but Coleridge. Hindsight does not remove the
fact of unpredictability; if it succeeds in tracing certain ele-
ments back to some antecedent source, it nonetheless leaves
us with the realization that it would have been impossible to
)^ foresee how, when, where, and why these elements would

I come together. In short, the enumeration of antecedents can
at best give us only the necessary but never the sufficient con-
ditions of a creative act. The introduction of the unconscious
as an explanation does not help the determinist at all. Far

Determinism and Novelty / 53

from it: for the unconscious, when it is truly creative, is far
more unpredictable than the conscious mind.

Indeed, artists repeatedly testify that when they sit down
to their job they themselves cannot predict the work that will
be produced. The psychological examples usually cited by
determinists reflect the more monotonous and routine aspects
of our behavior, as if to reinforce their general picture of the
world as a vast and dreamy machine. Professor Blanshard is
not of this kidney; he wisely refers to a "causality of a higher
level in the psychical realm." But something very odd happens
to the word "necessity" when he refers to "aesthetic necessity"
as a vis a tergo producing the unpredictable work, and I can-
not see that the term here has any resemblance to the meaning
it has in physics or in general deterministic schema he ad-
vanced earlier.

4. History. If in the previous sections I have merely given
indications of points, here — because of limitations of space —
I shall be able to give only indications of indications.

So far as I can see, recent historians (except the Marxists)
have become much more cautious, far more inclined to qual-
ify, in asserting determinism on any large scale in their works.
Determinism in history seems largely the historicism of the
nineteenth century; it is ideological rather than scientific, and
its technique is to impose, at the dictates of its ideology, large
simplifying patterns in place of the actual chain of events.
Anyone who wants a good example of this method might
compare Trotsky's History of the Russian Revolution, a work
composed thoroughly in accord with the nineteenth century
ideology of Marxism, with the treatment of the Russian
Revolution in Sidney Hook's The Hero in History. If you read
Trotsky's work carefully, you will notice that it abounds in
metaphors drawn from physics — the pressure and explosion
of gases, the fusion of metals, etc. — all intended to convey a
feeling of inevitable historic processes carried to their fore-
ordained conclusion. Actually, continual apologizing is en-
gaged in behind the scenes, because none of the cafe-intellec-
tual Marxists of 1914 predicted that the Revolution would
break out in Russia, and indeed their predictions pointed quite
to the contrary. If we aproach the facts in a more critical and
scientific spirit, without an ideological parti pris, then I think
we must conclude with Hook that the Revolution could not
have taken place without the individual figure of Lenin. And,
of course, tne introduction of human personality multiplies

54 / Determinism and Freedom

the factors of chance beyond the determinist's ability to press
them into any one of his ready-made schemes.

But the issues of history can perhaps be brought into
sharpest relief with a consideration of our position now, in
mid-twentieth century, as we face the future. This historical
future is thoroughly problematic and uncertain. Will Russia
or the United States be victorious? Or perhaps a third force?
Will Conmiunism conquer in the twentieth century? When
will the next war break out? When will the first bomb be
dropped? With such questions as these we face a future as
unknown to us as the events of the next millennium were to
an intelligent Greek of Thucydides' time, who must have felt
— such, at least, is our impression from reading what has been
called the first of modern histories — that the world would
change, that it would become very different from what it was
now, though in ways that could not be foreseen. Two thous-
and four hundred years have not made the least impression on
unpredictability here.

Such questions about the future bring us back to the pre-
phUosophical level from which this comment took off, the
level at which, as individuals, we all live and die. On this level
we encounter a vast, shaggy, amorphous mass of unpredicta-
bility on which our knowledge has made very little impres-
sion. Billions of dollars trade hands annually in games of
chance; people are anxious about their future, and fortune-
tellers ply a prosperous trade (there are currently eleven
astrologers in the Classified Telephone Directory — a curious
juxtaposition of ancient and modern); and where primitive
peoples scanned the skies for signs and portents, we do so for
unidentified aircraft, missiles, or flying saucers. Modern civili-
zation has stabilized social life and enormously increased
predictability in certain areas, but in another sense it has also
enormously increased the total quantity of contingency and
distributed it to other focal points in the body social.

Of course this final litany of contingency is not intoned as
proof — only as a reminder to the determinist that the spheres
in which determinism has so far been established are re-
stricted. When Professor Blanshard concludes by saying that
he has dealt with the three commonest objections to deter-
minism, he seems to have forgotten the simplest, most direct,
! and, to my mind, most overriding objection: that as a total
\ thesis determinism simply remains to be proved.

^

PART §§

Psfermlnism In Modern SeiarBc®

Chapter 1

Deferminism in Modern Selene©

Percy W. Bridgman, Harvard University

Anyone who is inclined to define science as the consensus of
competent scientists may well pause for second consideration
when he contemplates the present state of opinion in physics
with regard to the question of determinism as it is presented
by quantum mechanics. Seldom in the history of physics (and i
for the purposes of this exposition I shall take the "science"
of my title to be pretty nearly equivalent to physics) has there
been such a radical difference of fundamental outlook be-
tween the acknowledged leaders. The so-called "official" or
"orthodox" attitude is that of Niels Bohr and Heisenberg and
Max Born. But it is well known that during their lifetimes
Planck, who laid the foundations of the theory, and Einstein,
who very early developed some of its most important physical
consequences, were irreconcilably opposed to the official point/
of view. The intensity of Einstein's opposition amounted al-
most to intransigence. Among those living today, de Broglie,
who first had the fundamental insights of wave mechanics,
and whose first intuitive attitude was akin to the more tradi-
tional classical attitudes of Planck and Einstein, was blud-
geoned into twenty-five years of unwilling acceptance of the
official doctrines by Bohr's apparently irrefutable logic, but
has recently kicked over the traces again with his glimpse of
a possible way of refuting Bohr's argument. And it is well
known that Schrodinger, who gave wave mechanics its mathe-
matical form, has from the beginning been more or less of an
enfant terrible, refusing to accept such orthodox concepts as
quantum jumps. There is a host of other names not so closely
connected with the historical beginnings of the subject. In
this country Henry Margenau at Yale is well known to have
unorthodox points of view. David Bohm, also unorthodox,
has had a considerable following recently. In 1953 he wrote
a widely used book in which certain departures from the
orthodox view were suggested, and these have been further
developed and accentuated in recent writings. "Strife About
Complementarity," '^ a lively article by Mario Bunge, paints

^Brlt. low. Phil. Sci.. May 1955.

57

58 / Determinism and Freedom

a vivid picture of the present conflict of interests, viewpoints,
and temperaments.

The point at issue in all the discussions is itself somewhat
fuzzy. Such terms as "determinism," "causality," "natural
law," "prediction" turn up almost inevitably. These terms
are all connected in some way with the point or points at
issue, in spite of the fact that they are not fully equivalent
and that differences among them can be clearly specified if
one wants to make one's analysis meticulous enough. I shall
not find it necessary for the purposes of this exposition to
attempt to sharpen distinctions to the point of emphasizing
all recognizable differences among terms, but shall be able to
use them in a more or less intuitive common-sense fashion.
In addition to the four crudely synonymous terms mentioned
above, some other terms, without which the discussion never
gets very far, turn up: for example, "reality," "subjective,"
"objective." It is more difficult to find the meaning of these
terms than of the first four; one often has to try to reconstruct
the meaning by observing the use of the term in context,
which is unsatisfactory and may result in a multiplicity of
meanings. One may safely comment, however, that when
such terms begin to occur the discussion is heading away from
the narrow limits of concern with the experimental findings
of the laboratory toward something vaguer and more general,
which, for want of a better description, we may characterize
as "philosophical." In fact, it is difficult to keep the discus-
sion from entering the field recognized as philosophy, and
men like Bohr and Heisenberg and Born do not hesitate to
use the word "philosophical" in connection with their own
speculations.

However marked the failure of consensus with regard to
the philosophical aspects of physics, there is no such failure
with regard to the experimental situation in the laboratory.
Everyone is agreed that it is possible to set up in the labora-
j tory experimental systems in which events occur that we are
■^ at present completely unable to predict. This is another way
of saying that we can establish no unique causal connection
between the event and other events or situations, which is
another way of saying that up to the present we have been
able to formulate no law of nature according to which the
given event follows from other things. This again may be
described by saying that, as far as we can now see, the
event is undetermined. Whether or not the existence of such

Determinism in Modem Science / 59

undetermined events is closely connected with what the
philosopher understands by determinism and indeterminism,
we do not need to examine. There are doubtless valid dis-
tinctions to be made here, but for our purposes we need not
elaborate them.

The relevant experimental situations in which there is such
consensus are usually situations in the "microscopic" world
of electrons and individual quantum events. A typical exam-
ple would be the interference pattern produced by a slender
beam of light passing through a system of slits. The pattern
ordinarily seen is a pattern of light and dark bands with
smooth gradations from light to dark. But if the intensity of
light is made very low, the smooth pattern breaks down into
a pattern of individual spots, which mark the arrival of
individual photons of light and the excitation of individual
grains of the photographic emulsion. The place and time of
occurrence of any individual spot in this pattern are at pres-
ent absolutely unpredictable. It is not necessary, however, that
the unpredictable event should be on the microscopic scale; it
would be possible so to couple a disintegrating speck of some
radioactive compound to an atomic bomb as to blow up a
city at an absolutely unpredictable time. Such an unpredicta-
ble situation on the scale of daily life, however, seems always
to involve some reaching down into the "microscopic" and
bringing it up to the scale of daily life.

There is, then, no disagreement about the experimental
situation; disagreement begins with our interpretation of the
significance of the experimental facts, and in the program
for the future action that we draw up in its light. There is
in the first place the question of how best to talk about the
experimental situation as we now have it. This involves in
particular how best to devise a mathematics for describing
the present experimental situation with its at present unre-
solved fuzziness. Here again I think there is consensus that
the present mathematical machinery of wave mechanics is
competent to handle adequately all the phenomena at pres-
ent experimentally accessible — involving the reactions of
atoms, molecules, electrons, and photons — that it was con-
structed to handle. From one point of view this is all the
physicist is concerned with, and we might dismiss all the
disagreement as being outside the realm of physics and there-
fore of no proper concern to the physicists. But physicists are
obviously concerned. Physics is finding itself forced into a
position where it is increasingly concerned with problems out-

60 / Determinism and Freedom

side the realm of its traditional activity and verging toward
"philosophy."

The mathematical machinery, then, is generally accepted.
Bohr and Schrodinger would make the same calculations to
find, for example, how two hydrogen atoms combine to form
a hydrogen molecule. Differences arise in interpreting the
mathematics, which means perhaps that the question is how
we shall best talk about what we do when we use the mathe-
matics. It is to be anticipated that we shall have difficulties
in finding how best to talk about what we do, for the world
of newly discovered happenings is a strange and paradoxical
world in which the customary patterns according to which
ordinary events behave are no longer found. The first major
step in interpretation was taken by Max Born, who pointed
out that the psi function of Schrodinger's wave equation, an
equation that had already been successfully used in handling
atomic phenomena, could be described in terms of probabili-
ties. Specifically, in a problem dealing, for example, with
electrons, the square of the psi function at any place deter-
mines the probability that an electron will be found at that
place. The recognition of this possible interpretation of the
psi function, which is fundamental to the "orthodox" attitude
toward wave mechanics as formulated by Bohr and Heisen-
berg, is everywhere regarded as remarkable intuitive insight
on Born's part

There was a great advantage in an interpretation in terms
of probability because the mathematical machinery for han-
dling probability had been developed by mathematicians for
its own interest and had been previously applied by physicists
to such problems as the kinetic theory of gases and the statis-
tical interpretation of the second law of thermodynamics,
which made understandable the universal degradation of en-
ergy. At the same time it must be recognized that the funda-
mental role ascribed to probability is one of the reasons why
there is no consensus on the significance of wave mechanics.
For in spite of the fact that there is agreement as to the use
of statistical analysis in practical situations — all competent
persons would agree on how to draw up a table for a life
insurance company or how to bet in a complicated gambling
game — there is no agreement as to what probability "really"
is, and the matter is still controversial

Let us now turn to a more detailed examination of the
different points of view. First consider the orthodox interpre-
tation of Bohr and Heisenberg, which is probably that ac-

Determinism in Modem Science / 61

cepted by the majority of physicists. In this interpretation the
Heisenberg uncertainty relations and Bohr's principle of com-
plementarity play a leading role. According to the Heisenberg
principle there is a correlation between the fuzziness of a
position measurement and the fuzziness of a velocity meas-
urement such that when one fuzziness becomes less the other
becomes greater in proportion. It follows that it is at present
impossible to make measurements that will permit the accu-
rate calculation of the position of a particle at some future
time. In this sense, therefore, the future position of the parti-
cle is not determined, and the particle in its motion is not
subject to causality. This situation is often described by say-
ing that it is not possible to make simultaneous measurements
of position and velocity with sufficient precision to predict
the future position (or velocity) — the implication being that
for some reason "position" and "velocity," although "really
there," are not accessible to us simultaneously. Bohr cannot
insist too strongly that this is not a legitimate way of talking;
that, if "in principle" position and velocity cannot be known
simultaneously, then the object cannot simultaneously "have"
position and velocity. We can, however, choose which it is
to "have" by our choice of the measuring instrument. A great
deal of Bohr's analysis has been devoted to a detailed con-
sideration of the greatest variety of different methods of
measurement and to showing in detail that the fuzziness de-
scribed by Heisenberg's principle is a consequence of the
reaction between the measuring instrument and the object
being measured. It was an essentially new insight for the
physicist that the act of acquiring knowledge itself disturbs
the object of knowledge. The further fact that the disturbance
so introduced is uncontrollable as far as we can now see is
fundamental to the structure of wave mechanics.

Bohr's principle of complementarity, which is one of the
chief bones of contention, may be described as a working
out of the consequences of seriously accepting all the impli-
cations of the statement that a body cannot simultaneously
have position and velocity. We can use either method of
description as we choose, but having made our choice we
must abandon use of the other. The principle has more gen-
eral applications that just to position and velocity; the chief
consequence is that if we choose to describe a system in
conventional terms of space and time we must renounce the
possibility of describing it in causal terms, or if we choose to
describe it in causal terms we must renounce describing it in

62 / Detenninum and Freedom

terms of space and time. Involved in all this, in a way that is
perhaps not always very explicitly evident, is the insight that,
since an object never occurs naked but always in conjunction
with an instrument of measurement or the means by which
we acquire knowledge of it, the concept of "object," as some-
thing in and of itself, is an illegitimate one. Acceptance of
this insight with all its implications is obviously going to
react strongly on our idea of "reality," and many of the objec-
tions to the orthodox view arise precisely from unwillingness
to accept the altered view of "reality."

What is Bohr's own attitude toward his orthodox doctrine?
I can only guess, of course, but I think it safe to say that, if
in the future experimental detail is discovered of which we
now have no inkling, by means of which it would be possi-
ble to predict the individual photons in an interference pat-
tern, Bohr would simply accept the fact, as would every sci-
entist, and recognize that the present wave mechanics, as a
method of dealing with the complete physical situation,
would have to be abandoned. For practical purposes of cal-
culation within the appropriate range, however, wave me-
chanics could be retained, just as Newtonian mechanics has
its range of application even after the formulation of relativity
theory. But just as relativity theory forced a conceptual revo-
lution, so I think discovery of at present unsuspected experi-
mental detail would force a conceptual revolution in Bohr's
point of view, and I believe that Bohr would be the first to
recognize this. The question of Bohr's private opinion as to
the probability of the discovery of new kinds of experimental
facts is another matter. One may guess that he considers the
probability rather remote, or he would not have spent so
much time and thought on elaborating his present position.

The possibility of the discovery of new experimental facts
that would invalidate present quantum theory has often been
discussed. A roughly equivalent formulation is to ask whether
there may not be "concealed parameters" in addition to
those we now know and in terms of which we can adequately
describe present experimental knowledge. In this connection
there is an often quoted analysis by von Neumann, which
seems to be generally accepted as rigorous, to the effect that
it is not permissible to suppose that there may be "concealed
parameters" of such a kind that the principle of causality
would be saved by their introduction, because the present
quantum symbolism has no place for such new parameters,
which could not be introduced without sacrificing the sue-

Determinism in Modem Science / 63

cessful account of experiment that we now achieve with our
present parameters. In this analysis "causal" is understood
with the usual connotation of the physicist as being the sort
of complete causality that is inconsistent with Heisenberg's
relation. Von Neumann's argument does not apply to such
an emasculated definition of causality as is satisfied with a
psi function causally controlled. I digress here to remark on
an apparent inconsistency with a pronouncement of Poin-
car6's, in the days when Planck's h was still new, to the effect
that no system whatever, no matter how strange its be-
havior, can be peremptorily declared inconsistent with classi-
cal mechanics if one admits the possibility of sufficiently
complicated concealed parameters. For one can, if necessary,
imagine concealed inside each weirdly behaving atom, or
elsewhere, a robot with written instructions to act exactly
as the atom actually does. A robot, we may assume, is a
classically functioning mechanism. The joker is that the con-
cealed mechanisms reproducing the supposedly unmechanical
behavior must be completely isolated from outside contact or
access, so that there is no method whatever by which their
existence can be established, and for all practical purposes
we are dealing with a pure verbalism.

The question of Bohr's personal attitude is, after all, more
or less beside the point. We need not accept any more of
Bohr's opinions on the "philosophical" questions involved
here than are necessarily implicit in what Bohr has done in
distinction to what he has said. It seems to me that what Bohr
has done, essentially, is, in the first place, to take seriously
the thesis that we shall never discover experimental detail
that will restore causal control to atomic events and the in-
sight that nothing occurs except as part of a larger whole;
and, in the second place, to try to develop a logically con-
sistent way of thinking about what we find. I am not sure
Bohr would maintain that he had devised a system of thought
logically watertight in all conceivable aspects, but I think
we can see that he has gone a long way toward success and
that the problem he has set himself is one that cannot be
ignored.

One can describe the problem Bohr has set himself in an-
other equally noncommittal way: namely as the problem of
devising a method of talking about the presently known ex-
: perimental situation that will emphasize the occurrence of
events we cannot predict.

It seems to me that most of the criticisms of Bohr arise

64 / Detenninum and Freedom

from unwillingness to accept the conditions of the problem
Bohr has effectively set himself. Perhaps part of the unwill-
ingness arises from failure to appreciate the purely formal
aspects of the problem of devising methods of thinking about
a hypothetical physical situation. Regarding the problem in
this formal light, one should be able to attack it purely as an
intellectual exercise. But I believe the unwillingness actually
involves many other factors, some of them recognizably
emotional. The reason people will not see the problem in a
purely formal light is that the hypothetical state of affairs
basic to the problem is so much at variance with their con-
ventional and traditional pictures of what the experimental
situation must be that they will have none of it, and even re-
fuse to speculate how they might act if it actually existed. The
repugnance of different persons to accept the possibility that
the world is actually constructed according to the hypothesis
I of orthodox quantum theory varies greatly and may involve
iconsiderations closely approaching the religious. This is per-
haps most strikingly shown by Einstein, who could not bring
himself to accept the idea that chance plays a fundamental
role in the scheme of things, and who passionately exclaimed,
"Der Herr Gott wiirfelt nicht" ("The Lord God does not
throw dice"). Einstein's repugnance led him so far that, in-
stead of postulating that there might be experimental facts
not yet discovered, which is a perfectly tenable position and
all that he needed, he was convinced that no theory giving
a fundamental place to probability could be logically consist-
ent, and he spent a great deal of time trying to point out
logically untenable aspects of quantum theory, always to
be patiently refuted by Bohr.

Many alternatives for the orthodox interpretation have
been proposed, especially recently. I have neither the time nor
the competence to attempt to specify them in detail and to
classify them. There is an illuminating paper on this aspect
of the situation by Professor Margenau in Physics Today.^ I
shall content myself with some very general comments.

There are in the first place those who maintain that it is
possible to save the causality principle. A classified grouping
together of the various interpretations that attempt to maintain
the causality principle would cut across the lines of classifica-
tion proposed by Professor Margenau; he himself would be
found in this class, although in his own classification he stands

» 7, 6-13, 1954.

Detenninism in Modem Science / 65

alone. The causality principle is usually saved by transferring
its sphere of operation from concrete individual events to the
probabilities in terms of which the theory describes the sta-
tistics of many occurrences of the individual event. This is
possible because the psi function of wave mechanics, which
describes the statistics, satisfies a definite differential equation,
and therefore unfolds in an orderly and predictable way with
time. The psi function may therefore be said to be causal in
the sense that its future distribution is determined by its
present.

Now, no one can deny that the psi function satisfies an
equation according to which it unfolds systematically in time;
the question is, why do those who save the causality principle
in this way regard this behavior of the psi function as perti-
nent or significant, and why do they regard this way of saving
the causality principle with so much satisfaction? No one
who maintains the causality principle by this method would
claim that the individual event, such as the single photon in
an interference pattern, thereby becomes causally determined.
We are here in the presence of a dilemma, for, in spite of the \
fact that we apparently must give up causality for individual
atomic events, it is obvious that there is some sort of regu-
larity here, since the apparently haphazard individual photons
somehow build themselves into a regular pattern when there
are a great many of them. What is behind this regularity?
The fact of the regularity may be formally described by say- \
ing that there are laws of statistical behavior, although there \
are not laws for individual events. But how can this be? How
can individual events, each of which is completely haphazard,
combine into regular aggregates unless there is some factor of
control over their combination into aggregates, and what
kind of control can there be over a haphazard event?

The situation is almost imthinkable with the ordinary con-
notations of our words. It becomes rationally less repugnant,
perhaps, if we can rediscover on another level the regularity
that we have lost on the level of individual events. This I
think accounts for sfMne of the satisfaction of those who em-
phasize that the psi function, which describes the statistics
of physical situations, is itself not subject to chance, but is
controlled by an equation as deterministic as the equations
of classical physics. Such a causality is, however, but a poor
ghost of the robust causality of the classical conception, which
ascribes a cause (and an effect also) to every event. I believe,
however, that there is a deeper undercurrent in the thinking

w

66 / Determinism and Freedom

of such people than merely the desire to recover a moiety of
rationality; they feel that they can now say, "The world is
really deterministic and causal, after all, in spite of super-
ficial appearances to the contrary." Such a feeling usually
carries the further consequence that the psi function is as-
cribed an "objective reality," for what is the use of a really
determined world if the thing that is determined is not real?
This brings up the whole question of the meaning of
"reality," a term that has undergone a surprising renascence
in the usage of physicists after a period of conscious absten-
tion. The word is freely used by Schrodinger and de Broglie,
and even Max Born uses it on occasion. As would be ex-
pected, there is no clean-cut specification of the meaning of
"reality." One is often left to infer the meaning by observing
the usage, as when an author says, "Everyone would agree
that such-and-such a situation is real." I have yet to see, how-
ever, an attempt at a definition of reality that would justify
one in saying that a probability is objectively real.

Here we encounter one of those situations in which, in
spite of long discussion, there is not even yet consensus, for
there still are different schools of thought with regard to prob-
' ability. There are some who maintain that individual con-
1 Crete events "have" a probability, and that it is possible to
\find what this probability is. Professor Margenau is one of
these, and his conception of probability is basic to his inter-
pretation of wave mechanics. He defiJaes his probability in
terms of frequency, and he obtains the probability of a spe-
cific event by observing its frequency of occurrence. I think
he would say that any actual observation of frequency can
give only a rough value for the probability, because of the
finite size of the sample, but that nevertheless it is good
enough for practical purposes. Now, to me — and here I enter
controversial territory — any such view of probability that
allows one to say — and to attempt to prove that he is right
when he says it — that the probability of some concrete event
is some specific number rests simply on poor observation and
poor reporting of what he does when he applies the proba-
bility concept to some concrete situation. What can one pos-
sibly do to show that there was a one-sixth chance that a
particular throw of a die would yield a three when the actual
throw has been made and yields a three? It seems to me that
the meaning of probability must be found in another sphere
of activity. Inspection of what we do leads, I believe, to a
view of probability expressed by Max Born as follows:

Determinism in Modem Science / 67

"Chance can be understood only in regard to expectations of
a subject."^ Elsewhere, however, he says that probability
must be recognized to have some "objective" reference be-
cause we successfully apply probability to objective situations.
This sort of tenuous second-hand objectivity, however, is ap-
parently not what Professor Margenau and others have in
mind.

We return to the question of the frame of mind of those
who are so firmly resolved that there shall be an underlying
order that they are satisfied to find it even in probability. It
appears to me that their views often have some pretty deep
emotional roots. Many men apparently crave a friendly uni-
verse, one in which they can feel themselves at home at least
to the extent that they can believe the universe intrinsically
understandable by the human mind. Because a universe with-
out underlying regularities would not be understandable,
simply owing to the method of functioning of the human
mind, there follows the resolution that there shall be regu-
larities.

A slightly different twist is given to this point of view by
the German philosopher Ernst Cassirer in his book Deter-
minism and Indeterminism in Modern Physics,* recently
translated into English, with an introduction by Professor
Henry Margenau. For Cassirer the domain of natural law is
not the domain of objective things but of cognitions; natural
laws are defined by the ordering of cognitions that they make
possible. For him the causal relation is epistemologically nec-
essary. He says:

We find the essential significance of the causal relation, if
interpreted in a critical rather than a metaphysical sense, to be
that it contains a statement not immediately about things but
about experience, by which and in virtue of which alone things,
as objects of knowledge, can be given us. It expresses something
about the content of empirical knowledge.*

If one reads between the lines, particularly in the early part
of the book, where Cassirer says in one place, "Ignorabimus
is the only answer that science can give to the question of
the essence and origin of consciousness,"^ I think one is justi-
fied in believing that Cassuer displaces the domain of law and

» Proc. Roy. Soc, 66, 503, 1953.

* Trans, by O. Theodor Benfey (New Haven: Yale University Press, 1956).
»/fcid., p. 114.

• Ibid., p. 5.

68 / Determinism and Freedom

causality from the physical to the mental world because of
his belief in the existence of a mental world inaccessible by
the methods of science.

Whether or not this is a fair evaluation of Cassirer's atti-
tude, I think that one cannot read the many writings of
physicists on this question without feeling that they, too, have
injected something nonphysical into the situation. This is
usually unconscious, perhaps, but may break forth into articu-
late expression, as in the following quotation from Schrod-
inger:

. . . there is a tendency to forget that all science is bound up
with human culture in general, and that scientific findings, even
those which at the moment appear the most advanced and
esoteric and difficult to grasp, are meaningless outside their
cultural context. A theoretical science, unaware that those of its
concepts considered relevant and momentous are destined even-
tually to be framed in concepts and words that have a grip on
the educated community and become part and parcel of the
general world picture — a theoretical science I say, where this is
forgotten, and where the initiated continue musing to each other
in terms that are, at best, understood by a small group of fellow
travellers, will necessarily be cut off from the rest of cultural
mankind; in the long run it is bound to atrophy and ossify.''

There is more in a similar vein; one point of particular em-
phasis is that the physicist should not allow himself to forget
his organic connection with the past. Now, some of the impli-
cations of Schrodinger's attitude can be unhesitatingly ac-
cepted; nearly everyone would agree that society as a whole
has a vital stake in the scientific enterprise and that therefore
it is desirable to find the means for disseminating as widely
as possible an appreciation of the nature of science and a
knowledge of its factual findings. But it seems to me that
other and wider implications are unmistakable — the implica-
tion that it behooves the scientist in his theoretical construc-
tions to cast his thinking into idioms of his times and his
culture. In some ways this is a most surprising admonition.
I had supposed everyone recognized that as a matter of fact
the theories of the scientist are colored by contemporary cul-
ture, but that this was nevertheless something to be regarded
as undesirable, although unavoidable because rooted in hu-
man frailty. I had supposed that the ideal of the theorist was
to erect a structure determined as far as possible by the facts

''Brit. JouT. Phil. Set., 3, 109, 1952.

Determinism in Modem Science / 69

themselves functioning autonomously, and that, if he could
discern that his structure was being influenced by the intel-
lectual fashion of the times, scientific integrity demanded that
he resist as far as he could. I am old-fashioned enough to be
shocked at the suggestion that the scientist should cut his
cloth to the general intellectual fashion of his times, but I
very much fear that Schrodinger has put his finger on a
change in the intellectual atmosphere of the scientist. I find it i
hard, in reading the recent discussions of causality and de-
terminism, to resist the impression that many of the debaters
were influenced by extrascientific considerations. This influ-
ence is evident even in the work of Max Born, who says, in
combating Schrodinger's thesis that it is waves and not parti-
cles that are fundamental, "I think Schrodinger's suggestion is
impracticable and against the spirit of the time."* This is
perhaps unfair to Born, who may have been throwing back
into Schrodinger's teeth his demand that the theorist be aware
of his cultural background, but nevertheless one could wish
that Born had not said it.

Among writers of less scientific stature than Born the in-
fluence of extrascientific factors is unmistakable. This is par-
ticularly evident in Mario Bunge's paper "Strife About Com-
plementarity."^ The burden of Bunge's paper is that physicists
are at last awakening from the "dogmatic slumber" in which
they have accepted "the official philosophy of quantum
theory, which is essentially of a positivistic character" and
are embarking instead on "new, realistic, rationalistic, and de-
terministic trends." It seems to me that there is too little
argument in the paper and too much name-calling. It is as-
sumed that the reader will react negatively to such epithets as
"positivistic" and "empiristic" and positively to "realistic,"
"deterministic," and "scientific materialism." The assumption
that the reader will react in the expected way to these epithets
and that it is desirable that he should so react obviously does
not have its origin in any purely scientific experience. In par-
ticular it seems to me that a strong odor of Marxian dialectic
is detectable here.

Bunge lists four axioms "underlying every scientific en-
deavor and confirmed by its failures and successes": (a)
nature and every one of its parts is inexhaustible, actually as
well as potentially; (b) nature is an interconnected whole, so

8 Guthrie lecture, p. 507.

"Brit. Jour. Phil. Set., May 1955.

70 / Determinism and Freedom

that the complete specification of a single object would re-
quire the complete specification of the whole universe; (c)
knowledge is as inexhaustible as its objects; (c^) we are limited
to a finite number of variables, whereas the complete speci-
fication of every bit of matter would presumably require an
infinite number of variables. It is not quite clear what is meant
by calling these "axioms." The meaning cannot be that all
scientific workers do as a matter of fact follow these axioms,
for Bunge's contention is that the orthodox interpretation of
quantum mechanics does not follow these axioms and there-
fore cannot be correct. Neither can "axiom" be used in the
sense of mathematical postulate analysis, where an axiom is a
formally accepted preliminary to a deduction of the logical
consequences concealed in the axiom. Sfurely no one would
maintain that the physicist is any more restricted than the
mathematician in setting up any system of axioms that he
pleases and deducing he consequences. "Axiom" must mean
something else; its usage indicates that it ought to be de-
scribed, rather, as a precept or maxim that "should" guide the
scientist. We may well ask, Whence comes the authority
claimed for these precepts? Surely it does not come from any
scientific experience but from something extrascientific, some-
thing that is never formulated and can only be guessed at.

Considering the axioms in detail, one can have little objec-
tion to the first, if it is to be taken as advising the theorist
that he would do well to leave room in his thinking for the
discovery of at present unknown facts; but it is to be rejected
if it suggests that it is illegitimate for the physicist to specu-
late what the consequences would be if there were no new
facts to be discovered, which I think is essentially what Bohr
has done. The second axiom, that nature is so interconnected
that the complete specification of a single object would re-
quire the complete specification of the whole universe, is in
the spirit of the old Newtonian mechanics, which could set
up a formula giving numerically the gravitational interaction
of any two objects in the universe, no matter how small or
how remote from each other. But this axiom is completely di-
vorced from any possible experience and furthermore violates
the spirit of the first axiom, because it effectively says that
nature is so constituted that there is no minimum interaction
below which one object has no effect on another, and in so
saying presumes to dictate the structure of future experience.
Not only this, but it is inconsistent with the structure of our
present experience, for one of the consequences of Planck's

Determinism in Modem Science / 71

discovery of h is precisely that it showed that there are, as
a matter of fact, minimum interactions. I remember my own
almost audible relief when I saw that in discovering the ex-
istence of h Planck had absolved me of the necessity of be-
lieving that the slightest motion of an atom on this earth has
some effect on every atom in Sirius, an idea that always
seemed to me absurd to the point of grotesqueness. I never
was willing to push mathematics so far or to take it so seri-
ously. Yet the Newtonian view is the one to which the classi-
cal conception of causality would lead one.

Bunge's other two precepts, (c) and (d), may be taken in
the same spirit as the first in so far as they are not truisms,
namely as offering advice as to the direction that theory may
take in order to be probably profitable, but in no sense as
setting limits to the direction that it must take. I have doubt-
less devoted a disproportionate amount of space to Bunge's
paper, but it does seem to me to offer an illuminating example
of what may happen when we allow ourselves to be unduly
influenced by extrascientific considerations.

The attentive reader can perhaps detect an undercurrent of
disapproval in my ascription of extrascientific motives to
those who are led by their desire for a friendly universe to
^_d^mand determinism and to repudiate probability as an ulti-
~"mate. The same repudiation, however, may arise from con-
siderations not necessarily emotional at all. One of the funda-
mental drives of the scientist is to find explanations for the
things around him. If one accepts probability as ultimate, he |
hsis., ^^Y the yejy (jgfirijtion of probability, givenjjp Jhe jpossi- T
bility_of_explanation7T would concede thafgnelias a per|cct
right not to give*\ip the possibility of explanation except in
the very last resort, and to follow as long as he can a pro-
gram of theorizing that envisages the ultimate possibility of
finding some sort of "explanation." It seems to me that it is
in this spirit that de Broglie has returned, in his most recent
book, to a further examination of the consequences of as-
suming a sort of ghost mathematics back of the recognized
mathematics of wave mechanics — a mathematics in which
most of the present results are reproduced but that retains a
place for possible causality.

I should now like to present the conclusions to which I my-
self have come. There have been several guiding insights: In
the first place, the so-called microscopic world, which is com-
monly thought to be the arena of quantum theory, is really
nothing but the altered world of our own macroscopic expert-

72 / Determinism and Freedom

ence — altered because we have learned new sorts of macro-
scopic manipulations such as constructing microscopes or
Wilson cloud chambers or Geiger counters. It is consequently
to be expected that the roots of the difficulties revealed to
us by quantum theory are already present in the sphere of
ordinary life and should be discoverable by acute enough
analysis. Secondly, the process of acquiring information or
knowledge in any physical situation must interfere with and
to some extent alter the system about which we wish to ac-
quire information. This means that no object can ever be
strictly isolated and considered in and for itself, but must be
treated as a part of a larger system in which at least the instru-
ment by which we acquire knowledge is included. It is further
made plain that aspects of a situation about which we have
no actual or potential knowledge can be of no concern. Such
aspects can only be talked about, and then not without con-
cealed self-contradiction. Finally, our ultimate concern is with
events or happenings rather than things. With a background
of this sort it seems to me that we can face the world as re-
vealed to us by quantum theory without feeling that it does
not make sense.

Take in the first place the consideration that the object of
knowledge is not to be separated from the instrument of
knowledge. Two extreme limiting cases are to be recognized:
the instrument is very small compared with the object, and the
object is very small compared with the instrument. Since we
are in the last resort the instrument for acquiring knowledge,
the two limiting cases are: objects large compared with us and
objects small compared with us, or, in other words, simply
large and small objects. Now, it seems to me to make sense to
suppose that a small instrument can find out more about a
large object than a large instrument can find out about a small
object. The small instrument can explore the large object bit
by bit in a way impossible for a large instrument with a small
object. It also seems to me to make sense to suppose that more
perfect knowledge of the large object would make possible
more effective ways of dealing with it. In particular, I think
it makes sense to find that we can predict the future behavior
of a large object but not that of a small object. This is merely
another way of saying that it makes sense that causality should
fail for small objects. As part of the same picture, the Heisen-
berg uncertainty relation does not outrage my feeling of what
makes sense. Of course these very vague general considera-
tions can give no hint of the mathematical form of the Heisen-

Determinism in Modem Sdence / 73

berg relation or of the existence of Planck's h — this must
come from the detailed analysis of the actual experiments.

As another example, consider the often quoted paradox that
an electron has no individuality or identity. The paradox dis-
appears if one no longer thinks of an electron as an "object,"
but as an aspect of what happens in particular kinds of physi-
cal situations — including in the physical situation all the in-
strumentation by which the electron is detected or measured.
Under such conditions it need make no more sense to ask
whether the electron we now observe is the "same" as the
electron we observed a moment ago than to ask whether the
wind that now cools our cheek is the same as the wind that
blew yesterday. The paradox disappears even from such state-
ments as that we can "choose" whether the electron is to have
position or velocity. The electron in and by itself does not
"have" either position or velocity; these pertain to what hap-
pens in the whole situation, including the instrumentation. We
can set up a "position situation" or a "velocity situation" as
we please by suitably arranging the apparatus. There is surely
nothing paradoxical here.

I think quantum mechanics make sense, and I think that the
fact that it does make sense and hangs together logically
means, among other things, that a law of causality or a deter-
ministic structure for the universe or an underlying continu- /si
ity as distinguished from discreteness is not, as has often been s /
maintained, a necessary prerequisite to rational thought. But '^
whether quantum mechanics will survive is a different ques-
tion, a question to which the answer rests with experiment.
There is no present indication that experiment will decide
against it. I would like to suggest, however, two directions in
which I think future experiment might look for at present un-
discovered structure.

The first has to do with the significance of the successful use
of the mathematical apparatus of statistics and probability. In
this mathematical apparatus we had a ready-made tool admir-
ably adapted to dealing qualitatively with situations chracter-
ized by vagueness or experimental uncertainty or inaccuracy.
The experimental check so far leaves nothing to be desired;
nevertheless all our experimental checks have been of a rather
special kind in that they have been confined to regions of high
probability density. I should like to see developed an experi-
mental method for dealing more in detail with regions of low,
probability density. What inspires this suggestion is my lifelong j
repugnance to such statements as that of Bertrand Russell, that-

74 / Determinism and Freedom

if we only wait long enough we shall certainly see a pail of
water freeze on the fire. This may be conceded to be a rigorous
consequence of the mathematics, but to me it is utterly incred-
ible as a statement of physical fact. I would propose the ques-
tion: Can phenomena be found anywhere corresponding even
remotely to Russell's freezing of water on a fire? A more con-
servative proposal would be the investigation of the fluctua-
tions of fluctuation phenomena.

The second direction in which I would look for new types
of experimental evidence is suggested by our so-called free-
dom to decide whether the electron shall have position or
velocity. My question is : What are the limitations on the time
at which we have to decide which it shall be? Ordinarily we
make our choice long in advance; we set up a position ap-
paratus or a velocity apparatus and then wait to catch an elec-
tron in it. But what happens if we make our choice after we
have caught our electron? Specifically, imagine an interfer-
ence apparatus in which an electron passes through one hole,
presently meets a first screen with two holes, and still later
impinges on a second screen, where it forms part of an inter-
ference pattern. If there had been only one hole in the first
screen we would have got a pattern entirely different from that
obtained with a two-hole screen. My question now is: What
happens if we change from a one-hole screen to a two-hole
screen (or vice versa) after the electron has passed the first
hole and while it is in transit to the first screen? The same sort
of experiment might even more instructively be performed
with light quanta rather than with electrons. The general
problem might be to study in closed systems the fine structure
of macroscopic propagation phenomena that occur in times
less than the time required by light to travel across the system.
Such experiments have not yet been accessible to technique,
but I am told that they soon will be. They might throw light
on such problematic questions as the supposed instantaneous
collapse of the psi function when new information is ac-
quired at any point of a system.

Finally, I should like to suggest what seems to me a limita-
tion on all quantum theoretical analysis, or, for that matter,
on the criticisms of that analysis. It has been a new insight, at
\ least for the physicist, that the object of knowledge is not to
\ be separated from the instrument of knowledge. This has
\ I often been expressed by saying that an analysis of the role of
\ the observer is necessary to an understanding of quantum
\ theory. There the observer is identified with the instrument of

Determinism in Modem Science / 75

knowledge. It is well recognized that there is no sharp divid-
ing line between the instrument of knowledge and the object
of knowledge, and that for different purposes the line may be
drawn at different places. Thus it may be drawn at the photo-
graphic plate on which some event is recorded, or at the retina
of the eye of the experimenter who examines the plate, or at
the neural processes in the brain produced by the events at the
retina. Wherever we draw the line, we eventually end with
knowledge. Seldom, if ever, does the physicist ask what the
nature of this knowledge is, because for his purposes he does
not need to. If one wants a specific description of what the
physicist's purposes are, the answer can be found by examin-
ing what he does with his quantum theory: his use of it, for
the most part, is limited to the atomic events he manipulates
in his laboratory. But for purposes wider than those of the
physicists, in particular for the purpose of acquiring the full-
est possible understanding, we cannot forgo asking what this
knowledge is that for the physicists is the end product. Here
we obviously enter a field of great vagueness. From one point
of view a question has no meaning if we cannot formulate
what sort of an answer we should be willing to accept, and it
must be admitted that it is not easy to formulate the condi-
tions that should be satisfied if we are to feel that we had an
understanding of the nature of knowledge. But I think that
our groping quest may nevertheless have some direction.
What gives me most disquietude here, and what I think quan-
tum theory ignores in spite of its ostensible concern v/ith the
observer, it the simple observation that knowledge never oc-
curs except in conjunction with a nervous system that has
itself been subjected to a most elaborate preconditioning.
Furthermore, the nervous system is itself a physical system
of a complexity so formidable that we are only beginning to
have an inkling of how it functions. It seems to me that we
cannot permanently be satisfied with an analysis that purports,
as does quantum theory, to reduce to understandability the
functioning of the ultimate structural units of the universe,
when that understandability itself demands the cooperation of
superstructures of as yet unfathomed complexity.
' "-It seems to me that when we take full cognizance of the fact
that we cannot get away from our nervous systems, we shall
recognize intellectual limitations that we at present ignore.
With such a background I would say to those who think quan-
tum theory does not make sense that it at least makes sense
that it should not make sense.

Chapter 2

The Relativity of Defermlfiism

Milton K. Munitz, New York University

The lack of consensus in current discussions about determin-
ism in modern science, as Professor Bridgman has made
amply clear, is not in any way due to a disagreement either
about the experimental facts or about the mathematical calcu-
lations employed in various formulations of physical theory.
Professor Bridgman ascribes the defense of a deterministic
view on the part of some writers to extrascientific motives of
a specifically emotional sort, involving, among other things,
a longing for a "friendly universe," one in which regularities
are pervasively present. In briefly setting out his own views
he has indicated why it makes sense to say, not that the uni-
verse is unfriendly, but that causality, as he puts it, "should
fail for small objects." Professor Bridgman, it would seem,
finds a universe in which there is some objective element of
indeterminism, one in which he finds himself perfectly at
home.

Now, whether or not emotional considerations are relevant
factors in either accepting or rejecting a determinist view,
there is another basis of approach, to which Professor Bridg-
man alludes, that is of greater moment. He remarks that "we
cannot forgo asking what this knowledge is that for the physi-
cist is the end product." This is precisely the important kind
of question to raise. It indicates at once that what we are con-
cerned with is a philosophical question. Unfortunately, for
Professor Bridgman this means, as he says, that at this point
we "enter a field of great vagueness." Perhaps so. But whether
or not the issue of determinism is vague in principle because
it is a philosophical one, this issue, in so far as there is one,
is not in any case internal to science; nor is it to be settled
on emotional grounds. (I say "in so far as determinism is an
issue" because the term is sometimes used in a purely techni-
cal sense as a label for the specific type of differential equa-
tions and their solutions in classical Newtonian mechanics;
analogously, the equations of modern quantum mechanics,
different in certain crucial respects, are called "indetermin-

76

The Relativity of Determinism / 77

istic") The issue of determinism is more nearly indicated by
the problems that arise when one tries to decide whether to
agree with Professor Bridgman when he says: "a law of
causality or a deterministic structure for the universe or an
vmderlying continuity as distinguished from discreteness is
not ... a necessary prerequisite to rational thought." This
type of question cannot be settled simply by consulting the
above-mentioned "technical" meanings of determinism as
used to label certain types of equations in physics. If settled
at all, the issue of determinism is one to be resolved by trying
to clarify the cognitive status of various types of statements
in science, and among them a type of statement of which the
principle of indeterminacy is a particularly intriguing example.
It is the complex of differences in the philosophy of science,
not in science, that emerges here and that helps us focus on
the point where controversy genuinely arises.

Perhaps I can best bring out my main point if I call atten-
tion to what at first sight might seem to Professor Bridgman
and others a somewhat strange coupling of his own views with
that of a philosopher-scientist who, as he has reminded us,
has taken an almost intransigent view, wholly opposite to his
own, with respect to determinism: namely Einstein. I would
suggest that Professor Bridgman, like many other "indeter-
minists," has more in common with determinists like Einstein,
from one philosophical point of view, than is sometimes
recognized. What they have in common, indeed, is of more
interest and importance than their differences. In order to sup-
port this way of approaching the question, I should like to
suggest in barest outline what seems to me at least one other
entirely different way of looking at the matter, one that
serves as a major alternative to this first type of philosophy.
There is no question here of course of trying to prove the
superiority of one philosophy of science over another. At
best, what we may hint at is the manner in which certain diflS-
culties and controversies dissolve when we shift our perspec-
tive in a suflBciently fundamental way. The thesis I am
proposing, in short, is that the whole controversy between
determinism and indeterminism as it is customarily formu-
lated is not finally to be settled by taking sides and showing
the overwhelming merits of one of them. It is to be "settled"
by taking seriously a philosophy of science in which the terms
"determinism" and "indeterminism" not only imdergo radical
change, but in doing so, require all scientific knowledge in-
volving the use of theory to be deterministic.

78 / Determinum and Freedom

Various terms are at hand by which we may label these
rival philosophies. Karl Popper, for example, has recently
suggested the terms "essentialism" and "instrumentalism" as
suitable ones, while the terms "realism" and "conventional-
ism" may be preferred by others. Since, however, I wish to
stress that one whole group of views, whether "determinist"
or "indeterminist," has reference to what are alleged to be the
structural features of events or situations or objects as "dis-
closed" by scientific inquiry, I shall refer to this approach as
basically "ontological." According to the other approach,
"causality," "determinism," "necessity," "predictability," and
other associated terms have to do primarily with the connec-
tions between statements, which represent what is or is not
present in the inference patterns of the scientist. This I shall
call the "logical" approach.

Einstein's lifelong opposition to those who regarded quan-
tum physics as pointing to an element of genuine indetermin-
ism or unpredictability among some of nature's events was
based on a profound belief in what he thought of as the
intelligibility of the universe. It was not simply that he
thought, for example, that the program of field physics to
which he devoted aU his energies would be more successful as
a heuristic device for achieving better predictions or explana-
tions, or that it was simply a more "congenial" point of view.
Rather, he looked on nature in Spinozistic terms and referred
to his Spinozism as his "religion." He thought of nature, to
use his own favorite metaphor, as rather like a puzzle. The
scientist tries to fit the pieces together or obtain solutions to
his queries by seeking the uniquely correct way of specifying
what the structvu-e is. Einstein believed the program of field
physics to hold out the promise of eventually realizing this
goal to a superior degree in comparison with any other. Ein-
stein's dissatisfactions with quantum physics arose from the
conviction that its conceptual foundations were essentially in-
complete, that because of its statistical character quantum
theory referred to "ensembles of systems and not to individual
systems."

If one is going to criticize such a view philosophically, one
can perhaps best begin by showing that the "root metaphor"
on which it rests — that nature is like a puzzle with a unique
solution — is really one that turns out to be cramping rather
than liberating in its implications. Real puzzles are con-
structed by men for men, and we know in advance that there
is a favored solution. But why think of nature as a puzzle?

The Relativity of Determinism / 79

Men, being curious or troubled, find problems in nature, and
in their religious moments they may even wish to dwell on
an element of ineffable mystery connected with it. Given the
problems that scientists encounter and pick out, they can offer
a variety of possible answers. But why need we assume that
there is some one supremely satisfactory way of dealing with
these problems? Einstein's philosophy, one might say, is the
outcome of projecting into the allegedly antecedent and in-
dependent structure of facts what is at bottom a favored view
but still only one of many possible theoretical views of things.
His realistic rationalism is the result of reifying the concep-
tualizations of a legitimate yet limited perspective into the
necessities of things and events. The successes of the special
and general theory of relativity in yielding explanations and
predictions of a more adequate sort than those provided by
earlier or rival theories — these the working physicist recog-
nizes and gladly welcomes without having to subscribe to the
ontologism of Einstein's philosophy of science. Such a work-
ing physicist may rightfully insist on the use of quantum
physics, even if in its conceptual basis it is wholly different
from what Einstein would think of as desirable. And again he
would do so on the simple pragmatic ground that it is suc-
cessful in its predictions and explanations to a sufficiently high
degree to warrant his use of it. But need there be any logical
incompatibility in using side by side the wholly different
techniques of inference or different analogical modes of view-
ing things that constitute the foundations of these various
theories? Not, he may say, if we do not commit ourselves in
advance to the belief that nature has a structure that will
yield its secrets to one and only one of these formulations.
This is another way of saying that we have no compelling
reason to believe that the theories with which the physicist
works are, or aim to be, true, that they would disclose to us,
as Galileo put it, "the language in which the Book of Nature
is written." For nature is not really like a book written in a
language at all. Hence nature is neither logical nor illogical,
neither mathematical nor nonmathematical, neither rational
nor irrational, and so neither determined nor undetermined.
These predicates apply only to statements in human discourse,
and the way in which such statements fit or not into certain
accepted patterns of inference, patterns that in the case of
empirical science are set up by men for dealing with the facts
of observational experience.

Now, unlike Einstein's, Professor Bridgman's orientation to

80 / Determinism and Freedom

physics is primarily that of an experimentalist, and this fact
controls his interpretation of what is contained in the dis-
tinctive ideas of quantum physics. If we examine the manner
in which he expounds and illustrates the main principles of
this physics, it becomes clear that he reads them as basically
experimentally verifiable reports of what is found in the
laboratory. First, Professor Bridgman tends to equate the
physicist's process of acquiring knowledge with the process of
acquiring information through the use of appropriate instru-
mentation. It makes sense, he argues, on the submicroscopic
as well as on the macroscopic level that the very process of
acquiring information (knowledge) should interfere in some
way with the "objects" whose "traits" are being examined, so
that a more comprehensive perspective would include this
very interaction itself in all its complexity, extending even so
far as the still undisclosed intricacies of the human nervous
system. Secondly, Heisenberg's principle of uncertainty is
taken to describe the relation of relative fuzziness in the meas-
urements, experimentally arrived at, of the velocity and posi-
tion of subatomic particles. Finally, Professor Bridgman
interprets what we are to understand by such a phrase as the
"individuality of the electron" as not referring to some object
existing in itself apart from the observer's interaction with
"it," but as a more complex situational event. "The electron,"
he says, "in and by itself does not 'have' either position or
velocity; these pertain to what happens in the whole situation,
including the instrumentation."

Granted the obvious importance of the experimentalist's
instrumentation in obtaining information as a necessary and
relevant ingredient in the process of obtaining knowledge of
the physical world, it seems to me that the stress Professor
Bridgman gives to the virtual identification of information
and knowledge slurs over certain important distinctions be-
tween the roles of theory and observation. Now, theory is
primarily a matter of conceptual construction, having, as all
thought must, a neurological base; but to speak of this element
in knowledge as interacting with the object seems to me some-
what queer and indeed confusing. It results from trying to
put under the heading of apparatus what is actually not ap-
paratus at all. Moreover, this way of approaching the matter
has a further and perhaps more serious consequence. One as-
sumes that theory is at best only a report of what one finds in
experiment. The fact is, however, that Heisenberg's principle,
for example, is not a report of what one discovers as a result

The Relativity of Determinism / 81

of trying to perform certain measurements. It is, instead, a
defining element of quantum theory, and so, among other
things, a way of specifying what for the purposes of the
theory is to be regarded as a "particle." This is not the "parti-
cle" of traditional mechanics, since, even on the level of
theory, there is no possibility of simultaneous specification to
any desired degree of accuracy of both its position and vel-
ocity. What we have here, then, is a part of a diJB^erent and,
in some ways, more complex set of theoretical ideas by means
of which it is attempted to provide explanations and predic-
tions for observed or observable facts. Similarily, what is re-
ferred to as an "electron" or "photon" is not an object of
direct observation, but represents in summary fashion a
whole schema for interpretational devices and inferential
techniques of theory that are brought to bear in thinking or
reaching conclusions about observational materials. Professor
Bridgman, it seems to me, though appearing to rely exclu-
sively on what one finds in the laboratory, is actually letting
in a good deal in the way of constructive theory and is thus
perhaps giving a greater ontological weight to the terms of
theory — as if these corresponded to some structural features
of physical events — than need be done. He finds indeter-
minism in the universe not really because the measurements
of the laboratory force this interpretation, but because he is
willing to accept the orthodox interpretation of quantum
theory as describing the alleged structure of certain physical
events. In displaying this confidence in orthodox quantum
theory, he is expressing, I suspect, the same type of confidence
in what theory can accomplish as Einstein, although the
specific ontology, to be sure, is different.

Now, instead of letting the reifications of one's ontological
approach to scientific theory came in through either the front
door or the back door, one might say, in the present context
of discussion, that quantum theory is as deterministic as
classical physics, or indeed as any theory, by its very nature,
is bound to be. This means simply that a specific theory, what-
ever its conceptual tools or analogical base, offers us a way,
distinctive to itself, of making inferences from certain obser-
vationally identifiable facts to others, and of interpreting
what those experimental facts are. Any conclusions reached
in accordance with the rules of inference specified by the
theory are logically determined. Thus the principle of in-
determinacy, among others, is part of a total physical theory

82 / Determmism and Freedom

that in its functioning provides determined results of a type
it is competent to reach. If it proves successful in dealing with
observational facts, it is not because "causality fails for small
objects" but because man's creative intellectual ingenuity has
found one more means for dealing effectively through his
inferences with the data of his experience.

Chapter 3

The Case for Indeterminism

Alfred Lande, Ohio State University

Professor Bridgman has painted a lively picture of the strife
about determinism. It is a strife not so much about empirical
facts as about the interpretation of facts. Many consider de-
terminism the only rational interpretation — if not an aprioris-
tic category of the mind. They v/ould admit indeterminacy
in modern physics only as a temporary expedient adapted to
our present incapacity to predict microphysical events, an
incapacity that may be overcome by future technical develop-
ments. In opposition to this view I agree with Bridgman
when he said: "it makes sense that causality should fail for
small objects." But I would go even further and maintain that,
irrespective of any possible future technical developments,
determinism does not and never will make sense,^ in particu-
lar when applied to those random-like situations we know
from games of chance. The question is, of course, whether
there are true games of chance, i.e., random-like situations
that are irreducible to concealed causes in principle. In this
argument about principles it makes little difference whether
the random-like situations are those encovmtered in atomic
experiments, in dice games, or in the games insurance com-
panies play with their clients. Therefore I shall have to extend
our topic from determinism in modern science to natural
science in general. My excuse for this deviation is Bridgman's
own declaration: "It is to be expected that the roots of the dif-
ficulties revealed to us by quantum theory are already present
in the sphere of ordinary life, and should be discoverable by
acute enough analysis." Let us therefore analyze an example
taken from ordinary life.

Imagine a game of balls dropped through a chute onto a
knife edge. The angle of aim a may be adjusted by a screw,
and the knife edge may be sharp or slightly rounded. If the
chute is aimed at the right (left) of center, all balls will drop
to the right (left). Experience shows, however, that between

1 Cf . the critique of determinism by K. R. Popper, Brit. J. Philos. Science
1, Nos. 2 and 3.

83

84 / Determinism and Freedom

right- and left-hand aim there is always a small but finite
range Aa of aim within which an experimentally adjusted
angle a leads neither to all balls dropping to the right nor to
all balls dropping to the left but rather to both r- and /-balls
occurring at a certain frequency ratio. The latter varies from
100 : 0 to 0 : 100 when the aim is shifted from the right to
the left of the small range Ao. Primitive persons and other in-
determinists will interpret this as a sign of uncertainty, of
blind fate, with one and the same cause capable of being fol-
lowed by two different effects, r or /. Determinists will say,
however: "The distribution of the r- and /-results only appears
to be erratic. Actually each individual result has its particular
deterministic cause, be it a small deviation of the angle of
aim, or a small perturbation of the ball on its flight." (Simi-
larly, although insurance companies count on their frequency
tables, each individual "accident" is not an accident but has
its particular cause.)

I submit, however, that the hypothesis of (concealed) indi-
vidual causes behind individual effects r or / does not explain
the essential point of the observed situation in a deterministic
fashion. When the determinist ascribes the present final event
r to an r-producing chain . . . r r r reaching back into the in-
finite past, he merely shifts the problem r- and /-events to r-
and /-chains and, further, to the beginning of those chains, if
they have a beginning. We must ask him now for a determin-
istic explanation of the strange empirical observation that
those chains, or initial conditions, occur again and again at
Va definite frequency ratio and, furthermore, why even the
' fluctuations away from the average occur at a rate conforming
i with the mathematical theory of random as though by a pre-
-established harmony between fact and theory. It is this pre-
fstablished harmony that calls for explanation. Referring to
ithe infinite past, and saying that his harmony has always pre-
vailed, is an evasion rather than a deterministic explanation.
A stubborn determinist may defend his cause, however, by
means of the following argument: "Once upon a time there
was a demon who knew his mathematical random theory
and who deliberately went out to deceive the observer.
He first initiated two /--chains, then an /-chain, then four
r-chains; then realizing that he had given too much pre-
ponderance to r-chains, he thereupon started five /-chains
in a row, cleverly arranging the whole sequence with
averages and fluctuations so that a present-day scientist
might be lured away from the true deterministic faith."

The Case for Indeterminism / 85

There seems indeed to be only the following alternative:
Either the observed random-like distributions of final events
or chains or initial conditions in games of chance repre-
sent a basic and irreducible trait of nature. Or statistical dis-
tributions only feign an appearance of random, when in real-
ity there is, or has been, concerted deterministic action. Either
a deus ex machina or no deteministic explanation at all. Since
deceitful demons have no place in scientific theories, I have
reluctantly joined the party of indeterminacy pure and
simple. But I concede, that it is a party of renunciation with a
purely negative creed. Most of my partisans, including my-
self, suffer from a guilt complex that draws us toward our old
infatuation, determinism. This infatuation may have its roots
in a feeling of being ourselves demons who can deliberately
start deterministic chains. In other words, it may be that we
believe in strict determinism because we feel we have free wUl
— a somewhat paradoxical psychological hypothesis. But as a
scientist who observes games of chance, and who is unwilling
to admit a deus ex machina (at the beginning of time, if there
is a beginning, or a finite time ago), I must concede that the
deterministic interpretation fails; and this applies not only to
ordinary "games of chance" in which the statistical dispersion
is obvious, but in general to those cases where a similar dis-
persion of effects is revealed only by microphysical in-
struments.

Empirically it is a most surprising fact, which could not
have been foreseen a priori, that there are sequences of events
in harmony with mathematical random theory. This empirical
discovery was already made by cavemen when they gambled
for the best pieces of a slain bear. But when irreducible ran-
dom is once accepted, then it is a comparatively minor point
of dispute whether (a) each new experiment constitutes a new
game of chance (as quantum theory maintains), or {b) ran-
dom was set up once, a long or an infinite time ago, and ran-
dom distributions observed at present are but the deterministic
effects of that one initial "shuffling of the cards" (as classical
statistical mechanics maintains). The difference between the
two positions may be illustrated by the example of a great
number of arrows laid out as radii of a circle. Each individual
arrow has its definite x- and y-component, positive or nega-
tive. Suppose that the arrows are distributed over all directions
in a random-like fashion, with or without the help of a demon.
When viewing this ensemble through a circular glass disk, the
northern half of which is transparent and the southern half

86 / Determinism and Freedom

opaque, one observes a statistical 50 : 50 ratio between visible
and invisible arrows. The same ratio of visible to invisible ar-
rows will be observed when the disk is held in the NE-SW-
direction or in any other direction whatsoever. The frequency
ratios in all these instances are deterministic consequences of
the one original random distribution over all directions within
the circle. Also, when an individual arrow of given x- and
given y- component is viewed through our glass disk held in a
given orientation, then it is predictable whether this arrow will
be found belonging to the visible or the invisible group in that
particular orientation of the disk. This is an illustration of the
classical situation. Since it requires random mixing at least
once, it is a soft determinism at most.

Suppose now that our arrows are of microscopic magnitude
and represent the spin directions of electrons. In this case not
even a demon can lay out various spins in a variety of chosen
directions simultaneously. Electronic arrows (spin directions)
can only be laid out by means of a linear slit (a magnetic
field). When the slit is held in the N5-direction the arrows will
assume only the A^- and/ or 5-direction. If the arrows are now
subjected to a new orientation test by means of a slit of NE-
5fF-direction, each arrow has to jump from its original N- or
5-direction either to the NE- or to the 5Pr-direction. The in-
deterministic character of these jumps is proved, according to
our former argument, by the mere fact that each new slit ex-
periment actually yields a new statistical frequency ratio be-
tween opposite arrows that, together with averages and fluctu-
ations, conforms with the mathematical theory of random
events. In the present instance the subsequent N : S and NE :
SW ratios cannot be regarded as deterministic consequences
of an original random distribution over all directions, since
each new slit experiment collects all arrows within two op-
posite directions only. Each new test is an independent game
of chance in which it is unpredictable whether a given arrow
will choose one or the opposite direction. This is the obvious
and trivial inference from the fact, discovered by quantum
physicists, that there are mutually incompatible quantities
such as X- and j-components of an electronic spin vector (or
an j:-component of position and an a:-component of momen-
tum of any particle), so that one cannot provide an original
random-like outlay of spins over all directions.

Quantum theorists, however, do not seem to cherish such
simple and elementary considerations, since they always refer
to J. von Neumann's proof that quantum theory has no room

The Case for Indeterminism / 87

for concealed causal mechanisms. But what did von Neumann
do to prove this result? He first introduced the whole involved
apparatus of the quantum theory, thereby lifting us to a highly
exalted plane. In this lofty atmosphere he carried out certain
calculations, then brought his results back to earth again —
when he could have stayed there all the time. Yet when von
Neumann's conclusions were recently found to suffer from the
fault of circularity, the determinists took heart again. Let it be
said, however, that irrespective of von Neumarm and irrespec-
tive of the technicalities of quantum mechanics, the mere ex-
istence of mutually incompatible quantities a and b that re-
quire statistical redistributions in consecutive a- and 6-tests
leaves us no other alternative than either to assume a demon
who feigns pseudo-random, or to concede that the harmony
between statistical fact and random theory is deterministically
inexplicable.

The wide acclaim that von Neumann's (inconclusive) proof
has found among theoretical physicists is symptomatic of the
present vogue for considering quantum mechanics with all its
subtleties, wave-particle antinomies, and bewildering mathe-
matical prescriptions as the ultimate and irreducible ground
structure, the deepest bottom of theoretical analysis. This
fundamentalist dogma is quite in opposition to Bridgman's
and my own view that the enigmatic quantum laws ought to
be explainable as consequences of more fundamental laws
having their roots in the experiences of everyday life. In order
to support this view I should like to ask my determinist op-
ponents: "Do you accept the principle, abstracted from every-
day experience, that a finite effect calls for a finite cause, that
a finite change of effect can never be produced by an infinitely
small change of cause? If you accept this principle of con-
tinuity of Leibnitz, you have already lost your case!"

To return to our ball-knife game; If the chute is aimed at
the right (left) of the knife edge, then all balls will drop to the
right (left) of the knife. But suppose that the angle a of aim
is gradually shifted from right to left. Will this lead to an
abrupt change of effect, from all balls dropping to the left,
at the moment when the aim passes the center? If so, it would
constitute a violation of the postulate of continuity. According
to this postulate there ought to be a finite range Aa of transi-
tion from the one extreme case to the other — a range Ao
within a given angle a of aim leads to the dropping of a nu-
merical fraction of all balls to the right and of the remaining
fraction to the left. This is exactly the situation observed, dis-

88 / Determinism and Freedom

cussed previously. The postulate of continuity thus leads to
the same statistical situation by inference that we have ac-
cepted before simply as a strange matter of fact discovered by
cavemen.

A slight variant of the game of dropping balls on a knife
edge leads to a vi'ell-knovi'n phenomenon of quantum physics.
Consider a stream of particles incident on a filter of some
sort where they have to choose between passing and not pass-
ing. Let us define an /4 -filter as an instrument that passes the
incident particles when they are in the state A and rejects
them when they are in a state different from A . Suppose, first,
that our particles are in a state A and that we now modify
them to a state B differing from A by some amount, however
small. Are we to expect that this infinitely small change of
state will abruptly convert all passing particles into blocked
particles? According to the continuity principle there ought to
be a gradual transition from the case B = A, with all B's
passed, to the case B = A, with all B's rejected by the A-
filter. That is, there ought to be intermediate cases of a
"fractional equality," written B — A, in which neither all B's
pass nor all are blocked — cases, then, in which a "splitting"
of the incident B-state particles into a passing and a blocked
fraction takes place. The continuity postulate in application
to the sharp contrast between equality and inequality thus
leads to the same "splitting effect" by inference that quantum
physicists have discovered by delicate microphysical experi-
ments.

Let us further introduce the principle of reproducibility of
the result of a measurement. In the above experiment it re-
quires that a B -state particle that has passed the /4 -filter once
will pass an /I -filter again, the second time with certainty;
that is, the S-particle must have "jumped" in its first passage
to the new state A where it now remains (unless thrown into
another state by a third test by means of a B- or C-filter).
Similarly, a B-particle once rejected by the /i -filter will be
rejected again in another /l-filter test; that is, it must have
jumped from the state B to a state non-A, completely differ-
ent from A. The observation that there is a definite statistical
frequency ratio between passed and blocked particles shows,
according to our previous argument, that those "quantum
jumps" from B to A and/ or non-A represent a true game of
chance that cannot be referred to hidden deterministic causes.

One may supplement the postulate of continuity and repro-
ducibility by a few other postulates of a simple and elementary

The Case for Indeterminism / 89

sort and from there arrive by inference at the mathematical
and conceptual schema known today as the quantum theory ^
— thereby confirming Bridgman's opinion that the roots of
this theory should be discoverable in the sphere of ordinary
experience, and also confirming the statistical interpretation
of Bom, Heisenberg, and Bohr. When the first surprising
quantimi facts, such as E = hv, were brought to light, the
theorists tried to explain them on the grounds of accepted
principles of mechanics. When this attempt proved a failure,
quantum theory with its various rules and prescriptions was
finally accepted as not further reducible, as "fundamental." It
also seemed so very convenient to listen to the positivistic
siren song, that physicists should observe and describe, and
not question why things are as they are. The negative attitude
is accompanied today by ever repeated exhortations that the
principle of complementarity is the most profound and sub-
lime law of nature, applicable not only to microphysics but
also to biology, psychology, and lately even to the relation
between the sexes and among political power groups. In op-
position to this cult it is hoped that the recently revived search
for simple and general postulates underlying quantum theory
may bring us one step nearer the goal of aU scientific en-
deavor, which, according to a weU-coined phrase of Bridg-
man's, is the goal of satisfying our curiosity.

' A. Lande. Brit. J. Philos Set., 6, 300, 1956. Also Foundations of Quantum
Theory, A Study in Continuity and Symmetry (New Haven: Yale University
Press, 1955).

^er 4

Determinism and the Cosmos

Dennis W. Sciama, Trinity College, Cambridge

As A PHYSICIST I have found the following working hypothesis
very useful: violent controversy about a scientific problem
is a sign that some simple essential consideration is missing.
The polemic, as it were, tries to substitute for the missing
point, but of course it never can. I think for instance that this
has been so in discussions of Mach's principle of the origin
of inertia, and also of the problem deducing irreversible
macroscopic behavior from reversible microscopic laws.

Bridgman has reminded us that the physicists are con-
ducting a violent controversy about the meaning of quantum
mechanics. This situation is in striking contrast to that pre-
vailing in classical mechanics; for although classical me-
chanics is known to be false, there is no dispute as to its
meaning. It is only in quantum mechanics (which is known
to be true!) that there is such a dispute. Application of the
working hypothesis suggests that some simple point has still
to be made. My aim in these remarks is to propose one pos-
sibility for this simple point, a proposal based mainly on the
work of Dr. K. V. Roberts.

The basic way in which quantum mechanics differs from
classical mechanics is the following: our inferences about the
future must be expressed in terms of probabilities. This intro-
duction of probability would enable us to make the calcula-
tion.

With this state of affairs in mind, let us make a new
assumption. Let us suppose that in nature systems are deter-
ministic in the sense that we can calculate the state of a
system at time t if we know enough boundary conditions re-
ferring to times other than t; but let us differ from classical
mechanics by supposing that nature is so constructed that,
roughly speaking, half the boundary conditions must refer to
the past and half to the future of the moment t. In other
words, we assume that nature is such that "mixed" boundary
conditions are always needed.

Presumably a system with such properties would be called
deterministic. This is a matter of definition, of course; what
is really important is that the behavior of the system is as

90

Determinism and the Cosmos / 91

well defined and intelligible as that of a system obeying classi-
cal mechanics. But now we must ask: How would a "mixed"
system appear to an observer who himself is part of the
system?

Now, such an observer, for reasons that cannot be elabo-
rated here but that have to do with the second law of
thermodynamics, is acquainted only with the past. Hence if
he attempts to calculate the state of a system at a time t in
his future, he will find that he cannot do so, for he does not
know all the boundary conditions. His knowledge of the past
boundary conditions will delimit the possibilities considera-
bly, but it is clear that to the observer the system will appear
to contain indeterminate elements.

What sort of a theory wUl such an observer devise? In effect
he will be forced to average over all those future boundary
conditions that are compatible with his present knowledge.
(Of course, at first he will not realize that this is what he is
doing.) That is to say, he will be forced to introduce a
probability calculus to account for his observations. The
suggestion is that this probability calculus is just quantum
mechanics.

In this way the correctness of quantum mechanics can be
reconciled with a deterministic universe. In the language of
von Neumann, there are hidden variables; they escape his ban
because they refer to the future.

We are now in a position to answer the question: Is quan-
tum mechanical probability subjective or objective? We have
seen that the probability arises from the observer's ignorance
of some of the determining conditions. The probability is
therefore subjective.

So far the discussion has been academic in the sense that
no new physical results have emerged. However there is an
interesting possibility in this direction. For on the view pre-
sented here quantum mechanics is no longer a primitive
theory; it is a formalism that is derived from a more basic
theory. Now, Planck's constant, h, is a measure of the
"amount" of deviation from classical mechanics. In quantum
mechanics as it stands today the numerical value of this con-
stant is completely arbitrary. However, if quantum mechanics
is deducible from a more basic theory, then presumably h,
which is here a measure of our ignorance of the future, will be
expressed in terms of quantities fundamental to the basic
theory. Such a relation could be tested experimentally, and
so the theory could be checked.

PART ill

m L@w mnd Etiiics

Chapter 1

Legal Resporaslbiiify csnd Excuses

H. L. A. Hart, Oxford University

It is characteristic of our own and all advanced legal sys-
tems that the individual's liability to punishment, at any rate
for serious crimes carrying severe penalties, is made by law
to depend on, among other things, certain mental conditions.
These conditions can best be expressed in negative form as
excusing conditions : the individual is not liable to punishment
if at the time of his doing what would otherwise be a punish-
able act he is, say, unconscious, mistaken about the physical
consequences of his bodily movements or the nature or quali-
ties of the thing or persons affected by them, or, in some
cases, if he is subjected to threats or other gross forms of
coercion or is the victim of certain types of mental disease.
This is a list, not meant to be complete, giving broad descrip-
tions of the principal excusing conditions; the exact definition
of these and their precise character and scope must be sought
in the detailed exposition of our criminal law. If an individual
breaks the law when none of the excusing conditions are
present, he is ordinarily said to have acted of "his own free
wUl," "of his own accord," "voluntarily"; or it might be said,
"He could have helped doing what he did." If the determin-
ist ^ has anything to say on this subject, it must be because he

^ Earlier papers in this session will doubtless Iiave specified the variety of
theories or claims that shelter under the label "determinism." For many pur-
poses it is necessary to distinguish among them, especially on the question
whether the elements in human conduct that are said to be "determined"
are regarded as the product of sufficient conditions, or sets of jointly sufficient
conditions, which include the individual's character. I think, however, that the
defense I make in this paper of the rationality, moraUty, and justice of quali-
fying criminal responsibility by excusing conditions will be compatible with
any form of determinism that satisfies the two following sets of requirements.

A. The determinist must not deny (a) those empirical facts that at present
we treat as proper grounds for saying, "He did what he chose," "His choice
was effective," "He got what he chose," "That was the result of his choice,"
etc; (b) the fact that when we get what we chose to have, live our lives as
we have chosen, and particularly when we obtain by a choice what we have
judged to be the lesser of two evils, this is a source of satisfaction; (c) the
fact that we are often able to predict successfully and on reasonable evidence

95

96 / Determinism and Freedom

makes two claims. The first claim is that it may be true —
though we camiot at present and may never be able to show
that it is true — that human conduct (including in that expres-
sion not only actions involving the movements of the human
body but its psychological elements or components such as
decisions, choices, experiences of desire, effort, etc.) is sub-
ject to certain types of law, where law is to be understood in
the sense of a scientific law. The second claim is that, if
human conduct so understood is in fact subject to such laws
(though at the present time we do not know it to be so),
the distinction we draw between one who acts under excusing
conditions and one who acts when none are present becomes
imimportant, if not absurd. Consequently, to allow punish-
ment to depend on the presence or absence of excusing con-
ditions, or to think it justified when they are absent but not
when they are present, is absurd, meaningless, irrational, or
unjust, or immoral, or perhaps all of these.

My principal object in this paper is to draw attention to
the analogy between conditions that are treated by criminal
law as excusing conditions and certain similar conditions that
are treated in another branch of the law as invalidating cer-
tain civil transactions such as wills, gifts, contracts, and mar-
riages. If we consider this analogy, I think we can see that
there is a rationale for our insistence on the importance of
excusing conditions in criminal law that no form of determin-
ism that I, at any rate, can construct could impugn; and this
rationale seems to me superior at many points to the two
main accounts or explanations that in Anglo-American juris-
prudence have been put forward as the basis of the recogni-
tion of excusing conditions in criminal responsibility.

In this preliminary section, however, I want to explain why
I shall not undertake the analysis or elucidation of the mean-
ing of such expressions as "He did it voluntarily," "He acted
of his own free will," "He could have helped doing it," "He

that our choice will be eSectiye over certain periods in relation to certain
matters.

B. The determinlst does not assert and could not truly assert that we al-
ready know the laws that he says may exist or (in some versions) must exist.
Determinists diflEer on the question whether or not the laws are sufficiently
simple (a) for human beings to discover, (fc) for human beings to use for
the prediction of their own and others' conduct. But as long as It is not
asserted that we know these laws I do not think this difference of opinion
important here. Of course If we knew the laws and could use them for the
detailed and exact prediction of our own and others' conduct, deliberation
and choice would become pointless, and perhaps in such circumstances there
could not (logically) be "deliberation" or "choice."

Legal ResponsibUity and Excuses / 97

could have done otherwise," I do not, of course, think the an-
alysis of these terms unimportant: indeed I think we owe the
progress that has been made, at least in determining what
the "free will problem" is, to the work of philosophers who
have pursued this analysis. Perhaps it may be shown that
statements of the form "He did it of his own free will" or
"He could have done otherwise," etc., are not logically incom-
patible with the existence of the type of laws the determinist
claims may exist; if they do exist, it may not follow that
statements of the kind quoted are always false, for it may be
that these statements are true given certain conditions, which
need not include the nonexistence of any such laws.

Here, however, I shall not attempt to carry further any
such inquiries into the meaning of these expressions or to
press the view I have urged elsewhere, that the expression
"voluntary action" is best understood as excluding the pres-
ence of the various excuses. So I will not deal here with a
determinist who is so incautious as to say that it may be false
that anyone has ever acted "voluntarily," "of his own free
wUl," or "could have done otherwise than he did." It will
help clarify our conception of criminal responsibihty, I think,
if I confront a more cautious skeptic who, without commit-
ting himself as to the meaning of those expressions or their
logical or linguistic dependence on, or independence of, the
negation of those types of law to which the determinist refers,
yet criticizes our allocation of responsibility by reference to
excusing conditions. This more cautious determinist says that,
whatever the expressions "voluntary" etc. may mean, unless
we have reasonable grounds for thinking there are no such
laws, the distinctions drawn by these expressions cannot be
regarded as of any importance, and there can be neither rea-
son nor justice in allowing punishment to depend on the pres-
ence or absence of excusing conditions.

n

In the criminal law of every modern state responsibility for
serious crimes is excluded or "diminished" by some of the
conditions we have referred to as "excusing conditions." In
Anglo-American criminal law this is the doctrine that a "sub-
jective element," or "mens rea," is required for criminal re-
sponsibility, and it is because of this doctrine that a criminal
trial may involve investigations into the sanity of the accused,
into what he knew, believed, or foresaw; into the questions
whether or not he was subject to coercion by threats or

98 / Determinism and Freedom

provoked into passion, or was prevented by disease or transi-
tory loss of consciousness from controlling the movements of
his body or muscles. These matters come up under the heads
known to lawyers as Mistake, Accident, Provocation, Duress,
and Insanity, and are most clearly and dramatically exempli-
fied when the charge is one of murder or manslaughter.

Though this general doctrine underlies the criminal law, no
legal system in practice admits without qualification the prin-
ciple that all criminal responsibility is excluded by any of the
excusing conditions. In Anglo-American law this principle is
qualified in two ways. First, our law admits crimes of "strict
liability."^ These are crimes where it is no defense to show
that the accused, in spite of the exercise of proper care, was
ignorant of the facts that made his act illegal. Here he is
liable to punishment even though he did not intend to com-
mit an act answering the definition of the crime. These are
for the most part petty offences contravening statutes that
require the maintenance of standards in the manufacture of
goods sold for consumption; e.g., a statute forbidding the sale
of adulterated milk. Such offenses are usually punishable with
a fine and are sometimes said by jurists who object to strict
liability not to be criminal in any "real" sense. Secondly, even
in regard to crimes where liability is not "strict," so that
mistake or accident rendering the accused's action uninten-
tional would provide an excuse, many legal systems do not
accept some of the other conditions we have listed as ex-
cluding liability to punishment. This is so for a variety of
reasons.

For one thing, it is clear that not only lawyers but scientists
and plain men differ as to the relevance of some excusing con-
ditions, and this lack of agreement is usually expressed as a
difference of view regarding what kind of factor limits the
human capacity to control behavior. Views so expressed have
indeed changed with the advance of knowledge about the
human mind. Perhaps most people are now persuaded that it
is possible for a man to have volitional control of his muscles
and also to know the physical character of his movements
and their consequences for himself and others, and yet be
unable to resist the urge or temptation to perform a certain
act; yet many think this incapacity exists only if it is associ-

' For an Uluminating discussion of strict liability, see the opinion of Justice
Jackson in Mortsetts v. United States (1952) 342 U.S. 246; 96 L. Ed. 288; 78
S. Ct. 241. Also Sayre, "PubUc Welfare Offences," 33 Col. L. Rev. 58; Hall,
Principles of Criminal Law (Indianapolis: Bobbs-Merrill Co., 1947), chap. x.

Legal Responsibility and Excuses / 99

ated with well-marked physiological or neurological symp-
toms or independently definable psychological disturbances.
And perhaps there are still some who hold a modified form
of the Platonic doctrine that Virtue is Knowledge and believe
that the possession of knowledge ^ (and muscular control) is
per se a sufficient condition of the capacity to comply with
the law.*

Another reason limiting the scope of the excusing condi-
tions is difficulty of proof. Some of the mental elements in-
volved are much easier to prove than others. It is relatively
simple to show that an agent lacked either generally or on a
particular occasion volitional muscular control; it is somewhat
more difficult to show that he did not know certain facts
either about present circumstances (e.g., that a gun was
loaded) or the future (that a man would step into the line
of fire); it is much more difficult to establish whether or not
a person was deprived of "self-control" by passion provoked
by others, or by partial mental disease. As we consider these
different cases, not only do we reach much vaguer concepts,
but we become progressively more dependent on the agent's
own statements about himself, buttressed by inferences from
common-sense generalizations about human nature, such as
that men are capable of self-control when confronted with an
open till but not when confronted with a wife in adultery. The
law is accordingly much more cautious in admitting "defects
of the will" than "defect in knowledge" as qualifying or ex-
cluding criminal responsibility. Further difficulties of proof
may cause a legal system to limit its inquiry into the agent's
"subjective condition" by asking what a "reasonable man"
would in the circumstances have known or foreseen, or by
asking whether "a reasonable man" in the circimistances
would have been deprived (say, by provocation) of self-
control; and the system may then impute to the agent such
knowledge or foresight or control.'

' This view is often defended by the assertion that the mind is an "inte-
grated whole," that if the capacity for self-control is absent, knowledge must
also be absent. See Hall, op. cit., p. 524: "Diseased volition does not exist
apart from diseased intelligence"; also reference to the "integration theory,"
chap. xiv.

* English judges have taken different sides on the issue whether a man can
be said to have "lost self-control," and kiUed another while in that condition,
if he knew what he was doing and killed his victim intentionally. See Holmes
V. D.P.P. (1946) A.C. 597 (Lord Simon) and A.G. for Ceylon v. Kumaras-
Inghege v. Don John Perera (1953) A.C. 200 (Lord Goddard).

^ But see for a defense of the "reasonable man" test (in cases of alleged
provocation) Royal Commission on Capital Punishment, pp. 51-56 (S§
139-145). This defense is not confined to the difficulties of proof.

100 / Determinism and Freedom

For these practical reasons no simple identification of the
necessary mental subjective elements in responsibility, with
the fuU list of excusing conditions, can be made; and in all
systems far greater prominence is given to the more easily
provable elements of volitional control of muscular movement
and knowledge of circumstances or consequences than to the
other more elusive elements.

Hence it is true that legal recognition of the importance of
excusing conditions is never unqualified; the law, like every
other human institution, has to compromise with other values
besides whatever value is incorporated in the recognition of
some conditions as excusing. Sometimes, of course, it is not
clear, when "strict liability" is imposed, what value (social
welfare?) is triumphant, and there has consequently been
much criticism of this as an odious and useless departure from
proper principles of liability.

Modern systems of law are however also concerned with
most of the conditions we have listed as excusing conditions
in another way. Besides the criminal law that requires men
to do or abstain from certain actions whether they wish to or
not, all legal systems contain rules of a different type that
provide legal facilities whereby individuals can give effect to
their wishes by entering into certain transactions that alter
their own and/ or others' legal position (rights, duties, status,
etc.). Examples of these civil transactions (acts in the law,
Rechtgeschdfte) are wills, contracts, gifts, marriage. If a legal
system did not provide facilities allowing individuals to give
legal effect to their choices in such areas of conduct, it would
faU to make one of the law's most distinctive and valuable
contributions to social life. But here too most of the mental
conditions we have mentioned are recognized by the law as
important not primarily as excusing conditions but as invali-
dating conditions. Thus a will, a gift, a marriage, and (subject
to many complex exceptions) a contract may be invalid if
the party concerned was insane, mistaken about the legal
character of the transaction or some "essential" term of it, or
if he was subject to duress, coercion, or the undue influence
of other persons. These are the obvious analogues of mistake,
accident, coercion, duress, insanity, admitted by criminal law
as excusing conditions. Analogously, the recognition of such
conditions as invalidating civil transactions is qualified or
limited by other principles. Those who enter in good faith into
bilateral transactions of the kind mentionea with persons who
appear normal (i.e., not subject to any of the relevant invali-

Legal Responsibility and Excuses / 101

dating conditions) must be protected, as must third parties who
may have purchased interests originating from a transaction
that on the face of it seemed normal. Hence a technique has
been introduced to safeguard such persons. This includes
principles precluding, say, a party who has entered into a
transaction by some mistake from making this the basis of
his defense against one who honestly took his words at face
value and justifiably relied on them; there are also distinctions
between transactions wholly invalidated ad initio (void) and
those that are valid until denounced (voidable) to protect
those who have relied on the transaction's normal form.

m

The similarity between the law's insistence on certain
mental elements for both criminal responsibility and the valid-
ity of acts in the law is clear. Why, then, do we value a system
of social control that takes mental condition into account? Let
us start with criminal law and its excusing conditions. What
is so precious in its attention to these, and what would be lost
if it gave this up? What precisely is the ground of our dis-
satisfaction with "strict liability" in criminal law? To these
fundamental questions, there still are, curiously enough, many
quite discordant answers, and I propose to consider two of
them before suggesting an answer that would stress the anal-
ogy with civil transactions.

The first general answer takes this form. It is said that the
importance of excusing conditions in criminal responsibility is
derivative, and it derives from the more fundamental require-
ment that for criminal responsibility there must be "moral
culpability," which would not exist where the excusing condi-
tions are present. On this view the maxim actus non est reus
nisi mens sit rea means a morally evil mind. Certainly traces
of this view are to be found in scattered observations of Eng-
lish and American judges — in phrases such as "an evil mind
with regard to that which he is doing," "a bad mind," "there
must be an act done not merely unguardedly or accidentally,
without an evil mind."^ Some of these well-known formula-
tions were perhaps careless statements of the quite different
principle that mens rea is an intention to commit an act that
is wrong in the sense of legally forbidden. But the same view
has been reasserted in general terms in England by Lord
Justice Denning: "In order that an act should be punishable it

•Lord Esher in hee v. Dangar (1892) 2 Q.B. 337.

102 / Determinism and Freedom

must be morally blameworthy, it must be a sin."^ Most Eng-
lish lawyers would however now agree with Sir James Fitz-
James Stephen that the expression mens rea is unfortunate,
though too firmly established to be expelled, just because it
misleadingly suggests that in general moral culpability is es-
sential to a crime, and they would assent to the criticism ex-
pressed by a later judge that "the true translation of mens rea
is an intention to do the act which is made penal by statute or
common law."^ Yet, in spite of this, the view has been urged
by a distinguished American contemporary writer on criminal
law, Professor Jerome Hall, in his important and illuminating
Principles of Criminal Law, that moral culpability is the basis
of responsibility in crime. Again and again in Chapters V and
VI of his book Professor Hall asserts that, though the good-
ness or badness of the motive with which a crime is com-
mitted may not be relevant, the general principle of liability,
except of course where liability is unfortunately "strict" and
so any mental element must be disregarded, is the "intentional
or reckless doing of a morally wrong act."^ This is declared to
be the essential meaning of mens rea: "though mens rea differs
in different crimes there is one common essential element,
namely, the voluntary doing of a morally wrong act forbidden
by the Iaw."^° On this view the law inquires into the mind in
criminal cases in order to secure that no one shall be punished
in the absence of the basic condition of moral culpability. For
it is just only to "punish those who have intentionally com-
mitted moral wrongs proscribed by law."^^

Now, if this theory were merely a theory as to what the
criminal law of a good society should be, it would not be pos-
sible to refute it, for it represents a moral preference: namely
that legal punishment should be administered only where a
"morally wrong" act has been done — though I think such
plausibility as it would have even as an ideal is due to a con-
fusion. But of com^e Professor Hall's doctrine does not fit any
actual system of criminal law because in every such system
there are necessarily many actions (quite apart from the cases
of "strict liability") that if voluntarily done are criminally

T Denning, The Changing Law (London: Stevens, 1953), p. 12.

»Allard v. Selfrtdgc (1925) 1 K.B. 137. (Shearman.) This is quoted by
Glanville Williams in The Criminal Law (London: Stevens, 1953), p. 29, note
3, where the author comments that the judge should have added "reckless*
ness."

» Hall, op. ctt., p. 166.

i» Ibid., p. 167.

" Ibid., p. 149.

Legal Responsibility and Excuses / 103

punishable, although our moral code may be either silent as
to their moral quality, or divided. Very many offenses are
created by legislation designed to give effect to a particular
economic scheme (e.g., a state monopoly of road or rail trans-
port), the utility or moral character of which may be genu-
inely in dispute. An offender against such legislation can
hardly be said to be morally guilty or to have intentionally
committed a moral wrong, still less "a sin" proscribed by
law;i2 yet jf jjg ^^s broken such laws "voluntarily" (to use
Professor Hall's expression), which in practice means that he
was not in any of the excusing conditions, the requu-ements of
justice are surely satisfied. Doubts about the justice of the pun-
ishment would begin only if he were punished even though he
was at the time of the action in one of the excusing condi-
tions; for what is essential is that the offender, if he is to be
fairly punished, must have acted "voluntarily," and not that
he must have committed some moral offense. In addition to
such requirements of justice in the individual case, there is
of course, as we shall see, a different type of requirement as
to the general character of the laws.

It is important to see what has led Professor Hall and others
to the conclusion that the basis of criminal responsibility must
be moral culpability ("the voluntary doing of a morally wrong
act"), for latent in this position, I think, is a false dilemma.
The false dilemma is that criminal liability must either be
"strict" — that is, based on nothing more than the outward
conduct of the accused — or must be based on moral culpabil-
ity. On this view there is no third alternative and so there can
be no reason for inquiring into the state of mind of the ac-
cused— "inner facts," as Professor Hall terms them — except
for the purpose of establishing moral guilt. To be understood
all theories should be examined in the context of argument in
which they are advanced, and it is important to notice that
Professor Hall's doctrine was developed mainly by way of
criticism of the so-called objective theory of liability, which
was developed, though not very consistently, by Chief Justice
Holmes in his famous essays on common law.^^ Holmes as-

" "The criminal quality of an act cannot be discovered by intuition: nor
can it be discovered by any standard but one. Is the act prohibited with penal
consequences? Morality and criminality are far from coextensive nor is the
sphere of criminality part of a more exclusive field covered by morality unless
morals necessarily disapproves of the acts prohibited by the state, in which
case the argument moves in a circle." Lord Atkin, Proprietory Articles Trade
Association v. A.G. for Canada (1931) A.C. 324.

1* Holmes, The Common Law, Lecture II, "The Criminal Law."

104 / Determinum and Freedom

serted that the law did not consider, and need not consider, in
administering punishment what in fact the accused intended,
but that it imputed to him the intention that an "ordinary
man," equipped with ordinary knowledge, would be taken to
have had in acting as the accused did. Holmes in advocating
this theory of "objective liability" used the phrase "inner
facts" and frequently stressed that mens rea, in the sense of
the actual wickedness of the party, was unnecessary. So he
often identified "mental facts" with moral guilt and also iden-
tified the notion of an objective standard of liability with the re-
jection of moral culpability as a basis of liability. This termi-
nology was pregnant with confusion. It fatally suggests that
there are only two alternatives: to consider the mental condi-
tion of the accused only to find moral culpability or not to
consider it at all. But we are not impaled on the horns of any
such dilemma: there are independent reasons, apart from the
question of moral guilt, why a legal system should require a
voluntary action as a condition of responsibility. These rea-
sons I shall develop in a moment and merely summarize here
by saying that the principle ( 1 ) that it is unfair and unjust to
punish those who have not "voluntarily" broken the law is a
moral principle quite distinct from the assertion (2) that it is
wrong to punish those who have not "voluntarily committed a
moral viTong proscribed by law."

The confusion that suggests the false dilemma — either "ob-
jective" standards (strict liability) or liability based on the
"inner fact" of moral guilt — is, I think, this. We would all
agree that unless a legal system was as a whole morally de-
fensible, so that its existence was better than the chaos of its
collapse, and more good than evil was secured by maintaining
and enforcing laws in general, these laws should not be en-
forced, and no one should be punished for breaking them. It
seems therefore to follow, but does not, that we should not
punish anyone unless in breaking the law he has done some-
thing morally wrong; for it looks as if the mere fact that a
law has been voluntarily broken were not enough to justify
punishment; the extra element required is "moral culpability,"
at least in the sense that we should have done something mor-
ally wrong. What we need to escape confusion here is a dis-
tinction between two sets of questions. The first is a general
question about the moral value of the laws: Will enforcing
them produce more good than evil? If they do, then it is mor-
ally permissible to enforce them by punishing those who have
broken them, unless in any given case there is some "excuse."

Legal Responsibility and Excuses / 105

The second is a particular question concerning individual
cases: Is it right or just to punish this particular person? Is he
to be excused on account of his mental condition because it
would be unjust — in view of his lack of knowledge or control
— to punish him? The first, general question with regard to
each law is a question for the legislature; the second, arising
in particular cases, is for the judge. And the question of re-
sponsibility arises only at the judicial stage. One necessary
condition of the just application of a punishment is normally
expressed by saying that the agent "could have helped" doing
what he did, and hence the need to inquire into the "iimer
facts" is dictated not by the moral principle that only the do-
ing of an immoral act may be legally punished, but by the
moral principle that no one should be punished who could
not help doing what he did. This is a necessary condition (un-
less strict liability is admitted) for the moral propriety of legal
punishment and no doubt also for moral censure; in this
respect law and morals are similar. But this similarity as to
the one essential condition that there must be a "voluntary"
action if legal punishment or moral censure is to be morally
permissible does not mean that legal punishment is morally
permissible only where the agent has done something morally
wrong. I think that the use of the word "fault" in juristic dis-
cussion to designate the requirement that liability be excluded
by excusing conditions may have blurred the important dis-
tinction between the assertions that (1) it is morally permis-
sible to punish only voluntary actions and (2) it is morally
permissible to punish only voluntary commission of a moral
wrong.

IV

Let me now turn to a second explanation of the laws con-
cerned with the "inner facts" of mental life as a condition of
responsibility. This is a Benthamite theory that I shall name
the "economy of threats" and is the contention that the re-
quired conditions of responsibility — e.g., that the agent knew
what he was doing, was not subject to gross coercion or du-
ress, was not mad or a small child — are simply the conditions
that must be satisfied if the threat to punish announced by the
criminal law is to have any effect and if the system is to be
efficient in securing the maintenance of law at the least cost in
pain. This theory is stated most clearly by Bentham; it is also
to be found in Austin and in the report of the great Criminal
Law Commission of 1833 of which he was a member. In a

106 / Determinism and Freedom

refined form it is implicit in many contemporary attempted
"dissolutions" of the problem of free will. Many accept this
view as a common-sense utilitarian explanation of the impor-
tance that we attribute to excusing conditions. It appeals most
to the utilitarian and to the determinist, and it is interesting to
find that Professor Glanville Williams in his recent admirable
work on "The General Principles of Criminal Law,"i* when
he wished to explain the exemption of the insane from legal
responsibility compatibly with "determinism," did so by refer-
ence to this theory.

Yet the doctrine is an incoherent one at certain points, I
think, and a departure from, rather than an elucidation of, the
moral insistence that criminal liability should generally be
conditional on the absence of excusing conditions. Bentham's
best statement of the theory is in Chapter XIII of his Princi-
ples of Morals and Legislation: "Cases in Which Punishment
Must be Inefficacious." The cases he lists, besides those where
the law is made ex post facto or not adequately promulgated,
fall into two main classes. The first class consists of cases in
which the penal threat of punishment could not prevent a
person from performing an action forbidden by the law or
any action of the same sort; these are the cases of infancy and
insanity in which the agent, according to Bentham, has not
the "state or disposition of mind on which the prospect of evils
so distant as those which are held forth by the law" has the
effect of influencing his conduct. The second class consists of
cases in which the law's threat could not have had any effect
on the agent in relation to the particular act committed be-
cause of his lack of knowledge or control. What is wrong in
punishing a man under both these types of mental conditions
is that the punishment is wasteful; suffering is caused to the
accused who is punished in circumstances where it could do
no good.

In discussing the defense of insanity Professor Glanville
Williams applies this theory in a way that brings out its con-
sistency not only with a wholly utilitarian outlook on punish-
ment but with determinism.

For mankind in the mass it is impossible to tell whom the
threat of punishment will restrain and whom it will not; for
most it will succeed, for some it will fail. And the punishment
must then be applied to those criminals in order to maintain the
threat to persons generally. Mentally deranged persons, how-

1* Williams, op. cit., pp. 346-47.

Legal Responsibility and Excuses / 107

ever, can be separated from the mass by scientific tests, and
being a defined class their segregation from punishment does not
impair the eflBcacy of the sanction for people generally.^*

The point made here is that, if, for example, the mentally de-
ranged (scientifically tested) are exempted, criminals will not
be able to exploit this exemption to free themselves from
liability, since they cannot bring themselves within its scope
and so will not feel free to commit crimes with impunity. This
is said in order to justify the exemption of the insane con-
sistently with the "tenet" of determinism, in spite of the fact
that from a determinist viewpoint

every impulse if not in fact resisted was in those circumstances
irresistible. A so-called irresistible impulse is simply one in
which the desire to perform a particular act is not influenced by
other factors like the threat of punishment. ... on this definition
every crime is the result of an irresistible impulse.

This theory is designed not merely to fit a utilitarian theory
of punishment, but also the view that it is always false, if not
senseless, to say that a criminal could have helped doing what
he did. So on this theory when we inquire into the mental
state of the accused, we do not do so to answer the question,
Could he help it? Nor of course to answer the question. Could
the threat of punishment have been effective in his case? — for
we know that it was not. The theory presents us with a far
simpler conceptual scheme for dealing with the whole matter,
since it does not involve the seemingly counterfactual specu-
lation regarding what the accused "could have done." On this
theory we inquire into the state of mind of the accused simply
to find out whether he belongs to a defined class of persons
whose exemption from punishment, if allowed, will not
weaken the effect on others of the general threat of punish-
ment made by the law. So there is no question of its being
unjust or unfair to punish a particular criminal or to exempt
him from punishment. Once the crime has been committed
the decision to punish or not has nothing to do with any moral
claim or right of the criminal to have the features of his case
considered, but only with the causal efficacy of his punishment
on others. On this view the rationale of excuses is not (to put
it shortly) that the accused should in view of his mental con-
dition be excused whatever the effect of this on others, but
rather the mere fact that excusing him will not harm society

16 Williams, loc. clt.

108 / Determinism and Freedom

by reducing the efficacy of the law's threats for others. So
the relevance of the criminal's mental condition is purely the
question of the effect on others of his punishment or exemp-
tion from it.

This is certainly paradoxical enough. It seems to destroy
the entire notion that in punishing we must be just to the
particular criminal in front of us and that the purpose of
excusing conditions is to protect him from society's claim.
But apart from paradox the doctrine that we consider the
state of a man's mind only to see if punishment is required in
order to maintain the efficacy of threats for others is vitiated
by a non sequitur. Before a man does a criminal action we
may know that he is in such a condition that the threats can-
not operate on him, either because of some temporary con-
dition or because of a disease; but it does not follow — because
the threat of punishment in his case, and in the case of others
like him, is useless — that his punishment in the sense of the
official administration of penalties will also be unnecessary
to maintain the efficacy of threats for others at its highest. It
may very well be that, if the law contained no explicit exemp-
tions from responsibility on the score of ignorance, accident,
mistake, or insanity, many people who now take a chance in
the hope that they will bring themselves, if discovered, within
these exempting provisions would in fact be deterred. It is
indeed a perfectly familiar fact that pleas of loss of con-
sciousness or other abnormal mental states, or of the existence
of some other excusing condition, are frequently and some-
times successfully advanced where there is no real basis for
them, for the difficulties of disproof are often considerable.
The uselessness of a threat against a given individual or class
does not entail that the punishment of that individual or class
cannot be required to maintain in the highest degree the ef-
ficacy of threats for others. It may in fact be the case that to
make liability to punishment dependent on the absence of ex-
cusing conditions is the most efficient way of maintaining the
laws with the least cost in pain. But it is not obviously or
necessarily the case.

It is clear, I think, that if we were to base our views of
criminal resp)Onsibility on the doctrine of the economy of
threats, we should misrepresent altogether the character of
our moral preference for a legal system that requires mental
conditions of responsibility over a system of total strict liabil-
ity or entirely different methods of social control such as
hypnosis, propaganda, or conditioning.

Legal Responsibility and Excuses / 109

(^To make this intelligible we must cease to regard the law as
merely a causal factor in human behavior differing from
others only in the fact that it produces its effect through the
medium of the mind; for it is clear that we look on excusing
conditions as something that protects the individual against
the claims of the rest of society. Recognition of their excusing
force may lead to a lower, not a higher, level of efficacy of
threats; yet — and this is the point — we could not regard that
as sufficient ground for abandoning this protection of the
individual; or if we did, it would be with the recognition that
we had sacrificed one principle to another; for more is at
stake than the single principle of maintaining the laws at their
most efficacious level. We must cease, therefore, to regard the
law simply as a system of stimuli goading the individual by its
threats into conformity. Instead I shall suggest a mercantile
analogy. Consider the law not as a system of stimuli but as
what might be termed a choosing system in which individuals
can find out, in general terms at least, the costs they have to
pay if they act in certain ways. This done, let us ask what
value this system would have in social life and why we should
regret its absence. I do not of course mean to suggest that it
is a matter of indifference whether we obey the law or break
it and pay the penalty. Punishment is different from a mere
"tax on a course of conduct." What I do mean is that the
conception of the law simply as goading individuals into de-
sired courses of behavior is inadequate and misleading; what
a legal system that makes liability generally depend on ex-
cusing conditions does is to guide individuals' choices as to
behavior by presenting them with reasons for exercising
choice in the direction of obedience, but leaving them to
choose.

It is at this point that I would stress the analogy between
the mental conditions that excuse from criminal responsibility
and the mental conditions that are regarded as invalidating
civil transactions such as wills, gifts, contracts, marriages, and
the like. The latter institutions provide individuals with two
inestimable advantages in relation to those areas of conduct
they cover. These are (1) the advantage to the individual of
determining by his choice what the future shall be and (2)
the advantage of being able to predict what -the future wUl
be. For these institutions enable the individual ( 1 ) to bring
into operation the coercive forces of the law so that those
legal arrangements he has chosen shall be carried into effect
and (2) to plan the rest of his life with a certainty or at least

110 / Determinism and Freedom

the confidence (in a legal system that is working normally)
that the arrangements he has made will in fact be carried out.
By these devices the individual's choice is brought into the
legal system and allowed to determine its future operations in
certain areas, thereby giving him a type of indirect coercive
control over, and a power to foresee the development of,
official life. This he would not have "naturally"; that is, apart
from these legal institutions.

In brief, the function of these institutions of private law is
to render effective the individual's preferences in certain
areas. It is therefore clear why in this sphere the law treats
the mental factors of, say, mistake, ignorance of the nature
of the transaction, coercion, undue influence, or insanity as
invalidating such civil transactions. For a transaction entered
into under such conditions will not represent a real choice:
the individual might have chosen one course of events and
by the transaction procured another (cases of mistake, ig-
norance, etc.), or he might have chosen to enter the trans-
action without coolly and calmly thinking out what he wanted
(undue influence), or he might have been subjected to the
threats of another who had imposed his choices (coercion).

To see the value of such institutions in rendering effective
the individual's considered and informed choices as to what
on the whole shall happen, we have but to conduct the experi-
ment of imagining their absence: a system where no mental
conditions would be recognized as invalidating such trans-
actions and the consequent loss of control over the future
that the individual would suffer. That such institutions do
render individual choices effective and increase the powers of
individuals to predict the course of events is simply a matter
of empirical fact, and no form of "determinism," of course,
can show this to be false or illusory. If a man makes a will
to which the law gives effect after his death, this is not, of
course, merely a case of post hoc: we have enough empirical
evidence to show that this was an instance of a regularity
sufficient to have enabled us to predict the outcome with rea-
sonable probability, at least in some cases, and to justify us,
therefore, in interpreting this outcome as a consequence of
making the will. There is no reason why we should not de-
scribe the situation as one where the testator caused the out-
come of the distribution made. Of course the testator's choice
in his example is only one prominent member of a complex
set of conditions, of which all the other members were as
necessary for the production of the outcome as his choice.

Legal Responsibility and Excuses / 111

Science may indeed show ( 1 ) that this set of conditions also
includes conditions of which we are at the present moment
quite ignorant and (2) that the testator's choice itself was the
outcome of some set of jointly sufficient conditions of which
we have no present knowledge. Yet neither of these two sup-
positions, even if they were verified, would make it false to
say that the individual's choice did determine the results, or
make illusory the satisfaction got (a) from the knowledge that
this kind of thing is possible, (b) from the exercise of such
choice. And if determinism does not entail that satisfactions
(o) or (b) are obtainable, I for one do not understand how it
could affect the wisdom, justice, rationality, or moraUty of the
system we are considering.

If with this in mind we turn back to criminal law and its
excusing conditions, we can regard their function as a mech-
anism for similarly maximizing within the framework of
coercive criminal law the efficacy of the individual's informed
and considered choice in determining the future and also his
power to predict that future. We must start, of course, with
the need for criminal law and its sanctions as at least some
check on behavior that threatens society. This implies a belief
that the criminal law's threats actually do diminish the fre-
quency of antisocial behavior, and no doubt this belief may
be said to be based on inadequate evidence. However, we
must clearly take it as our starting point: if this belief is
wrong, it is so because of lack of empirical evidence and not
because it contradicts any form of determinism. Then we can
see that by attaching excusing conditions to criminal responsi-
bility, we provide each individual with something he would
not have if we made the system of criminal law operate on
a basis of total "strict liability." First, we maximize the in-
dividual's power at any time to predict the likelihood that the
sanctions of the criminal law will be applied to him. Secondly,
we introduce the individual's choice as one of the operative
factors determining whether or not these sanctions shall be
applied to him. He can weigh the cost to him of obeying the
law — and of sacrificing some satisfaction in order to obey —
against obtaining that satisfaction at the cost of paying "the
penalty." Thirdly, by adopting this system of attaching ex-
cusing conditions we provide that, if the sanctions of the
criminal law are applied, the pains of punishment will for
each individual represent the price of some satisfaction ob-
tained from breach of law. Ths, of course, can sound like a
very cold, if not immoral, attitude toward the criminal law.

112 / Determinism and Freedom

general obedience to which we regard as an essential part of
a decent social order. But this attitude seems repellent only
if we assume that all criminal laws are ones whose operation
we approve. To be realistic we must also think of bad and
repressive criminal laws; in South Africa, Nazi Germany,
Soviet Russia, and no doubt elsewhere, we might be thankful
to have their badness mitigated by the fact that they fall only
on those who have obtained a satisfaction from knowingly do-
ing what they forbid.

Again, the value of these three factors can be realized if
we conduct the Gedankenexperiment of imagining criminal
law operating without excusing conditions. First, our power of
predicting what will happen to us wiU be immeasurably di-
minished; the likelihood that I shall choose to do the for-
bidden act (e.g., strike someone) and so incur the sanctions
of the criminal law may not be very easy to calculate even
under our system: as a basis for this prediction we have in-
deed only the knowledge of our own character and some
estimate of the temptations life is likely to offer us. But if
we are also to be liable if we strike someone by accident, by
mistake, under coercion, etc., the chance that we shall incur
the sanctions are immeasurably increased. From our knowl-
edge of the past career of our body considered as a thing, we
cannot infer much as to the chances of its being brought into
violent contact with another, and under a system that dis-
pensed with the excusing condition of, say, accident (implying
lack of intention), a collision alone would land us in jail.
Secondly, our choice would condition what befalls us to a
lesser extent. Thirdly, we should suffer sanctions without
having obtained any satisfaction. Again, no form of determin-
ism that I, at least, can construct can throw any doubt on, or
show to be illusory, the real satisfaction that a system of
criminal law incorporating excusing conditions provides for
individuals in maximizing the effect of their choices within the
framework of coercive law. The choices remain choices, the
satisfactions remain satisfactions, and the consequences of
choices remain the consequences of choices even if choices
are determined and other "determinants" besides our choices
condition the satisfaction arising from their being rendered
effective in this way by the structure of the criminal law.

It is now important to contrast this view of excusing con-
ditions with the Benthamite explanation I discussed in Part
III of this paper. On that view excusing conditions were
treated as conditions imder which the law's threat could op-

Legal ResponsibUity and Excuses / 113

erate with maximum efficacy. They were recognized not be-
cause they ensured justice to individuals considered separately,
but because sanctions administered under those conditions
were believed more effective and economical of pain in secur-
ing the general conformity to law. If these beliefs as to the
efficacy of excusing conditions could be shown false, then all
reasons for recognizing them as conditions of criminal re-
sponsibility would disappear. On the present view, which I
advocate, excusing conditions are accepted as independent of
the efficacy of the system of threats. Instead it is conceded
that recognition of these conditions may, and probably does,
diminish that efficacy by increasing the number of conditions
for criminal liability and hence giving opportunities for pre-
tense on the part of criminals, or mistakes on the part of
tribunals.

On this view excusing conditions are accepted as something
that may conflict with the social utility of the law's threats;
they are regarded as of moral importance because they pro-
vide for all individuals alike the satisfactions of a costing
system. Recognition of excusing conditions is therefore seen
as a matter of protection of the individual against the claims
of society for the highest measure of protection from crime
that can be obtained from a system of threats. In this way
the criminal law respects the claims of the individual as such,
or at least as a choosing being, and distributes its coercive
sanctions in a way that reflects this respect for the individual.
This surely is very central in the notion of justice and is one,
though no doubt only one, among the many strands of prin-
ciple that I think lie at the root of the preference for legal
institutions conditioning liability by reference to excusing
conditions.

I cannot, of course, by unearthing this principle claim to
have solved everyone's perplexities. In particular, I do not
know what to say to a critic who urges that I have shown only
that the system in which excusing conditions are recognized
protects the individual better against the claims of society
than one in which no recognition is accorded to these factors.
This seems to me to be enough; yet I cannot satisfy his com-
plaint, if he makes it, that I have not shown that we are
justified in punishing anyone ever, at all, under any condi-
tions. He may say that even the criminal who has committed
his crime in the most deliberate and calculating way and has
shown himself throughout his life competent in maximizing
what he thinks his own interests will be little comforted when

114 / Determinism and Freedom

he is caught and punished for some major crime. At that stage
he will get little satisfaction if it is pointed out to him ( 1 ) that
he has obtained some satisfaction from his crime, (2) that
he knew that it was likely he would be punished and that he
had decided to pay for his satisfaction by exposing himself to
this risk, and (3) that the system imder which he is punished
is not one of strict liability, is not one under which a man
who accidentally did what he did would also have suffered
the penalties of the law.

I will add four observations ex abundante cautela.

1. The elucidation of the moral importance of the mental
element in responsibihty, and the moral odium of strict lia-
bility that I have indicated, must not be mistaken for a psy-
chological theory of motivation. It does not answer the
question, Why do people obey the law? It does not assert that
they obey only because they choose to obey rather than pay
the cost. Instead, my theory answers the question, Why should
we have a law with just these features? Human beings in the
main do what the law requires without first choosing between
the advantage and the cost of disobeying, and when they obey
it is not usually from fear of the sanction. For most the
sanction is important not because it inspires them with fear
but because it offers a guarantee that the antisocial minority
who would not otherwise obey will be coerced into obedience
by fear. To obey without this assurance might, as Hobbes saw,
be very foolish: it would be to risk going to the wall. How-
ever, the fact that only a few people, as things are, consider
the question, Shall I /obey or pay? does not in the least mean
that the standing pdssibility of asking this question is unim-
portant: for it secures just those values for the individual that
I have mentioned.

2. I must of course confront the objection the Marxist
might make, that the excusing conditions, or indeed mutatis
mutandis the invalidating conditions, of civil transactions are
of no use to many individuals in society whose economic or
social position is such that the difference between a law of
strict liability and a law that recognizes excusing conditions is
of no importance.

It is quite true that the fact that criminal law recognizes
excusing mental conditions may be of no importance to a
person whose economic condition is such that he cannot
profit from the difference between a law against theft that is

Legal Responsibility and Excuses / 115

strict and one that incorporates excusing conditions. If starva-
tion "forces" him to steal, the values the system respects and
incorporates in excusing conditions are nothing to him. This
is of course similar to the claim often made that the freedom
that a political democracy of the Western type offers to its
subjects is merely formal freedom, not real freedom, and
leaves one free to starve. I regard this as a confusing way of
putting what may be true under certain conditions: namely,
that the freedoms the law offers may be valueless as playing
no part in the happiness of persons who are too poor or weak
to take advantage of them. The admission that the excusing
conditions may be of no value to those who are below a
minimum level of economic prosperity may mean, of course,
that we should incorporate as a further excusing condition the
pressure of gross forms of economic necessity. This point,
though valid, does not seem to me to throw doubt on the
principle lying behind such excusing conditions as we do
recognize at present, nor to destroy their genuine value for
those who are above the minimum level of economic pros-
perity, for a difference between a system of strict liability and
our present system plays a part in their happiness.

3. The principle by reference to which I have explained the
moral importance of excusing conditions may help clarify an
old dispute, apt to spring up between lawyers on the one hand
and doctors and scientists on the other, about the moral basis
of punishment.

From Plato to the present day there has been a recurrent
insistence that if we were rational we would always look on
crime as a disease and address ourselves to its cure. We would
do this not only where a crime has actually been committed
but where we find well-marked evidence that it will be. We
would take the individual and treat him as a patient before
the deed was done. Plato, ^^ it will be remembered, thought it
superstitious to look back and go into questions of responsi-
bility or the previous history of a crime except when it might
throw light on what was needed to cure the criminal.

Carried to its extreme, this doctrine is the program of
Erewhon where those with criminal tendencies were sent by
doctors for indefinite periods of cure; punishment was dis-
placed by a concept of social hygiene. It is, I think, of some
importance to realize why we should object to this point of
view, for both those who defend it and those who attack it

i» Plato, Protagoras, 324; Laws 861, 865.

116 / Determinism and Freedom

often assume that the only possible consistent alternative to
Erewhon is a theory of punishment under which it is justified
simply as a return for the moral evil attributable to the ac-
cused. Those opposed to the Erewhonian program are apt to
object that it disregards moral guilt as a necessary condition
of a just punishment and thus leads to a condition in which
any person may be sacrificed to the welfare of society. Those
who defend an Erewhonian view think that their opponents'
objection must entail adherence to the form of retributive
punishment that regards punishment as a justified return for
the moral evil in the criminal's action.

Both sides, I think, make a common mistake: there is a
reason for making punishment conditional on the commission
of crime and respecting excusing conditions, which are quite
independent of the form of retributive theory that is often
urged as the only alternative to Erewhon. Even if we regard
the over-all purpose of punishment as that of protecting
society by deterring persons from committing crimes and in-
sist that the penalties we inflict be adapted to this end, we can
in perfect consistency and with good reason insist that these
punishments be applied only to those who have broken a law
and to whom no excusing conditions apply. For this system
will provide a measure of protection to individuals and will
maximize their powers of prediction and the efficacy of their
choices in the way that I have mentioned. To see this we have
only to ask ourselves what in terms of these values we should
lose (however much else we might gain) if social hygiene
and a system of compulsory treatment for those with detecta-
ble criminal tendencies were throughout substituted for our
system of punishment modified by excusing conditions. Surely
the realization of what would be lost, and not a retributive
theory of punishment, is all that is required as a reason for
refusing to make the descent into Erewhon.

4. Finally, what I have written concerns only legal re-
sponsibility and the rationale of excuses in a legal system in
which there are organized, coercive sanctions. I do not think
the same arguments can be used to defend moral responsi-
bility from the determinist, if it is in any danger from that
source.

Chapter 2

Paul Edwards, New York University

In ffls ESSAY "The Dilemma of Determinism," William James
makes a distinction that will serve as a point of departure for
my remarks. He there distinguishes between the philosophers
he calls "hard" determinists and those he labels "soft" deter-
minists. The former, the hard determinists, James tells us,
"did not shrink from such words as fatality, bondage of the
will, necessitation and the like." He quotes a famous stanza
from Omar Khayyam as representing this kind of determin-
ism:

With earth's first clay they did the last man knead.
And there of the last harvest sowed the seed.
And the first morning of creation wrote
What the last dawn of reckoning shall read.

Another of Omar's verses expresses perhaps even better the
kind of theory that James has here in mind:

Tis all a checker-board of nights and days,
Where destiny with men for pieces plays;
Thither and thither moves, and metes, and slays.
And one by one back to the closet lays.

James mentioned no names other than Omar Khayydm. But
there is little doubt that among the hard determinists he
would have included Jonathan Edwards, Anthony Collins,
Holbach, Priestley, Robert Owen, Schopenhauer, Freud, and
also, if he had come a little earlier, Clarence Darrow.

James of course rejected both hard and soft determinism,
but for hard determinism he had a certain respect: the kind of
respect one sometimes has for an honest, straightforward ad-
versary. For soft determinism, on the other hand, he had
nothing but contempt, calling it a "quagmire of evasion."
"Nowadays," he writes, "we have a soft determinism which
abhors harsh words, and repudiating fatality, necessity, and

117

118 / Determinism and Freedom

even predetermination, says that its real name is 'freedom.' "
From his subsequent observations it is clear that he would in-
clude among the evasionists not only neo-Hegelians like
Green and Bradley but also Hobbes and Hume and Mill; and
if he were alive today James would undoubtedly include
Schlick and Ayer and Stevenson and Noel-Smith, not to men-
tion some of the philosophers present in this room.

The theory James calls soft determinism, especially the
Hume-Mill-Schlick variety of it, has been extremely fashion-
able during the last twenty-five years, while hardly anybody
can be found today who has anj^hing good to say for hard
determinism. In opposition to this contemporary trend, I
should like to strike a blow on behalf of hard determinism in
my talk today, I shall also try to bring out exactly what is
really at issue between hard and soft determinism. I think
the nature of this dispute has frequently been misconceived
chiefly because many writers, including James, have a very
inaccurate notion of what is maintained by actual hard deter-
minists, as distinct from the bogey men they set up in order
to score an easy victory.

To begin witih, it is necessary to spell more fully the main
contentions of the soft determinists. Since it is the dominant
form of soft determinism at the present time, I shall confine
myself to the Hume-Mill-Schlick theory. According to this
theory there is in the first place no contradiction whatsoever
between determinism and the proposition that human beings
are sometimes free agents. When we call an action "free" we
never in any ordinary situation mean that it was uncaused;
and this emphatically includes the kind of action about which
we pass moral judgments. By calling an action "free" we mean
; that the agent was not compelled or constained to perform it.
Sometimes people act in a certain way because of threats or
because they have been drugged or because of a posthypnotic
suggestion or because of an irrational overpowering urge such
as the one that makes a kleptomaniac steal something he does
not really need. On such occasions human beings are not free
agents. But on other occasions they act in certain ways be-
cause of their own rational desires, because of their own un-
impeded efforts, because they have chosen to act in these
ways. On these occasions they are free agents although their
actions are just as much caused as actions that are not deemed
free. In distinguishing between free and unfree actions we
do not try to mark the presence and absence of causes but
attempt to indicate the kind of causes that are present.

Hard and Soft Detenmnism / 119

* Secondly there is no antithesis between determinism and
moral responsibility. When we judge a person morally respon-
sible for a certain action, we do indeed presuppose that he
was a free agent at the time of the action. But the freedom
presupposed is not the contracausal freedom about which in-
determinists go into such ecstatic raptures. It is nothing more
than the freedom already mentioned- — the ability to act ac-
cording to one's choices or desires. Since determinism is com-
patible with freedom in this sense, it is also compatible with
moral responsibility. In other words, the world is after all
wonderful: we can be determinists and yet go on punishing
our enemies and owe children, and we can go on blaming our-
selves, all without a bad intellectual conscience.

\f Mill, who was probably the greatest moralizer among the
soft determinists, recognized with particular satisfaction the
influence or alleged influence of one class of human desires.
Not^nly, for example, does such lowly desire as my desire to
get a new car influence my conduct. It is equally true, or so
at least Mill believed, that my desire to become a more vir-
tuous person does on occasion influence my actions. By
suitable training and efforts my desire to change my char-
acter may in fact bring about the desired changes. If Mill were
alive today he might point to contemporary psychiatry as an
illustration of his point. Let us suppose that I have an intense
desire to become famous, but that I also have an intense de-
sire to become a happier and more lovable person who,
among other things, does not greatly care about fame. Let us
suppose, furthermore, that I know of a therapy that can
transform fame-seeking and unlovable into lovable and fame-
indifferent character structures. If, now, I have enough
money, energy, and courage, and if a few other conditions
are fulfilled, my desire may actually lead to a major change
in my character. Since we can, therefore, at least to some
extent, form our own character, determinism according to.
Millis_cprapatible not only with judgments of moral responsi-^'
bility about this or that particular action flowing from an
unimpeded desire, but also, within limits, with moral judg-
ments about the character of human beings.

I think that several of MUl's observations were well worth
making and that James's verdict on his theory as a "quagmire
of evasion" is far too derogatory. I think hard determinists
have occasionally written in such a way as to suggest that
they deny the causal efficacy of human desires and efforts.
Thus Holbach wrote:

120 / Determinism and Freedom

You will say that I feel free. This is an illusion, which may be
compared to that of the fly in the fable, who, lighting upon the
pole of a heavy carriage, applauded himself for directing its
course. Man, who thinks himself free, is a fly who imagines he
has power to move the universe, while he is himself unknow-
ingly carried along by it

There is also the following passage in Schopenhauer:

Every man, being what he is and placed in the circumstances
which for the moment obtain, but which on their part also arise
by strict necessity, can absolutely never do anything else than
just what at that moment he does do. Accordingly, the whole
course of a man's life, in all its incidents great and small, is as
necessarily predetermined as the course of a clock.

Voltaire expresses himself in much the same way in the article
on "Destiny" in the Philosophical Dictionary.

Everything happens through immutable laws, . . . everj^thing
is necessary. . . , "There are," some persons say, "some events
which are necessary and others which are not." It would be very
comic that one part of the world was arranged, and the other
were not; that one part of what happens had to happen and that
another part of what happens did not have to happen. If one
looks closely at it, one sees that the doctrine contrary to that of
destiny is absurd; but there are many people destined to reason
badly; others not to reason at all, others to persecute those who
reason. . . .

... I necessarily have the passion for writing this, and you
have the passion for condemning me; both of us are equally
fools, equally the toy of destiny. Your nature is to do harm,
mine is to love truth, and to make it public in spite of you.

Furthermore there can be little doubt that Hume and Mill
and Schlick were a great deal clearer about the relation be-
tween motives and actions than the hard determinists, who
either conceived it, like Collins, as one of logical necessity or,
like Priestley and Voltaire and Schopenhauer, as necessarily
involving coercion or constraint.

But when all is said and done, there remains a good deal of
truth in James's charge that soft determinism is an evasion.
For a careful reading of their works shows that none of the
hard determinists really denied that human desires, efforts,
and choices make a difference in the course of events. Any
remarks to the contrary are at most temporary lapses. This,
then, is hardly the point at issue. If it is not the point at issue.

Hard and Soft Determinism / 121

what is? Let me at this stage imagine a hard determinist re-
plying to a champion of the Hume-MUl theory: "You are
right," he would say, "in maintaining that some of our actions
are caused by our desires and choices. But you do not pursue
the subject far enough. You arbitrarily stop at the desires and t
volitions. We must not stop there. We must go on to ask '
where they come from; and if determinism is true there can i
be no doubt about the answer to this question. Ultimately our
desires and our whole character are derived from our in-
herited equipment and the environmental influences to which
we were subjected at the beginning of our lives. It is clear that
we had no hand in shaping either of these." A hard determin-
ist could quote a number of eminent supporters. "Our voli-
tions and our desires," wrote Holbach in his little book Good
Sense, "are never in our power. You think yourself free, be-
cause you do what you will; but are you free to will or not to
will; to desire or not to desire?" And Schopenhauer expressed
the same thought in the following epigram: "A man can
surely do what he willsto do, but he cannot determine what

LeTme turn once more to the topic of character transfor-
mation by means of psychiatry to bring out this point with
full force. Let us suppose that both A and B are compulsive
and suffer intensely from their neuroses. Let us assume that
there is a therapy that could help them, which could materially
change their character structure, but that it takes a great deal
of energy and courage to undertake the treatment. Let us sup-
pose that A has the necessary energy and courage while B
lacks it. A undergoes the therapy and changes in the desired
way. B just gets more and more compulsive and more and
more miserable. Now, it is true that A helped form his own .
later character. But his starting point, his desire to change, his'j
energy and courage, were already there. They may or may not
have been the result of previous efforts on his own part. But
there must have been a first effort, and the effort at that time
was the result of factors that were not of his making.

The fact that^ person's character is ultimately the product
of factors over which he had no control 4s not denied by the
soft determinists, though many of them don't like to be re-
minded of it when they are in a moralizing mood. Since the
hard determinists admit that our desires and choices do on
occasion influence the course of our lives, there is thus no
disagreement between the soft and the hard determinists about
the empurical facts. However, some hard determinists infer

122 / Determinum and Freedom

from some of these facts that human beings are never morally
responsible for their actions. The soft determinists, as already
stated, do not draw any such inference. In the remainder of
my paper I shall try to show just what it is that hard determin-
ists are inferring and why, in my opinion, they are justified in
their conclusion.

I shall begin by adopting for my purposes a distinction intro-
duced by C. A. Campbell in his extremely valuable article "Is
Free Will a Pseudo-Problem?"^ in which he distinguishes be-
tween two conceptions of moral responsibility. Different per-
sons, he says, require different conditions to be fulfilled before
holding human beings morally responsible for what they do.
First, there is what Campbell calls the ordinary unreflective
person, who is rather ignorant and who is not greatly con-
cerned with the theories of science, philosophy, and religion.
If the unreflective person is sure that the agent to be judged
was acting under coercion or constraint, he will not hold him
responsible. If, however, he is sure that the action was per-
formed in accordance with the agent's unimpeded rational
desire, if he is sure that the action would not have taken place
but for the agent's decision, then the unreflective person will
consider ascription of moral responsibility justified. The fact
that the agent did not ultimately make his own character will
either not occur to him, or else it will not be considered a
sufficient ground for withholding a judgment of moral
responsibility.

In addition to such unreflective persons, continues Camp-
bell, there are others who have reached "a tolerably advanced
level or reflection."

Such a person will doubtless be acquainted with the claims
advanced in some quarters that causal law operates universally;
or/and with the theories of some philosophies that the universe
is throughout the expression of a single supreme principle; or/
and with the doctrines of some theologians that the world is
created, sustained and governed by an Omniscient and Omni-
potent Being.

Such a person will tend to require the fulfillment of a further
condition before holding anybody morally responsible. He will
require not only that the agent was not coerced or constrained
but also — and this is taken to be an additional condition — that
he "could have chosen otherwise than he actually did." I

^Mtnd, 1951.

Hard and Soft Determinism / 123

should prefer to put this somewhat differently, but it will not
affect the main conclusion drawn by Campbell, with which I
agree. The reflective person, I should prefer to express it, re-
quires not only that the agent was not coerced; he also re-
quires that the agent originally chose his own character — the
character that now displays itself in his choices and desires
and efforts. Campbell concludes that determinism is indeed
compatible with judgments of moral responsibility in the un-
reflective sense, but that it is incompatible with judgments of
moral responsilsility in the reflective sense.

Although I do not follow Campbell in rejecting determin-
ism, I agree basically with his analysis, with one other quali-
fication. I do not think it is a question of the different senses in
which the term is used by ignorant and unreflective people, on
the one hand, and by those who are interested in science,
religion, and philosophy, on the other. The very same persons,
whether educated or uneducated, use it in certain contexts in
the one sense and in other contexts in the other. Practically
all human beings, no matter how much interested they are in
science, religion, and philosophy, employ what Campbell calls
the unreflective conception when they are dominated by vio-
lent emotions like anger, indignation, or hate, and especially
when the conduct they are judging has been personally injuri-
ous to them. On the other hand, a great many people, whether
they are educated or not, will employ what Campbell calls the
reflective conception when they are not consumed with hate
or anger — when they are judging a situation calmly and re-
flectively and when the fact that the agent did not ultimately
shape his own character has been vividly brought to their at-
tention. Clarence Darrow in his celebrated pleas repeatedly
appealed to the jury on precisely this ground. If any of you,
he would say, had been reared in an environment like that of
the accused or had to suffer from his defective heredity, you
would now be standing in the dock. I cannot refrain at this
stage from reading a poem written by the hard determinist,
A. E. Housman, which Darrow recited on such occasions. Its
title is "The Culprit," and it is the soliloquy of a boy about to
be hanged.

The night my father got me
His mind was not on me;

He did not plague his fancy
To muse if I should be
The son you see.

124 / Determinum and Freedom

The day my mother bore me

She was a fool and glad.
For all the pain I cost her.

That she had borne the lad

That borne she had.

My mother and my father

Out of the light they lie;
The warrant could not find them.

And here 'tis only I

Shall hang so high.

Oh let not man remember

The soul that God forgot.
But fetch the county kerchief

And noose me in the knot.

And I will rot.

For so the game is ended

That should not have begun.
My father and my mother

They had a likely son,

And I have none.2

Darrow nearly always convinced the jury that the accused
could not be held morally responsible for his acts; and cer-
tainly the majority of the jurors were relatively uneducated.

I have so far merely distinguished between two concepts of
moral responsibility. I now wish to go a step farther and claim
that only one of them can be considered, properly speaking,
a moral concept. This is not an easy point to make clear, but
I can at least indicate what I mean. We do not normally con-
sider just any positive or negative feeling a "moral" emotion.
Nor do we consider just any sentence containing the words
"good" or "bad" expressions of "moral" judgment. For ex-
ample, if a man hates a woman because she rejected him, this
would not be counted as a moral emotion. If, however, he dis-
approves, say, of Senator McCarthy's libelous speech against
Adlai Stevenson before the 1952 election because he disap-
proves of slander in general and not merely because he likes
Stevenson and dislikes McCarthy, his feeling would be
counted as moral. A feeling or judgment must in a certain
sense be "impersonal" before we consider it moral. To this I

»From The Collected Poems oi A. E. Housman. Copyright, 1922, 1940 by
Henry Holt and Company, Inc. Copyright, 1950, by Barclays Bank, Ltd. By
permission of the publishers.

Hard and Soft Determinism / 125

would add that it must also be independent of violent emo-
tions.') Confining myself to judgments, I would say that a
judgment was "moral" only if it was formulated in a calm and
reflective mood, or at least if it is supported in a calm and
reflective state of mind. If this is so, it follows that what
Campbell calls the reflective sense of "moral responsibiUty" is
the only one that qualifies as a properly moral use of the term.
Before I conclude I wish to avoid a certain misunderstand-
ing of my remarks. From the fact that human beings do not
ultimately shape their own character, I said, it follows that
they are never morally responsible. I do not mean that by
reminding people of the ultimate causes of their character one
makes them more charitable and less vengeful. Maybe one
does, but that is not what I mean. I mean "follow" or "imply"
in the same sense as, or in a sense closely akin to, that in
which the conclusion of a valid syllogism follows from the
premises. The effectiveness of Darrow's pleas does not merely
show, I am arguing, how powerfully he could sway the emo-
tions of the jurors. His pleas also brought into the open one of
the conditions the jurors, like others, consider necessary on
reflection before they hold an agent morally responsible. Or
perhaps I should say that Darrow committed the jurors in
their reflective nature to a certain ground for the ascription of
moral responsibility.^

» Author's Note. This paper was written in the hope of stimulating dis-
cussion of a position which has not received adequate attention in recent
years. The position was stated rather bluntly and without the necessary quali-
fications because of limitations of time. I hope to return to the subject at
greater length In the near future, and on that occasion to present a more
balanced treatment which wiU attempt to meet criticisms made in the dis-
cussion. {December 1957.)

Chapter 3

What M@aii§ This Freedom?

John Hospers, Brooklyn College

I AM IN AGREEMENT to a vcry large extent with the conclu-
sions of Professor Edwards' paper, and am happy in these
days of "soft determinism" to hear the other view so force-
fully and fearlessly stated. As a preparation for developing my
own views on the subject, I want to mention a factor that I
think is of enormous importance and relevance: namely, un-
conscious motivation. There are many actions — not those of
an insane person (however the term "insane" be defined), nor
of a person ignorant of the effects of his action, nor ignorant
of some relevant fact about the situation, nor in any obvious
way mentally deranged — for which human beings in general
and the courts in particular are inclined to hold the doer re-
sponsible, and for which, I would say, he should not be held
responsible. The deed may be planned, it may be carried out
in cold calculation, it may spring from the agent's character
and be continuous with the rest of his behavior, and it may be
perfectly true that he could have done differently // he had
wanted to; nonetheless his behavior was brought about by
unconscious conflicts developed in infancy, over which he had
no control and of which (without training in psychiatry) he
does not even have knowledge. He may even think he knows
why he acted as he did, he may think he has conscious control
over his actions, he may even think he is fully responsible for
them; but he is not. Psychiatric casebooks provide hundreds of
examples. The law and common sense, though puzzled some-
times by such cases, are gradually becoming aware that they
exist; but at this early stage countless tragic blunders still
occur because neither the law nor the public in general is
aware of the genesis of criminal actions. The mother blames
her daughter for choosing the wrong men as candidates for
husbands; but though the daughter thinks she is choosing
freely and spends a considerable amount of time "deciding"
among them, the identification with her sick father, resulting
from Oedipal fantasies in early childhood, prevents her from
caring for any but sick men, twenty or thirty years older than
herself. Blaming her is beside the point; she cannot help it,

126

What Means This Freedom? / 127

and she cannot change it. Countless criminal acts are thought ^
out in great detail; yet the participants are (without their own ,
knowledge) acting out fantasies, fears, and defenses from ;
early childhood, over whose coming and going they have no
conscious control.

Now, I am not saying that none of these persons should be
in jails or asylums. Often society must be protected against
them. Nor am I saying that people should cease the practices
of blaming and praising, punishing and rewarding; in general
these devices are justified by the results — although very often
they have practically no effect; the deeds are done from inner
compulsion, which is not lessened when the threat of punish-
ment is great. I am only saying that frequently persons we
think responsible are not properly to be called so; we mis-
takenly think them responsible because we assume they are
like those in whom no unconscious drive (toward this type of
behavior) is present, and that their behavior can be changed
by reasoning, exhorting, or threatening.

I have said that these persons are not responsible. But what
is the criterion for responsibility? Under precisely what con-
ditions is a person to be held morally responsible for an ac-
tion? Disregarding here those conditions that have to do with
a person's ignorance of the situation or the effects of his
action, let us concentrate on those having to do with his "inner
state." There are several criteria that might be suggested:

1 . The first idea that comes to mind is that responsibility is
determined by the presence or absence of premeditation — the
opposite of "premeditated" being, presumably, "unthinking"
or "impulsive." But this will not do — ^both because some acts
are not premeditated but responsible, and because some are
premeditated and not responsible.

Many acts we call responsible can be as unthinking or im-
pulsive as you please. If you rush across the street to help the
victim of an automobile collision, you are (at least so we
would ordinarily say) acting responsibly, but you did not do
so out of premeditation; you saw the accident, you didn't
think, you rushed to the scene without hesitation. It was like
a reflex action. But you acted responsibly: unlike the knee
jerk, the act was the result of past training and past thought
about situations of this kind; that is why you ran to help in-
stead of ignoring the incident or rurming away. When some-
thing done originally from conviction or training becomes

128 / Detenninism and Freedom

habitual, it becomes like a reflex action. As Aristotle said,
virtue should become second nature through habit: a virtuous
act should be performed as // by instinct; this, far from de-
tracting from its moral worth, testifies to one's mastery of the
desired type of behavior; one does not have to make a moral
effort each time it is repeated.

There are also premeditated acts for which, I would say,
the person is not responsible. Premeditation, especially when
it is so exaggerated as to issue in no action at all, can be the
result of neurotic disturbance or what we sometimes call an
emotional "block," which the person inherits from long-past
situations. In Hamlet's revenge on his uncle (I use this ex-
ample because it is familiar to all of us), there was no lack,
but rather a surfeit, of premeditation; his actions were so
exquisitely premeditated as to make Freud and Dr. Ernest
Jones look more closely to find out what lay behind them. The
very premeditation camouflaged unconscious motives of which
Hamlet himself was not aware. I think this is an important
point, since it seems that the courts often assume that pre-
meditation is a criterion of responsibility. If failure to kill his
uncle had been considered a crime, every court in the land
would have convicted Hamlet. Again: a woman's decision to
stay with her husband in spite of endless "mental cruelty" is,
if she is the victim of an unconscious masochistic "will to
punishment," one for which she is not responsible; she is the
victim and not the agent, no matter how profound her convic-
tion that she is the agent; she is caught in a masochistic web
(of complicated genesis) dating back to babyhood, perhaps
a repetition of a comparable situation involving her own
parents, a repetition-compulsion that, as Freud said, goes "be-
yond the pleasure principle." Again: a criminal whose crime
was carefully planned step by step is usually considered re-
sponsible, but as we shall see in later examples, the over-
whelming impulse toward it, stemming from an unusually
humiliating ego defeat in early childhood, was as compulsive
as any can be.

2. Shall we say, then, that a person is not responsible for his
act unless he can defend it with reasons? I am afraid that this
criterion is no better than the previous one. First, intellectuals
are usually better at giving reasons than nonintellectuals, and
according to this criterion would be more responsible than
persons acting from moral conviction not implemented by
reasoning; yet it is very doubtful whether we should want to
say that the latter are the more responsible. Second, the giving

What Means This Freedom? / 129

of reasons itself may be suspect. The reasons may be rationali-
zations camouflaging unconscious motives of which the agent
knows nothing. Hamlet gave many reasons for not doing what
he felt it was his duty to do: the time was not right, his uncle's
soul might go to heaven, etc. His various "reasons" contra-
dicted one another, and if an overpowering compulsion had
not been present, the highly intellectual Hamlet would not
have been taken in for a moment by these rationalizations.
The real reason, the Oedipal conflict that made his uncle's
crime the accomplishment of his own deepest desire, binding
their fates into one and paralyzing him into inaction, was un- |
conscious and of course unknown to him. One's intelligence tl
and reasoning power do not enable one to escape from un- p
consciously motivated behavior; it only gives one greater facil- |
ity in rationalizing that behavior; one's intelligence is simply |
used in the interests of the neurosis — it is pressed into service
to justfy with reasons what one does quite independently of
the reasons.

If these two criteria are inadequate, let us seek others.

3. Shall we say that a person is responsible for his action
unless it is the result of unconscious forces of which he knows
nothing? Many psychoanalysts would probably accept this
criterion. If it is not largely reflected in the language of re-
sponsibility as ordinarily used, this may be due to ignorance
of fact: most people do not know that there are such things
as unconscious motives and unconscious conflicts causing hu-
man beings to act. But it may be that if they did, perhaps they
would refrain from holding persons responsible for certain
actions.

I do not wish here to quarrel with this criterion of re-
sponsibility. I only want to point out the fact that if this criter-
ion is employed a far greater number of actions will be ex-
cluded from the domain of responsibility than we might at
first suppose. Whether we are neat or untidy, whether we are
selfish or unselfish, whether we provoke scenes or avoid them,
even whether we can exert our powers of will to change our
behavior — all these may, and often do, have their source in
our unconscious life.

4. Shall we say that a person is responsible for his act un-
less it is compelled? Here we are reminded of Aristotle's as-
sertion {Nicomachean Ethics, Book III) that a person is
responsible for his act except for reasons of either ignorance
or compulsion. Ignorance is not part of our problem here (un-
less it is unconsciously induced ignorance of facts previously

130 / Determinism and Freedom

remembered and selectively forgotten — in which case the for-
getting is again compulsive), but compulsion is. How will
compulsion do as a criterion? The difficulty is to state just
what it means. When we say an act is compelled in a psycho-
logical sense, our language is metaphorical — which is not to
say that there is no point in it or that, properly interpreted, it
is not true. Our actions are compelled in a literal sense if
someone has us in chains or is controlling our bodily move-
ments. When we say that the storm compelled us to jettison
the cargo of the ship (Aristotle's example), we have a less
literal sense of compulsion, for at least it is open to us to go
down with the ship. When psychoanalysts say that a man was
compelled by unconscious conflicts to wash his hands con-
stantly, this is also not a literal use of "compel"; for nobody
forced his hands under the tap. Still, it is a typical example
of what psychologists call compulsive behavior: it has uncon-
scious causes inaccessible to introspection, and moreover noth-
ing can change it — it is as inevitable for him to do it as it
would be if someone were forcing his hands under the tap. In
this it is exactly hke the action of a powerful external force; it
is just as little within one's conscious control.

In its area of application this interpretation of responsibility
comes to much the same as the previous one. And this area
. is very great indeed. For if we cannot be held responsible for
I the infantile situations (in which we were after all passive
victims), then neither, it would seem, can we be held respon-
sible for compulsive actions occurring in adulthood that are
inevitable consequences of those infantile situations. And,
psychiatrists and psychoanalysts tell us, actions fulfilling this
description are characteristic of all people some of the time
and some people most of the time. Their occurrence, once the
infantile events have taken place, is inevitable, just as the ex-
plosion is inevitable once the fuse has been lighted; there is
simply more "delayed action" in the psychological explosions
than there is in the physical ones.

(I have not used the word "inevitable" here to mean
"causally determined," for according to such a definition every
event would be inevitable if one accepted the causal principle
in some form or other; and probably nobody except certain
philosophers uses "inevitable" in this sense. Rather, I use
"inevitable" in its ordinary sense of "cannot be avoided." To
the extent, therefore, that adult neurotic manifestations can
be avoided, once the infantile patterns have become set, the
assertion that they are inevitable is not true.)

What Means This Freedom? / 131

5. There is still another criterion, which I prefer to the
previous ones, by which a man's responsibility for an act can
be measured: the degree to which that act can (or could have
been) changed by the use of reasons. Suppose that the man
who washes his hands constantly does so, he says, for hygienic
reasons, believing that if he doesn't do so he will be poisoned
by germs. We now convince him, on the best medical author- f,
ity, that his belief is groundless. Now, the test of his responsi-
bility is whether the changed belief will result in changed be- '
havior. If it does not, as with the compulsive hand washer, he
is not acting responsibly, but if it does, he is. It is not the use
of reasons, but their efficacy in changing behavior, that is
being made the criterion of responsibility. And clearly in
neurotic cases no such change occurs; in fact, this is often
made the defining characteristic of neurotic behavior: it is
unchangeable by any rational considerations.

n

I have suggested these criteria to distinguish actions for
which we can call the agent responsible from those for which
we cannot. Even persons with extensive knowledge of psychi-
atry do not, I think, use any one of these criteria to the exclu-
sion of the others; a conjunction of two or more may be used
at once. But however they may be combined or selected in
actual application, I believe we can make the distinction along
some such lines as we have suggested.

But is there not still another possible meaning of "responsi-
bility" that we have not yet mentioned? Even after we have
made all the above distinctions, there remains a question in
our minds whether we are, in the final analysis, responsible for
any of our actions at all. The issue may be put this way: How
can anyone be responsible for his actions, since they grow out
of his character, which is shaped and molded and made what
it is by influences — some hereditary, but most of them stem-
ming from early parental environment — that were not of his
own making or choosing? This question, I believe, still
troubles many people who would agree to all the distinctions
we have just made but still have the feeling that "this isn't all."
They have the uneasy suspicion that there is a more ultimate
sense, a "deeper" sense, in which we are not responsible for
our actions, since we are not responsible for the character out
of which those actions spring. This, of course, is the sense
Professor Edwards was describing.

Let us take as an example a criminal who, let us say.

132 / Determinism and Freedom

strangled several persons and is himself now condemned to
die in the electric chair. Jury and public alike hold him fully
responsible (at least they utter the words "he is responsible"),
for the murders were planned down to the minutest detail, and
the defendant tells the jury exactly how he planned them. But
now we find out how it all came about; we learn of parents
who rejected him from babyhood, of the childhood spent in
one foster home after another, where it was always plain to
him that he was not wanted; of the constantly frustrated early
desire for affection, the hard shell of nonchalance and bitter-
ness that he assumed to cover the painful and humiliating fact
of being unwanted, and his subsequent attempts to heal these
wounds to his shattered ego through defensive aggression.

The criminal is the most passive person in this world, helpless
as a baby in his motorically inexpressible fury. Not only does he
try to wreak revenge on the mother of the earliest period of his
babyhood; his criminality is based on the inner feeling of being
incapable of making the mother even feel that the child seeks
revenge on her. The situation is that of a dwarf trying to annoy
a giant who superciliously refuses to see these attempts. . . .
Because of his inner feeling of being a dwarf, the criminotic
uses, so to speak, dynamite. Of that the giant must take cogni-
zance. True, the "revenge" harms the avenger. He may be
legally executed. However, the primary inner aim of forcing
the giant to acknowledge the dwarf's fury is fulfilled.^

The poor victim is not conscious of the inner forces that exact
from him this ghastly toll; he battles, he schemes, he revels in
pseudo-aggression, he is miserable, but he does not know
what works within him to produce these catastrophic acts of
crime. His aggressive actions are the wriggling of a worm on
a fisherman's hook. And if this is so, it seems difficult to say
any longer, "He is responsible." Rather, we shall put him be-
hind bars for the protection of society, but we shall no longer
flatter our feeling of moral superiority by calling him person-
ally responsible for what he did.

Let us suppose it were established that a man commits
murder only if, sometime during the previous week, he has
eaten a certain combination of foods — say, tuna fish salad at
a meal also including peas, mushroom soup, and blueberry
pie. What if we were to track down the factors common to all
murders committed in this country during the last twenty

1 Edmund Bergler, The Basic Neurosis (New York: Grune and Stratton,
1949), p. 305.

What Means This Freedom? / 133

years and found this factor present in all of them, and only
in them? The example is of course empirically absurd; but
may it not be that there is some combination of factors that
regularly leads to homicide, factors such as are described in
general terms in the above quotation? (Indeed the situation in
the quotation is less fortunate than in our hypothetical ex-
ample, for it is easy to avoid certain foods once we have been
warned about them, but the situation of the infant is thrust
on him; something has already happened to him once and for
all, before he knows it has happened.) When such specific
factors are discovered, won't they make it clear that it is fool-
ish and pointless, as well as immoral, to hold human beings
responsible for crimes? Or, if one prefers biological to psycho-
logical factors, suppose a neurologist is called in to testify
at a murder trial and produces X-ray pictures of the brain of
the criminal; anyone can see, he argues, that the cella turcica
was already calcified at the age of nineteen; it should be a
flexible bone, growing, enabling the gland to grow.^ All the
defendant's disorders might have resulted from this early cal-
cification. Now, this particular explanation may be empirically
false; but who can say that no such factors, far more complex,
to be sure, exist?

When we know such things as these, we no longer feel so
much tempted to say that the criminal is responsible for his
crime; and we tend also (do we not?) to excuse him — not
legally (we still confine him to prison) but morally; we no
longer call him a monster or hold him personally responsible
for what he did. Moreover, we do this in general, not merely
in the case of crime: "You must excuse Grandmother for
being irritable; she's really quite ill and is suffering some pain
all the time." Or: "The dog always bites children after she's
had a litter of pups; you can't blame her for it: she's not feel-
ing well, and besides she naturally wants to defend them." Or:
"She's nervous and jumpy, but do excuse her: she has a severe
glandular disturbance." ,

Let us note that the more thoroughly and in detail we know
the causal factors leading a person to behave as he does, the
more we tend to exempt him from responsibility. When we
know nothing of the man except what we see him do, we say
he is an ungrateful cad who expects much of other people and
does nothing in return, and we are usually indignant. When
we leam that his parents were the same way and, having no

= Meyer Levin, Compulsion (New York: Simon and Schuster, 1956), p.
403.

134 / Determinism and Freedom

guilt feelings about this mode of behavior themselves, brought
him up to be greedy and avaricious, we see that we could
hardly expect him to have developed moral feelings in this
direction. When we learn, in addition, that he is not aware of
being ungrateful or selfish, but unconsciously represses the
memory of events unfavorable to himself, we feel that the
situation is unfortunate but "not really his fault." When we
know that this behavior of his, which makes others angry, oc-
curs more constantly when he feels tense or insecure, and that
he now feels tense and insecure, and that relief from pressure
will diminish it, then we tend to "feel sorry for the poor guy"
and say he's more to be pitied than censured. We no longer
want to say that he is personally responsible; we might rather
blame nature or his parents for having given him an unfortun-
ate constitution or temperament.

In recent years a new form of punishment has been imposed
on middle-aged and elderly parents. Their children, now in their
twenties, thirties or even forties, present them with a modem
grievance: "My analysis proves that you are responsible for my
neurosis." Overawed by these authoritative statements, the poor
tired parents fall easy victims to the newest variations on the
scapegoat theory.

In my opinion, this senseless cruelty — ^which disinters educa-
tional sins which had been burned for decades, and uses them as
the basis for accusations which the victims cannot answer — is
unjustified. Yes "the truth loves to be centrally located" (Mel-
ville), and few parents — since they are human — ^have been per-
fect. But granting their mistakes, they acted as their neurotic
difficulties forced them to act. To turn the tables and declare
the children not guilty because of the impersonal nature of their
own neuroses, while at the same time the parents are personally
blamed, is worse than illogical; it is profoxmdly unjust.^

And so, it would now appear, neither of the parties is respon-
sible: "they acted as their neurotic difficulties forced them to
act." The patients are not responsible for their neurotic mani-
festations, but then neither are the parents responsible for
theirs; and so, of course, for their parents in turn, and theirs
before them. It is the twentieth-century version of the family
curse, the curse on the House of Atreus.

"But," a critic complains, "it's immoral to exonerate people
indiscriminately in this way. I might have thought it fit to

• Edmund Bergler, The Superego (New York: Onme and Stratton, 1952),
p. 320.

What Means This Freedom? / 135

excuse somebody because he was born on the other side of the
tracks, if I didn't know so many bank presidents who were
also born on the other side of the tracks." Now, I submit that
the most immoral thing in this situation is the critic's carica-
ture of the conditions of the excuse. Nobody is excused
merely because he was born on the other side of the tracks.
But if he was born on the other side of the tracks and was a
highly narcissistic infant to begin with and was repudiated
or neglected by his parents and . . . (here we list a finite num-
ber of conditions), and if this complex of factors is regularly
followed by certain behavior traits in adulthood, and more-
over unavoidably so — that is, they occur no matter what he
or anyone else tries to do— then we excuse him morally and
say he is not responsible for his deed. If he is not responsible '
for A, a series of events occurring in his babyhood, then
neither is he responsible for B, a series of things he does in
adulthood, provided that B inevitably — ^that is, unavoidably
— follows upon the occurrence of A. And according to psychi-
atrists and psychoanalysts, this often happens. )

But one may still object that so far we have talked only
about neurotic behavior. Isn't nonneurotic or normal or not
unconsciously motivated (or whatever you want to call it)
behavior still within the area of responsibility? There are
reasons for answering "No" even here, for the normal person
no more than the neurotic one has caused his own character,
which makes him what he is. Granted that neurotics are not
responsible for their behavior (that part of it which we call
neurotic) because it stems from undigested infantile conflicts
that they had no part in bringing about, and that are external to
them just as surely as if their behavior had been forced on
them by a malevolent deity (which is indeed one theory on
the subject); but the so-called normal person is equally the
product of causes in which his volition took no part. And if,
unlike the neurotic's, his behavior is changeable by rational
considerations, and if he has the will power to overcome the
effects of an unfortunate early environment, this again is no
credit to him; he is just lucky. If energy is available to him in
a form in which it can be mobilized for constructive purposes,
this is no credit to him, for this too is part of his psychic leg-
acy. Those of us who can discipline ourselves and develop
habits of concentration of purpose tend to blame those who
cannot, and call them lazy and weak-willed; but what we fail
to see is that they literally cannot do what we expect; if their
psyches were structured like ours, they could, but as they are

136 / Determinism and Freedom

burdened with a tyrannical super-ego (to use psychoanalytic
jargon for the moment), and a weak defenseless ego whose
energies are constantly consumed in fighting endless charges
of the superego, they simply cannot do it, and it is irrational
to expect it of them. We cannot with justification blame them
for their inability, any more than we can congratulate our-
selves for our ability. This lesson is hard to learn, for we con-
stantly and naively assume that other people are constructed
as we ourselves are.

For example: A child raised under slum conditions, whose
parents are socially ambitious and envy families with money,
but who nevertheless squander the little they have on drink,
may simply be unable in later life to mobilize a drive sufficient
to overcome these early conditions. Common sense would
expect that he would develop the virtue of thrift; he would
make quite sure that be would never again endure the grind-
ing poverty he had experienced as a child. But in fact it is not
so: the exact conditions are too complex to be specified in de-
tail here, but when certain conditions are fulfilled (concerning
the subject's early life), he will always thereafter be a spend-
thrift, and no rational considerations will be able to change
this. He will listen to the rational considerations and see the
force of these, but they will not be able to change him, even
if he tries; he cannot change his wasteful habits any more than
he can lift the Empire State Building with his bare hands. We
moralize and plead with him to be thrifty, but we do not see
bow strong, how utterly overpowering, and how constantly
with him, is the opposite drive, which is so easily manageable
with us. But he is possessed by the all-consuming, all-encom-
passing urge to make the world see that he belongs, that he
has arrived, that he is just as well oflE as anyone else, that the
awful humiliations were not real, that they never actually oc-
curred, for isn't he now able to spend and spend? The humilia-
tion must be blotted out; and conspicuous, fleshy, expensive,
and wasteful buying will do this; it shows the world what
the world must know! True, it is only for the moment; true,
it is m the end self-defeating, for wasteful consumption is the
best way to bring poverty back again; but the person with an
overpowering drive to mend a lesion to his narcissism cannot
resist the avalanche of that drive with his puny rational con-
sideration. A man with his back against the wall and a gun at
his throat doesn't think of what may happen ten years hence.
(Consciously, of course, he knows nothing of this drive; all
that appears to consciousness is its shattering effects; he

WThiat Means This Freedom? / 137

knows only that he must keep on spending — not why — and
that he is unable to resist.) He hasn't in him the psychic ca-
pacity, the energy to stem the tide of a drive that at that mo-
ment is all-powerful. We, seated comfortably away from this
flood, sit in judgment on him and blame him and exhort him
and criticize him; but he, carried along by the flood, cannot
do otherwise than he does. He may fight with all the strength
of which he is capable, but it is not enough. And we, who are
rational enough at least to exonerate a man in a situation of
"overpowering impulse" when we recognize it to be one, do
not even recognize this as an example of it; and so, in addition
to being swept away in the flood that childhood conditions
rendered inevitable, he must also endure our lectures, our
criticisms, and our moral excoriation. -s,^

But, one will say, he could have overcome his spendthrift]
tendencies; some people do. Quite true: some people do. They
are lucky. They have it in them to overcome early deficiencies >
by exerting great effort, and they are capable of exerting the '
effort. Some of us, luckier still, can overcome them with but
little effort; and a few, the luckiest, haven't the deficiencies to
overcome. It's all a matter of luck. The least lucky are those
who can't overcome them, even with great effort, and those
who haven't the ability to exert the effort. ~^

But, one persists, it isn't a matter simply of luck; it is a
matter of effort. Very well then, it's a matter of effort; without X
exerting the effort you may not overcome the deficiency. But
whether or not you are the kind of person who has it in himj
to exert the effort is a matter of luck.

All this is well known to psychoanalysts. They can predict,
from minimal cues that most of us don't notice, whether a
person is going to turn out to be lucky or not. "The analyst,"
they say, "must be able to use the residue of the patient's un-
conscious guilt so as to remove the symptom or character trait
that creates the guilt. The guilt must not only be present, but
available for use, mobilizable. If it is used up (absorbed) in
criminal activity, or in an excessive amount of self-damaging
tendencies, then it cannot be used for therapeutic purposes,
and the prognosis is negative." Not all philosophers will relish
the analyst's wav of putting the matter, but at least as a physi-
cian he can soon detect whether the patient is lucky or un-
lucky— and he knows that whichever it is, it isn't the patient's
fault. The patient's conscious volition cannot remedy the
deficiency. Even whether he will co-operate with the analyst
is really out of the patient's hands: if he continually projects

138 / Determinism and Freedom

the denying-mother fantasy on the analyst and unconsciously
identifies him always with the cruel, harsh forbidder of the
nursery, thus frustrating any attempt at impersonal observa-
tion, the sessions are useless; yet if it happens that way, he
can't help that either. That fatal projection is not under his
control; whether it occurs or not depends on how his uncon-
scious identifications have developed since his infancy. He can
try, yes — but the ability to try enough for the therapy to have
effect is also beyond his control; the capacity to try more than
just so much is either there or it isn't — and either way "it's in
the lap of the gods."

The position, then, is this: if we can overcome the effects of
early environment, the ability to do so is itself a product of
the early environment. We did not give ourselves this ability;
and if we lack it we cannot be blamed for not having it. Some-
times, to be sure, moral exhortation brings out an ability that
is there but not being used, and in this lies its occasional util-
ity; but very often its use is pointless, because the ability is not
there. The only thing that can overcome a desire, as Spinoza
said, is a stronger contrary desire; and many times there
simply is no wherewithal for producing a stronger contrary
desire. Those of us who do have the wherewithal are lucky.

There is one possible practical advantage in remembering
this. It may prevent us (unless we are compulsive blamers)
from indulging in righteous indignation and committing the
sin of spiritual pride, thanking God that we are not as this
publican here. And it will protect from our useless moralizings
those who are least equipped by nature for enduring them.
As with responsibility, so with deserts. Someone commits a
crime and is punished by the state; "he deserved it," we say
self-righteously — as if we were moral and he immoral, when
in fact we are lucky and he is unlucky — forgetting that there,
but for the grace of God and a fortunate early environment,
go we. Or, as Clarence Darrow said in his speech for the de-
fense in the Loeb-Leopold case:

I do not believe that people are in jail because they deserve to
be. ... I know what causes the emotional life. ... I know it is
practically left out of some. Without it they cannot act with the
rest. They cannot feel the moral shocks which safeguard others.
Is [this man] to blame that his machine is imperfect? Who is
to blame? I do not know. I have never in my life been interested
so much in fixing blame as I have in relieving people from
blame. I am not wise enough to fix it.*

* Levin, op. cit., pp. 439-40, 469.

What Means This Freedom? / 139

I want to make it quite clear that I have not been arguing
for determinism. Though I find it difficult to give any sense to
the term "indeterminism," because I do not know what it
would be like to come across an uncaused event, let us grant
indeterminists everything they want, at least in words —
influences that suggest but do not constrain, a measure of
acausality in an otherwise rigidly causal order, and so on —
whatever these phrases may mean. With all this granted, ex-
actly the same situation faces the indeterminist and the deter-
minist; all we have been saying would still hold true. "Are
our powers innate or acquired?" <^

Suppose the powers are declared innate; then the villain may
sensibly ask whether he is responsible for what he was bom
with. A negative reply is inevitable.Are they then acquired? Then
the ability to acquire them — ^was that innate? or acquired? It is
innate? Very well then. . . .^

The same fact remains — ^that we did not cause our characters,
that the influences that made us what we are are influences
over which we had no control and of whose very existence we
had no knowledge at the time. This fact remains for "deter-
minism" and "indeterminism" alike. And it is this fact to
which I would appeal, not the specific tenets of traditional
forms of "determinism," which seem to me, when analyzed,
empirically empty.

"But," it may be asked, "isn't it your view that nothing
ultimately could be other than it is? And isn't this determin-
istic? And isn't it deterministic if you say that human beings
could never act otherwise than they do, and that their desires
and temperaments could not, when you consider their ante-
cedent conditions, be other than they are?"

I reply that all these charges rest on confusions.

1. To say that nothing could be other than it is, is, taken
literally, nonsense; and if taken as a way of saying something
else, misleading and confusing. If you say, "I can't do it," this
invites the question, "No? Not even if you want to?" "Can"
and "could" are power words, used in the context of human
action; when appBeH^o nature they are merely anthropomor-

* This section of Professor riospers' paper was not read In Its present form
at the conference. — Ed.

" W. I. Matson, "The Irrelevance of Free-wiU to Moral Responsibility,"
Uind, LXV (October 1956), p. 495.

140 / Determinism and Freedom

phic. "Could" has no application to natxire — unless, of course,
it is uttered in a theological context: one might say that God
could have made things different. But with regard to inani-
mate nature "could" has no meaning. Or perhaps it is in-
tended to mean that the order of nature is in some sense
necessary. But in that case the sense of "necessary" must be
specified. I know what "necessary" means when we are talk-
ing about propositions, but not when we are talking about the
sequence of events in nature.

2. What of the charge that we could never have acted
otherwise than we did? This, I submit, is simply not true. Here
the exponents of Hume-Mill-Schlick-Ayer "soft determinism"
are quite right. I could have gone to the opera today instead
of coming here; that is, if certain conditions had been differ-
ent, I should have gone. I could have done many other things
instead of what I did, if some condition or other had been
different, specifically if my desire had been different. I repeat
that "could" is a power word, and "I could have done this"
means approximately "I should have done this if I had wanted
to." In this sense, all of us could often have done otherwise
than we did. I would not want to say that I should have done
differently even if all the conditions leading up to my action
had been the same (this is generally not what we mean by
"could" anyway) ; but to assert that I could have is empty, for
if I did act different from the time before, we would auto-
matically say that one or more of the conditions were differ-
ent, whether we had independent evidence for this or not,
thus rendering the assertion immune to empirical refutation.
(Once again, the vacuousness of "determinism.")

3. Well, then, could we ever have, not acted, but desired
otherwise than we did desire? This gets us once again to the
heart of the matter we were discussing in the previous section.

/ Russell said, "We can do as we please but we can't please as
we please." But I am persuaded that even this statement con-

\ ceals a fatal mistake. Let us follow the same analysis through.

j "I could have done X" means "I should have done AT if I had

I wanted to." "I could have wanted AT" by the same analysis
would mean "I should have wanted X if I had wanted to" —
which seems to make no sense at all. (What does Russell
want? To please as he doesn't please?)
fl 1/ What does this show? It shows, I think, that the only
'^•' ' meaningful context of "can" and "could have" is that of
action. "Could have acted differently" makes sense; "could

What Means This Freedom? / 141

have desired differently," as we have just seen, does not. Be-
cause a word or phrase makes good sense in one context, let
us not assume that it does so in another. ^

I conclude, then, with the following suggestion: that we|/T|
operate on two levels of moral discourse, which we shouldn't'
confuse; one (let's call it the upper level) is that of actions; ^l'
the other (the lower, or deeper, level) is that of the springs
of action. Most moral talk occurs on the upper level. It is
on this level that the Hume-Mill-Schlick-Ayer analysis of
freedom fully applies. As we have just seen, "can" and
"could" acquire their meaning on this level; so, I suspect, does
"freedom." So does the distinction between compulsive and
noncompulsive behavior, and among the senses of "responsi-
bility," discussed in the first section of this paper, according
to which we are responsible for some things and not for
others. All these distinctions are perfectly valid on this level
(or in this dimension) of moral discourse; and it is, after all,
the usual one — we are practical beings interested in changing
the course of human behavior, so it is natural enough that 99
per cent of our moral talk occurs here.

But when we descend to what I have called the lower level
of moral discourse, as we occasionally do in thoughtful mo-
ments when there is no immediate need for action, then we
must admit that we are ultimately the kind of persons we are
because of conditions occurring outside us, over which we had
no control. But while this is true, we should beware of extend-
ing the moral terminology we used on the other level to this
one also. "Could" and "can," as we have seen, no longer have
meaning here. "Right" and "wrong," which apply only to
actions, have no meaning here either. I suspect that the same
is true of "responsibility," for now that we have recalled often
forgotten facts about our being the product of outside forces,
we must ask in all seriousness what would be added by saying
that we are not responsible for our own characters and tem-
peraments. What would it mean even? Has it a significant
opposite? What would it be like to be responsible for one's
own character? What possible situation is describable by this
phrase? Instead of saying that it is false that we are responsi-
ble for our own characters, I should prefer to say that the
utterance is meaningless — meaningless in the sense that it de-
scribes no possible situation, though it seems to because the
word "responsible" is the same one we used on the upper
level, where it marks a real distinction. If this is so, the result
is that moral terms — at least the terms "could have" and "re-

H2 / Determinism and Freedom

sponsible" — simply drop out on the lower level. What re-
mains, shorn now of moral terminology, is the point we tried
to bring out in Part 11: whether or not we have personality
disturbances, whether or not we have the ability to overcome
deficiences of early environment, is like the answer to the
question whether or not we shall be struck down by a dread
disease: "it's all a matter of luck." It is important to keep this
in mind, for people almost always forget it, with consequences
in human intolerance and unnecessary suffering that are in-
calculable.

PART iV

DI§€ussion

Chapter 1

"Excusing Conditions" and
Moral Responsibility

Elizabeth Lane Beardsley, Lincoln University

Mr. Hart presents an extremely interesting justification of
the use of "excusing conditions" as removing liability to legal
penalties on the ground that the individual's powers of pre-
diction and choice are thereby maximized. That is, the in-
dividual can be more confident of being able to predict and
control the possibility that legal penalties wUl be applied to
him that he could in a society that abolished excusing condi-
tions. As I understand Mr. Hart's thesis, this consideration
is deemed sufficient to justify the use of excusing conditions.
But, although this line of thought seems to shed considerable
light on the problem why we approve of excusing conditions,
I am not quite convinced that this can be the entire truth of
the matter.

It seems clear that if I lived in a society in which all liability
to legal penalties was of the "strict" kind, I should stand in
danger of suffering such penalties without having really
chosen to commit the act that brought them upon me. Here
the absence of excusing conditions is a necessary condition of
my standing in such danger. The extent of the danger, how-
ever, depends also on the likelihood of my ever being in one
of the conditions that in other societies would be regarded as
excusing. And some of these conditions, such as mental dis-
ease and psychopathic states in general, and also ignorance,
appear capable, in principle at least, of being reduced in inci-
dence by fairly direct action.

If we now perform the kind of mental experiment so skill-
fully carried out by Mr. Hart, we may compare the situation
in two different hypothetical societies, 51 and 52. In 51, all
liability to legal sanctions is of the strict kind, but the chances
of anyone's suffering such sanctions because of insanity or
neurosis or ignorance are small, because the incidence of these
conditions is low. In 52, on the other hand, insanity, neurosis,

145

146 / Determmism and Freedom

and ignorance are much more prevalent, but here they consti-
tute excusing conditions. (A member of S2 stands in some
slight danger of suffering legal penalties for acts committed
when in these conditions, because of the difficulties of estab-
lishing the presence or absence of the conditions.) In both 51
and 52, the danger of suffering legal penalties for acts com-
mitted because of mental aberration or ignorance is small —
let us assume that it is equally small. The reduction of this
danger is certainly an achievement to which we give moral
approval. But in 51 and 52 it has been accomplished by differ-
ent means, and it seems to me that our moral appraisal of
the methods used by the two societies is not the same. There
is an element of moral value in the method used by 52 that
is absent in what has been done in 51. This seems to indicate
that something more than the power to maximize the indi-
vidual's power of prediction and choice underlies our ap-
proval of the use of at least certain ones of the excusing
conditions.

^

n

The thesis set forth by C. A. Campbell,^ and Edwards, that
people differ with regard to their reflectiveness concerning the
problem of relating the concepts of cause and moral blame, is
helpful. It seems to me, however, that Mr. Edwards' two-rung
hierarchy is a truncated one, and that it needs to be extended
at both ends. I should like to distinguish four degrees or levels
of reflectiveness on this matter. On the bottom level would be
persons who do not even require that the immediate causes of
an act be of a certain kind, i.e., that the act be voluntary (free
from "coercion"), before the agent is blamed for it. I do not
know whether any such persons now exist. Perhaps they do
not, even in our simplest societies, though even complex so-
cieties count among their members those who sometimes
direct judgments of blame toward agents without first ascer-
taining that their acts were voluntary. In any case, it is
certainly plausible to assume that the hxmian race, in the
evolution of its moral thinking, once passed through the stage
of failing to recognize the moral significance of the distinction
between voluntary and nonvoluntary acts. To point this out
helps us remember that, however little reflective Mr. Edwards'
"unreflective" individuals may be, it would be possible to be
even less so. The individuals termed "unreflective" by Mr.

^Mlnd, 1951.

"Excusing Conditions" and Moral Responsibility / 147

Edwards occupy the second level in my proposed hierarchy.
For these persons the voluntariness of an act is a condition of
the blameworthiness of the agent. This means that they ask
certain questions regarding the immediate causes of an act;
but they do not go on to ask about the causes of those causes,
nor do they feel impelled to bring into line their views about
the scope of causality in general, on the one hand, and their
beliefs and practices regarding moral blame, on the other.

The third degree of reflectiveness is reached by Mr. Ed-
wards' "reflective" persons. These persons, as I interpret their
state of mind, are disturbed by several considerations that do
not trouble their second-level associates. The former agree
that men are responsible only for their voluntary acts, but
they see that the process of asking for the causal conditions of
a given act can be carried much farther than is required
merely to establish that the act is voluntary, or chosen by the
agent; and they do not see any reason why further questions
about more remote causal conditions leading to the choice it-
self should be ruled out in an appraisal of the blameworthiness
of the agent for the act chosen. Moreover, persons at this
third level of reflectiveness are aware that when these further
questions are asked, the answers will very soon begin to in-
clude references to factors that are not themselves voluntary
acts of the agent, and for which the agent cannot be held
"responsible" from any point of view. The reaction of such
persons is then to wonder how one can justify regarding any
agent as blameworthy for an act, however "voluntary" it
may seem after a limited inquiry, when it is clear that a
further inquiry would disclose causal antecedents of the act
that are not voluntary acts. It appears that persons at this
level of reflectiveness make one of two drastic moves. They ]
may feel so sure that the notions of moral blame and responsi-
bility must be retained that they abandon the thesis that all
acts of choosing have causal antecedents (Campbell's view),
or they may feel so sure that all events, .including acts of
choosing, are caused that they abandon the notions of moral
blame and responsibility (the "hard determinism" of Edwards'
view).

Many moral philosophers of a determinist bent will hope
that there is a still higher level of reflectiveness here, to which
we may aspire. I think that there is, although I do not know
how to describe its basic characteristics very precisely. But I
think that such a fourth level of reflectiveness can perhaps be
reached by those who see both the force and the flaws of the

148 / Determinism and Freedom

arguments that seem so persuasive at the third level. Such per-
sons would see that the fact that the concepts of moral re-
sponsibility and blame have limits in their application does
not mean that they have no application. That they do have
limits is demonstrated by the considerations advanced at the
third level. Surely these show that there are indeed some ways
of feeling about other people, some acts of total rejection of
their characters, some judgments of unlimited condemnation
for their acts, that are wholly unjustified, no matter what
monstrous things they may do. Whatever degree of condem-
nation would be properly reserved for a first cause of the
doing of evil is certainly to be withheld from men. But this
is not to say that no condemnation whatsoever is to be applied
to men. Third-level theorists apparently believe that the de-
cision not to blame the doers of acts that include factors
other than voluntary acts among their causal antecedents is
reached by a mere extension of the very same reasoning that
leads us not to blame the doers of nonvoluntary acts; but
this is not the case. The two premises, (1) that we are not
morally responsible for acts unless they are voluntary, and
(2) that for many of the causes of our acts we are not morally
responsible, since these are not themselves voluntary acts, are
not sufficient to yield the conclusion that we are not morally
responsible for our voluntary acts. But those who wish to
avoid this conclusion, without renouncing determinism in the
process, cannot content themselves with the negative task of
looking for loopholes in third-level arguments. Some rather
complicated analyses of the standards that actually govern our
judgments of moral praise and blame will be needed, along
with much other work. The construction of a position at the
fourth level of reflectiveness presents a difficult challenge; but
the stimulating papers of Edwards and Hospers, as well as
the spirited discussion that followed, have provided a strong
incentive for continuing to try to meet that challenge.

Chapter 2

P®l^@rmmiSffs @rad the Jystafo^baiit/

Richard Brandt, Swarthmore College

People often say that someone's act was "reprehensible" or
"morally blameworthy" or "admirable" or "praiseworthy";
and they often have correspondingly favorable or unfavor-
able attitudes toward individuals on account of their acts.
Furthermore, they often make very similar remarks about
the character of persons, or about persons on account of their
character; and again they sometimes have favorable or un-
favorable attitudes toward persons on account of their char-
acter. We might sum all this up by saying that people
sometimes engage in "blaming" or "praising." (A person need
not say anything aloud in order to blame; it is enough if he
makes a mental appraisal and takes up a corresponding atti-
tude.)

When philosophers say that human beings are "morally
responsible" for their actions, what they apparently mean —
although one perhaps does them an injustice if one supposes
anything specific is meant — is that it is right and proper,
sometimes, to engage in blaming and praising as defined above.
People are sometimes fittingly, deservedly, praised and
blamed. These philosophers are not just making a causal
statement, such as "Human volitions are sometimes uncaused
beginnings of causal series," although some such causal propo-
sition may be part of their reason for saying that human
beings are "morally responsible." (Sometimes, too, when
philosophers say that people "act freely" they are not assert-
ing any definite causal proposition, but rather, simply, that
those causal propositions are true about human behavior that
are not inconsistent with the fittingness of blaming and prais-
ing behavior.)

A great many distinguished philosophers have held that,
if determinism is a correct theory of all human psychological
processes, then people are not "morally responsible" for their
actions or character; vis., blaming and praising are not really
fitting. I am unconvinced by their reasoning, however, and

149

150 / Determinism and Freedom

shall now explain why I think determinism is not inconsistent
with moral responsibility.

It is convenient to begin by ascribing to these philosophers,
who think determinism requires serious revision of ordinary
moral thinking, a rather radical thesis, as follows. Tradition-
ally, both in law and in morals, some acts have been regarded
as excusable because of certain specific conditions, and other
acts have been regarded as inexcusable in view of the absence
of such specific conditions. Now, the radical thesis to be con-
sidered is this: that no actions are ever inexcusable, that all
actions are excusable in view of their having been caused.

It would be unfair to suggest that any philosophers advo-
cate this view in such a sweeping and unqualified form. So
we must consider how this thesis should be complicated if
it is to be seriously defended.

First, nobody holds that society has no need for a system of
criminal justice. Moreover, given such a system with legal re-
quirements for conduct and sanctions in case of infractions,
there must be some actions that are inexcusable as far as the
law is concerned. Everyone is agreed that misconduct mvist be
subject to punishment, as a condition of the protection of the
rights of all, as a condition of a well-ordered society where
people live in security. It is also agreed, I suppose, that any ac-
ceptable system of criminal justice will excuse antisocial and
forbidden behavior when it is unintentional, manifests itself
under duress, etc., and that it wUl not excuse such behaviw
when it is deliberate, uncompelled — in short, when none of
the standard defenses against a criminal charge apply. This
distinction will stand, I think, irrespective of improvements
in the system of criminal justice due to advances in psychol-
ogy. So far, it is not clear that the acceptance of determinism
indicates any modifications.

Second, the determinist must assent to a further utility of
the distinction between excusable and inexcusable actions.
It would be agreed today that up to a point it is correct to
view moral criticism and accusations as an informal extension
of the system of criminal justice. Moral criticism is a mild
form of punishment; its occurrence is a sanction, the opera-
tion of which can warn and teach just as do criminal codes
and criminal proceedings. Moral criticism, like legal sanc-
tions, is a device for social control that is justified — at any
rate, among other things — by its good effects. And, as in
criminal law, it is a good thing for moral criticism to recog-
nize certain antisocial and forbidden behavior as excusable

Determinism and the Justifiability of Moral Blame / 151

under specific conditions similar in general type to those
under which a person is not liable for his conduct before the
law; and it is a good thing for moral criticism to recognize
misconduct as inexcusable when none of these defenses are
available. Again, then, it is not clear that the acceptance of
determinism calls for modifications in the exercise of moral
criticism.

If the determinist finds the above two points congenial, it
is not easy to see how he will avoid going still farther. We
must remember that, if we are to advocate moral criticism as
a means of social control for the sake of the general welfare,
we must specify the conditions necessary for the application
of moral criticism. If moral criticism is to be effective it must
be sincere, and if it is to be sincere the critic must be so con-
structed that he genuinely disapproves of the behavior or
character trait he is criticizing; the capacity to be unfavorably
excited toward persons who misbehave must be built in.
Moreover, if, as suggested, what is wanted is a system of
criticism that recognizes a distinction between excusable and
inexcusable behavior, it appears we must also approve of
people's recognizing in their own minds and consciences a
distinction between excusable and inexcusable behavior.

In view of these considerations we may well ask philoso-
phers who think determinism has sweeping importance for
moral philosophy: What exactly is it that the truth of deter-
minism renders indefensible? Are there some practices — of
moral criticism, etc. — that the consistent determinist must
abandon, and if so, which? In what sense can it be said that,
if determinism is true, no actions are inexcusable?

Perhaps what these philosophers are arguing is this: that
whereas, even if determinism is true, praising and blaming
can be justified to some extent by the above utilitarian appeal
— so far the distinction between the excusable and the in-
excusable stands up — nevertheless no further justification can
be given for these practices. And it may be thought, in view
of the essential point or significance of blaming or praising,
some further justification must be given, if these practices
are to be accepted as fitting and proper. I agree that the utili-
tarian justification is not enough, but I am unconvinced that
further satisfactory justification is impossible.

Let us be clear what is considered unfitting by these phi-
losophers: (1) that people should feel disgust, contempt,
anger, etc., toward others on account of their misbehavior
(instead of excusing them) and judge correspondingly; (2)

152 / Determinism and Freedom

that they should approve of making human beings suffer for
their past deeds, when other things are equal (as they perhaps
seldom are, since, for example, suffering is something to be
avoided); (3) that they should feel remorseful about their
own past deeds and not excuse themselves; (4) that they
should take pride in having done what they think they ought
when it was hard to do so; (5) that they should feel admira-
tion or respect for others because they did what they ought
when few would have been able to do so; and (6) that they
should approve of rewarding people for special achievements,
when they did what they ought though it was hard. Perhaps
we are unfairly including too much. But what ought we to
exclude, on the view proposed?

I am suggesting that the philosophers who say that deter-
minism implies that people are not "morally responsible" are
saying, in part, that the above activities, although they can be
justified to some extent by their utility, cannot, in view of the
truth of determinism, be justified in other respects, and that
the utilitarian justification is not enough.

On what argument do these philosophers rely in order to
show that, considerations of utility aside, moral criticism and
the distinction between excusable and inexcusable behavior
are unjustified if determinism is true? Professor Edwards ar-
gues, in effect, that a person who was convinced of the truth
of determinism would (if he were disregarding the suggested
utilities) never practice the various forms of moral criticism
— or approve of such practice — if he were an impartial person
in a calm frame of mind. I agree with him that, if such feel-
ings would not occxir if one were calm and impartial, they are
not moral feelings when they do occur, and that then, if deter-
minism is true, we probably ought not to say that people are
blameworthy or admirable. (I have some qualms, though,
since a person who thought of the utilities might still approve
of such critical reactions.) And it is conceivable — although
here I would very much question whether in fact it has come
about, even among "reflective" people — ^that "blameworthy"
and "admirable" should be so used that a necessary condition
for their applicability to an act or character would be that the
latter be undetermined.

But is it the case that we incline to stop rendering — or ap-
proving the rendering of — ^moral judgments in our impersonal,
nonviolent moments when we bring to mind the fact that all
our behavior is determined and disregard the question of util-
ity? Are we inclined to be less provoked with ourselves if we

Determinism and the Justifiability of Moral Blame / 153

notice, for instance, how we have given way to an envious
thought or motive — that we have been expounding a shoddy
argument we should have more carefully scrutinized — when
we think of the way all this was determined? Or suppose we
hear reliably that one of our colleagues bears a baseless
grudge against a student, that he marks the student's examina-
tions low for no reason, that he refuses to Usten to his ques-
tions, that he refuses to give him a recommendation necessary
for medical school. Do we, when we reflect on the determin-
ism in human behavior, find our indignation melted — in the
way it is melted if we hear that his attitude was based on some
serious misunderstanding, or even that he has been under
severe emotional strain on account of personal difficulties? Or
does our admiration for someone who has stood up for a
principle at the risk of losing his job evaporate when we re-
flect that after all, given his make-up, such behavior was in
the cards — as it does when we learn that our man knew from
the start that his job was safe? It is not obvious that "No" is
a wrong answer to these questions.

Further debate on these important issues is doubtless called
for. Possibly it can be shown that a person who is unimpressed
by the plea that some selfish action should be excused be-
cause it was determined somehow does not have a clear view
of what it means for an action to be caused. Or again it might
be shown that there is no formulable principle that will dis-
tinguish conditions under which we count misbehavior as ex-
cusable — e.g., ignorance or incapacity — from conditions
under which we should refuse to regard misbehavior as excus-
able. (If we do not have such a principle ready, we might still
decide that it is more plausible to keep on looking for a satis-
factory formulation than to give up a distinction between
tJTses of cases that strike us as very different.)

My comments may be summarized as follows. First, the
distinction between excusable and inexcusable misbehavior
cannot be abandoned altogether. It is necessary for a desirable
system of criminal law, and there are strong reasons of utility
for its preservation in overt moral criticism and perhaps in the
private moral thinking and feelings of mankind, irrespective
of the truth of determinism. Second, it remains to be shown
that "reprehensible" €«■ "morally admirable" entails "was un-
determined" in the usage of reflective people, or that the mak-
ing of judgments of blame or the presence of corresponding
feelings is causally incompatible with believing that the act in
question was determined, even in a calm and impersonal

154 / Determinism and Freedom

frame of mind. And therefore the judgments and feelings
whereby we distinguish between excusable and inexcusable be-
havior are not undermined as being unfitting, even if the truth
of determinism is granted. My conclusion is that the implica-
tions of determinism for ordinary thinking about praise-
worthiness and blameworthiness are by no means as serious
as some philosophers suggest.

The above remarks have been concerned only with the con-
sistency of judgments of praise and blame with determinism.
But they bear also on the consistency with determinism of
judgments of duty and obligation if, as I believe to be the case,
saying that a certain action is one's duty is to say that one will
be morally to blame if one fails to do it unless a valid excuse
can be offered.

Chapter 3

Determinism and Punishment

Percy W. Bridgman, Harvard University

At the present time there seems to be one question that
looms as most important in the minds of many who are con-
cerned with the question of punishment. This question is
whether punishment is an acceptable line of conduct for
society in the context provided by psychoanalysis, which pic-
tures every single act of every individual as fully determined
by factors over which he has no control. Thus stated, the
problem whether to punish or not becomes a rather special
subcase of the much more general problem of reconciling two
patently inconsistent points of view, that of determinism and
"free will." It is obviously of the utmost practical importance
that we find a workable solution of this problem. Otherwise
we become victims of a cancerous confusion leading to vacil-
lation, like that of the donkey between two bales of hay.

It seems to me that we have to recognize clearly that there
are two levels of operation. There is the level of daily life and
social interaction, i.e., the level of "free will," and there is the
deterministic level. So far as the deterministic level has con-
crete reference beyond the purely verbal it is the level of
scientific activity. At this level, so far as present achievement
goes, determinism has the status merely of a program to direct
inquiry, a program applicable to the overwhelming majority
of the phenomena of the world about us, including biological
phenomena. It is a simple description of the attitude of many
scientists to say that they can see nothing on the present hori-
zon that would make this an impossible program, and that in
many fields they regard finding methods of carrying out this
program as the most promising line of scientific attack. It must
be emphasized, however, that every biologist, and particularly
every psychologist, would admit that at present we are fan-
tastically far from being able to carry out such a program.

The other level is the common-sense level of everyday life,
the level of "free will." On this level we have to devise a
practical method of dealing with situations in which we can-
not control or predict. Such situations occur predominantly in
dealing with organisms. In particular, no one can find in his

155

156 / Determinism and Freedom

own consciousness or outside it factors that would enable him
to predict his own future behavior. We develop a language to
describe this situation, in which our inability to foresee our
own future and that of our fellows is reflected in the concept
of "free will," and we further develop a whole related vocabu-
lary for situations in which we make no attempt at control or
prediction. It is a mere statement of fact that there are situa-
tions in which we make no attempt at control or prediction,
and any ultimate possibility of such control is disregarded.
Whether or not we believe that we might at some time in the
future achieve such control becomes, in this context,
irrelevant.

There is, and can be, no sharp dividing line between the
vocabulary of determinism and that of daily life. Use of the
vocabulary of daily life is an art, and the wisdom of all ages
is necessary to use it effectively.

It seems to me that much of the current unwillingness to
use the instrument of punishment under conditions that would
be acceptable to enlightened social opinion stems from a doc-
trinaire insistence that our verbal edifice be a single logically
consistent unit. It is in the nature of things impossible to erect
a single consistent verbal structure, logically watertight in all
respects. To insist on acting as if we could is in the first place
self-defeating — for by what logic can the man who argues that
punishment is unjustified expect his argument to affect the ac-
tions of his opponent, when both his argument and the re-
sponse to it were already rigidly predetermined? Pushed still
further, the insistence that punishment is unjustified can lead
only to social catastrophe. At present the only technique we
have for dealing with our fellows is to act as if they were the
same sort of creatures as we ourselves. We disregard deter-
minism when dealing with ourselves — we have to disregard
it, within reason, in our everyday contacts with others. Too
many of us take our verbal structures with a deadly serious-
ness— a certain tough-mindedness and small sense of humor
might provide an antidote.

Chapter 4 ■

Responsibility and Avoidability

Roderick W. Chisholm, Brown University

Edwards and Hospers hold that there is an important sense
in which we may be said not to be morally responsible for
any of our acts or choices. I propose the following as an
explicit formulation of their reasoning:

1. If a choice is one we could not have avoided making,
then it is one for which we are not morally responsi-
ble.

2. If we make a choice under conditions such that, given
those conditions, it is (causally but not logically) im-
possible for the choice not to be made, then the
choice is one we could not have avoided making,

3. Every event occvu"s under conditions such that, given
those conditions, it is (causally but not logically)
impossible for that event not to occur.

4. The making of a choice is the occurrence of an event.

5. We are not morally responsible for any of our
choices.

If we wish to reject the conclusion (5) — and for most of us
(5) is difficult to accept — we must reject at least one of the
premises.

Premise ( 1 ) , I think, may be interpreted as a logical truth.
If a man is responsible for what he did, then we may say,
"He could have done otherwise." And if we may say, "He
couldn't help it," then he is not responsible for what he did.

Many philosophers would deny (2), substituting a weaker
account of avoidability. A choice is avoidable, they might say,
provided only it is such that, // the agent had reflected further,
or had reflected on certain things on which in fact he did not
reflect, he would not have made the choice. To say of a choice
that it "could not have been avoided," in accordance with this

157

158 / Determinism and Freedom

account, would be to say that, even if the agent had reflected
further, on anything you like, he would all the same have
made the choice. But such conditional accounts of avoidability
("An act or choice is avoidable provided only it is such that,
// the agent were to do so-and-so, the act or choice would not
occur") usually have this serious defect: the antecedent
clause ("if the agent were to do so-and-so") refers to some
act or choice, or to the failure to perform some act or to
make some choice; hence we may ask, concerning the occur-
rence or nonoccurrence of this act or choice, whether or not
it is avoidable. Thus one who accepted (5) could say that, if
the agent's failure to reflect further was itself unavoidable, his
choice was also unavoidable. And no such conditional ac-
count of avoidability seems adequate to the use of "avoid-
able" and "unavoidable" in questions and statements such as
these.

If we accept a conditional account of avoidability, we may
be tempted to say, of course, that it would be a misuse of
"avoidable" to ask whether the nonoccurrence of the ante-
cedent event ("the agent does so-and-so") is avoidable. But
the philosopher who accepts (5) may well insist that, since
the antecedent clause refers to an act or a choice, the use of
"avoidable" in question is not a misuse.

What, then, if we were to deny (3)? Suppose that some
of our choices do not satisfy (3) — that when they are made
they are not made under any conditions such that it is (caus-
ally) impossible (though logically possible) for them not to
be made. If there are choices of this sort, then they are
merely fortuitous or capricious. And if they are merely for-
tuitous or capricious, if they "just happen," then, I think, we
may say with Blanshard that we are not morally responsible
for them. Hence denying (3) is not the way to avoid (5).

We seem confronted, then, with a dilemma: either our
choices have sufficient causal conditions or they do not; if
they do have sufficient causal conditions they are not avoid-
able; if they do not, they are fortuitous or capricious; and
therefore, since our choices are either unavoidable or fortui-
tous, we are not morally responsible for them.

There are philosophers who believe that by denying the
rather strange-sounding premise (4) we can escape the dilem-
ma. Insisting on something like "the primacy of practical rea-
son," they would say that since we are certain that (5) is false
we must construct a metaphysical theory about the self, a
theory denying (4) and enabling us to reconcile (3) and the

Responsibility and Avoidability / 159

denial of (5). I say "metaphysical" because it seems to be
necessary for the theory to replace (4) by sentences using
such terms as "active power," "the autonomy of the will,"
"prime mover," or "higher levels of causality" — terms desig-
nating something to which we apparently need not refer when
expressing the conclusions of physics and the natural sciences.
But I believe we cannot know whether such theories enable us
to escape our dilemma. For it seems impossible to conceive
what the relation is that, according to these theories, holds
between the "will," "self," "mover," or "active power," on the
one hand, and the bodily events this power is supposed to con-
trol, on the other — the relation between the "activities" of
the self and the events described by physics.

I am dissatisfied, then, with what philosophers have pro-
posed as alternatives to premises (1) through (4) above, but
since I feel certain that (5) is false I also feel certain that at
least one of the premises is false.

Chapter 5

Determinism, Freedom, and
Responsibility

C. J. Ducasse, Brown University

Several speakers at this conference appeared to take it for
granted that determinism and freedom are incompatible, and
hence that the questions in need of being answered were only,
first, whether, or how far and where in particular, determin-
ism or freedom in fact obtains; and second, what bearing
various answers to the first question would have on practical
issues in philosophy, science, law, and ethics.

Underlying this conception of the points at issue is a tacit
assumption that the concepts determinism, freedom, indeter-
minism, and contingency are quite clear, or at least clear
enough to make possible definite answers to the questions
mentioned. I believe on the contrary that this assumption is
largely mistaken, and that the inconclusiveness of the present
discussions, as well as of innumerable other discussions of
the same questions in the past, has been due to the fact that
they were engaged in without adequate preliminary analysis
of the concepts employed in them. The remarks to follow
will therefore attempt to distinguish among several of the
senses in which the key terms mentioned are often used; to
clarify each of those senses; and then to indicate what does or
does not follow as regards some of the issues in discussions of
which those terms are commonly employed.

1. Determinism as theoretically universal predictability. In
science, and also in certain other contexts, determinism is
employed to mean theoretically universal predictability; that
is, it is used to signify that, on the basis of knowledge of (a)
the state of the world at any given time and (b) the laws ac-
cording to which its state at any time is related to its states
at other times, it would be possible to infer what the state of
the world was, or will be, at any earlier or later time. Lap-
lace's famous statement formulates a determinism so con-
ceived:

An intelligence knowing, at a given instant of time, all forces
acting in nature, as well as the momentary positions of all

160

Determinism, Freedom and Responsibility / 161

things of which the universe consists, would be able to compre-
hend the motions of the largest bodies of the world and those
of the smallest atoms in one single formula, provided it were
sufficiently powerful to subject all data to analysis. To it, noth-
ing would be uncertain, both future and past would be present
before its eyes.^

2. Critique of Laplacian determinism. This account of de-
terminism, however, is open to fatal criticism on several
grounds.

a. For one thing, it assumes that the physical world is the
whole of the world; and this leaves out of account mental
events in general and volitions in particular, unless it defines
them as themselves purely physical events — e.g., as molecular
events in the tissues of the brain. But so to define them is not
legitimate, since it amounts to asserting that the term "mental
events" does not denote by means of it, but denotes instead
certain quite different events — such a contention being as
paradoxical as would be the parallel one that what we intend
to denote — i.e., to point at — when using the word "cabbages"
is not cabbages but, say, tigers. Of course, it might conceiv-
ably be true that all mental events are dependent on bodily
events of some sort. But to be "dependent on" and to be
"identically the same as" are two different relations; the first,
being at least dyadic, precludes the second, since it is mo-
nadic. Hence the physical world is not the whole world.

b. But further, determinism as conceived by Laplace as-
sumes that observation would yield precise knowledge of the
state of the physical world at the time; and this, since the days
of Laplace, has been disproved. It has been shown that to
observe both the position and the velocity of a particle at a
given time is not simply difficult but inherently impossible,
because to attempt it is automatically to alter one or the
other.

c. Again, Laplacian determinism asstunes that the only
forces at work in the world are those of classical mechanics.
But this is not known to be true, and it is dubious in particu-
lar in the case of biological processes, and still more so in
the case of consciously purposive action,^

In addition, the determinism of Laplace ignores the possi-
bility that some events or entities are wholly or in part sui

^Theorte analytique des probabilit6s (3rd ed.; Paris. 1820).

2 On this point interesting material is to be foimd in an address by H. S.
Jennings on "Some Implications of Emergent Evolution," Science, January
14, 1927; in E. S. Russell's The Directiveness of Organic Activities (1946);
and in recent writings of a number of other biologists.

162 / Determinism and Freedom

generis — a possibility that would make the notion of laws
governing their occurrence incongruous, since laws obtain
with regard to events or entities only in so far as these are
instances of a kind, but not in so far as they are individually
unique. Indeed no event or entity is completely similar to any
other; no matter how great may be the similarity of one to
another, its being "other" means at least that its spatial and/ or
temporal relations are somewhat different; and this entails that
an individual residuum, unpredictable because unprecedented,
is an ultimate constituent of every occurring event or existent
thing.

d. The upshot, then, is that determinism in the sense of
theoretically universal predictability, whether of all physical
events only, or of events of other kinds, too — for instance,
of mental events in general and of volitions or decisions in
particular — is not only not known to be true, but much rather
is known to be false. Hence the status of determinism in the
sense of theoretically universal predictability is only that of a
pious but bigoted article of scientistic faith.

3. Freedom and determinism conceived as theoretically
universal predictability. What, now, is entailed as regards free-
dom by the indefensibility of determinism conceived as
theoretically universal predictability? Consider a given event
— say, the choice a given person makes on a given occasion
among the alternatives open to his choice. Obviously, the fact
that the choice he makes was not certainly predictable — even
perhaps by himself — does not entail that it was "free" in any
sense other than "free from the possibility of being pre-
dicted"! In particular, it does not entail that the choice he
made was uncaused in any sense of "uncaused" other than
that of "unpredictable"; nor does it entail that the choice was
made "freely" rather than perhaps under duress, threat, or
pressure.

4. Determinism conceived as fatalism. Certain events, e.g.,
eclipses, tides, earthquakes, volcanic eruptions, etc., are be-
yond present or prospective control by man's will. Others,
such as the movements of his limbs, and also many effects
that these cause directly or indirectly under specifiable condi-
tions, are within man's control when those conditions obtain.
Fatalism, however, in effect contends that even under those
conditions man's volitions are not inherently efficacious to
their intent, but that their efficacy or inefficacy is preordained,
permitted, or perhaps implemented, as man's clumsiness or
resourcefulness may render necessary, by some mysterious

Determinism, Freedom and Responsibility / 163

purposeful agency called Fate, the Gods, or Destiny; and that,
until its decrees are fulfilled, men — other than perhaps cer-
tain prophets — do not and cannot know what "Fate" had
decreed.

No evidence, however, that would tend to show fatalism to
be true is available, for when the events man attempts to
cause is of a kind he is capable of causing at will under cir-
cumstances of certain kinds known to him, his failure to cause
it in a given case is always explicable as resulting simply from
his not having known that the circumstances then existing did
in fact differ in some essential respect from those under which
his action would have been efficacious to its intent. Fatalism
thus arises only out of man's naive tendency to assume, ani-
mistically, that causes which thwart or unexpectedly promote
his purposes must themselves be purposive!

5. Determinism as universality of causation. The thesis of
determinism conceived as universality of causation is that
every event that occurs has some cause and has some effect.
Whether this thesis is true or false, and what is truth or
falsity entails as to man's freedom depends on the nature of
the causality relation.

a. Hvmie offers two definitions of causality — one objective
and the other in part subjective. According to both our judg-
ments that certain sequences of events are causal sequences
mean that in our experience those sequences have been
constant, i.e., regular. Hume insists, however, both after and
before stating his two definitions, that both of them are
"drawn from circumstances foreign to the cause . . . from
something extraneous and foreign to it," But although he
acknowledges that this is inconvenient, he declares that it
cannot be remedied.

If, however, as some writers since Hume have done, one
takes experienced regularity of sequence to be all that caus-
ality consists in, then universality of causation would mean
that every event of kind E we have experienced was in our
experience preceded by an event of kind C and followed by
one of kind F. And this, of course, has not in fact been the
case. Moreover, even if it had been the case, it would entail
nothing about events of a kind that we have experienced no
instances of; nor even about instances we have not experi-
enced, of a kind that we have experienced some (other)
instances of.

Anyway, the definition of causality simply as empirical
regularity of sequence would, as Thomas Reid and others

164 / Detcrminisin and Freedom

have pointed out, require us to pronounce causal certain se-
quences in our experience that are regular, but that we con-
fidently deny to be causal. On the other hand, we sometimes
pronounce a sequence causal without waiting to observe
whether or not it is constant. That is, we pronounce it causal
on grounds other than constancy, which, according to Hume's
definition, alone constitutes causaUty.

Thus the regularity of a sequence never in itself answers,
but on the contrary always raises, the question whether the
sequence is a causal one. That a given sequence is causal en-
tails that repetition of its first term and of its circumstances
will be regularly followed by repetition of its second term;
but regularity of a sequence does not entail, but only sug-
gests, that it is perhaps a causal one. To decide whether or
not it is indeed causal, we need have available, and in fact
use, a different definition of causality.

b. Causality is the relation that obtains between the three
factors of a perfect experiment; i.e., between a given state of
affairs S and only two changes (whether simple or complex)
in it — one a change C at a time Tl, and the other a change E
at an immediately sequent time T2. If this relation obtains
among S, C, and E, then C is, by definition, the proximate
cause of E under the circumstances S. The causality relation
is thus not dyadic only, but irreducibly triadic: the circum-
stances S cannot be regarded as a part of the cause C, be-
cause C consists of a change occurring in S.

c. That change C in 5 was the proximate cause of change
£ in 5 does not presuppose that S, C, and E ever occurred
before or ever will occur again, but only that C and E were
the only two changes in S; but it entails that, should S, and C
in S, ever have occurred before or ever occur again, then E
in S did, or will, follow in every such case. For in the defini-
tion of etiological sufficiency of C in 5 to E in 5, i.e., of caus-
ation of the latter by the former, no particular date is
specified, but only posteriority of the time of E to the time of
C Moreover, what occurs in S upon occurrence of C in S is
an intrinsic element of the nature of S; hence to suppose that
S and that C in 5 recur, but that £ in 5 docs not then recur,
is to suppose contradictorily that in the second case the state
of affairs in question is, to that extent at least, different from
S.

The fact that causality is defined in terms of a single case of
sequence (of the type specified above) entails that causal laws
are causal not because they are laws (since some laws, to

Determinism, Freedom and Responsibility / 165

wit, some empirical regularities, are not causal) but because
they are generalizations from cases each of which was in its
own individual right a case of causation; if two or more se-
quences of the type specified resemble one another in that,
and only in that, in each of them the state of affairs is of a
certain kind S, the cause of a certain kind C, and the effect
of a certain kind E, then the "method of single agreement,"
employed as a method of generalization by abstraction, war-
rants the generalization that in any state of affairs of kind S,
an only change of kind C immediately causes a single change
of kind E.

d. The foregoing analysis of causality makes evident that
the canon of the so-called "method of single difference" is a
description of the causality relation itself, not of a relation
other than causality, constituting only a sign of the presence
of causality. That is, the "method of single difference" is a
method only in the strained sense in which the description or
photograph of a person can be said to be a method by which
to identify him if one happens to meet him. Causality is, and
is nothing but, the relation between S, C, and E described in
what precedes. It constitutes etiological sufficiency of C in 5
to E in S, and conversely, etiological necessitation of £ in 5
by C in 5.

e. From this analysis of causality it follows analytically
that every event has a cause and an effect. For, given any state
of affairs S and any change E that is at its time the only
change in S, there is always some immediately anterior change
in S that qualifies as cause of £ in 5, i.e., that is the only
change in S at its time. This follows from the fact that without
some change there is no time; and hence that to suppose
either that S endured or that S changed prior to E is to sup-
pose that some sort of "clock" was "ticking" then. And should
no other change than the "ticking" of that "clock" have oc-
curred in S prior to E, then the "ticking" itself would qualify
as cause of E under the definition.

For a corresponding reason any change C in any S that is
the only change in S at its time causes some effect E in S.
Also, the specification that C and E be the only two changes
in S makes it superfluous to specify that C and E are con-
tiguous in time. Their contiguity is entailed by the specifica-
tion that they are the only two changes in S, since, if time
elapsed, i.e., if any "clock" were "ticking," between C and E,
its "ticks" would constitute changes in S additional to C and
E.

166 / Determinism and Freedom

/. To these remarks it should be added that in the case of
causality as in that of sincerity or divinity or gravitation, etc.,
to define its nature is one task; and to decide whether some-
thing concrete — in the case of causality, a concrete relation
that we observe — is really an instance of what the definition
specifies is quite another task and one that, theoretically, is
never performed with complete certainty. The acceptability or
nonacceptabUity of the decision in a given case turns on
pragmatic considerations, just as the answers to the question
whether the weight of a given book is exactly the same as, or
is a trifle more or a trifle less than, that of the standard pound
— i.e., of the piece of metal whose weight is 1 lb. by defini-
tion— turns on the purpose that governs at the time. If one's
purpose is to maU the book, then the scale at the post oflBce,
and the clerk's reading of what it marks, are authoritative.

On the other hand, if in the attempt to identify empirically
instances of what a definition defines, all pragmatic concerns
are put aside, then one automatically rules out all possibility
of success; for some pragmatic concern is the only thing that
gives any empirical meaning to the question whether a pro-
posed operation of identification and a proposed criterion of
the success of the operation are absurd or on the contrary
rational.

.' g. The conclusion, then, of the present section is that de-
terminism, in the sense that every event has some cause and
some effect, is analytically true. This, however, does not at
all entail that every event is completely and certainly pre-
dictable even theoretically, i.e., even on the supposition of
exact and exhaustive knowledge of the past history of the
universe; for complete and certain prediction would be pos-
sible only on the basis of strict sameness of present and past
data, whereas what we actually get is only similarity in vary-
ing degrees. At all times in the history of the world, an
element of novelty, whether great or small, is present and
accounts for what has been called the emergence of such
novelties as "life" and "mind"; which, because novel, i.e.,
unprecedented, were inherently unpredictable.

To designate the sense of "determinism" considered in this
section, which entails that universality of causation is an-
alytically true and yet which has room for the occurrence of
novel and therefore unpredictable "emergents," we may adopt
the name by which the late biologist H. S. Jennings referred
to his own conception of determinism, to wit, "radically
experimental determinism.'^,,

Determinism, Freedom and Responsibility / 167

6. Freedom. Freedom is more often conceived negatively,
i.e., in terms of indeterminism or of exception to determin-
ism, than in terms of a positive analysis. The same is true of
the terms chance and contingency.

a. What has been said up to this point will have made clear
that to say of a given event — whether physical or mental —
that it was "free," or more or less synonj^nously, that it was
"contingent" or "a matter of chance," may mean different
things. For example, it may mean (1) that the event was
practically unpredictable, i.e., unpredictable on the basis of
the data we had or could get at the time; or (2) that it was
unpredictable even theoretically, for one or another of the
reasons mentioned in our critique of Laplace-type determin-
ism; or (3) that it had no cause; or (4) that it was not
logically necessary, i.e., that supposition of the event's non-
occurrence, or of its having been more or less different from
what it was, implies no contradiction. And of course, so long
as these diverse meanings of the term are not distinguished
in discussions of "freedom," no possibility exists of really
establishing anything.

Anyway, none of these four senses seem to correspond to
the meaning "free" is intended to have in the instances people
commonly offer when asked to give examples of the exercise
of that "free wUl" they believe they have; for they usually
offer illustrations such as that, at the moment, they are free
to raise their arm, or not, as they will; or that, when offered a
choice of, say, apple pie or cherry pie, they are free to choose
the one they prefer. An analysis of the meaning of "freedom"
as used on these and similar occasions is what I now submit.

b. That a person P, under circumstances K, "can," or
synonymously "is free to," do A or not to do it means that
volition (or, pace Professor Ryle, "decision") by P io do A
or not do A, would, in those circumstances, be sufficient to
cause A to occur or not to occur. And if it should be objected
that this is only freedom to act, but not freedom to will, then
the reply is that that analysis applies whether ^4 is a bodily
act or itself an act of will; that is, it applies irrespective of
whether P's present decision is, say, to raise his arm, or
whether his present decision is, perhaps, to decide tomorrow
whether to buy a house or rent one. ,

c. That under the circumstances that obtained P was, in
the sense just stated, free to do or not do A does not presup-
pose that his decision had no motive or other cause; nor does
the fact that neither he nor anyone else could have on that

168 / Determinism and Freedom

occasion predicted what his decision would be constitute any
evidence that it had no cause.

d. Nor does the fact that his decision, like any other event,
had some cause — causation, as defined in Section 5b, is uni-
versal— mean that his decision was not free but compelled;
for a decision one is caused to make is not describable as
"compelled" unless it is a decision to do something to which
one is averse, even if less averse than to the alternatives then
open to choice. The decision to hand over one's money to the
holdup man who confronts one with the choice "Your money
or your life" is not made freely but under compulsion;
whereas the decision to choose, say, apple pie rather than
cherry pie in the restaurant is — assuming that one likes apple
pie — an example of what is called a free decision.
,- Thus the assumption that tacitly underlies most discussions
of "free will" and determinism — the assumption, namely, of
. incompatibility between, on the one hand, determinism in
' the sense of universality of causation and, on the other, free-
dom in the sense in which man certainly does have freedom
in many cases — is an altogether erroneous assumption. It
arises from failure to analyze the meaning that the terms in
question have in the specific contexts in which they are em-
ployed.

7. Determinism and moral responsibility. The final ques-
tion to consider is what determinism, in the sense that every
event has some cause and some effect, entails concerning
moral responsibility or lack of it.

a. It is essential here to distinguish between moral responsi-
bility and the legal responsibility that in practice may or may
not go with it. Also, to distinguish between either moral or
legal responsibility and merely etiological responsibility, which
consists simply in the fact that our acts, besides the effects
they aim at, have many others of which we are not aware,
and which may or may not harmonize with our intention.

b. That a person P is now morally responsible for his vol-
untary acts — or was not so but can now be made morally
responsible for his future ones — means simply that to praise
or blame him or otherwise reward or punish him for some
thing he now does or did will tend to cause him to act, or
tend to inhibit him from acting, in a similar manner on simi-
lar future occasions.

c. Hence, that a person P is not now morally responsible
for what he voluntarily does or did means that praise or
blame, or other forms of reward or pxmishment, would have

Determinism, Freedom and Responsibility / 169

no such effect on him. In others words, it means that he is
not now capable of moral education, and hence that what he
needs is psychiatric treatment. In such a case to inflict punish-
ment for a morally wrong act is either simply stupid or
sadistic or both.

d. This analysis of what constitutes "moral" responsibility
entails that such responsibility not only is not incompatible
with determinism, but on the contrary presupposes determin-
ism. In other words, that an agent is morally responsible pre-
supposes that an awareness on his part that something he
contemplates doing would be morally wrong or, as the case
may be, morally right (his notion of what constitutes moral
lightness or wrongness is irrelevant here) will, other motives
being equal, cause him to refrain from acting, or on the con-
trary cause him to act. Without such causation there is no
moral responsibility.

©r 6

Some Hefiecfaons en "The Case
for Determinism"

Carl G. Hempel, PriBceton University

1. On defining determinism. There appears to be a dis-
crepancy between two characterizations of determinism that
Professor Blanshard offers at the beginning of his admirably
lucid and stimulating paper. In the third paragraph determin-
ism is defined as the view that every event A has at least one
temporally later necessary consequent B. According to this
view, then, the cause ^ is a sufficient condition of the effect B;
and determinism asserts that every event causes some (later)
event. But clearly this assertion is by no means equivalent to
the thesis that every event is caused by some (earlier) event,
or briefly that "ail events are caused" — a. formulation by
which Mr. Blanshard characterizes determinism in the second
paragraph of his essay.

Evidently it is the latter assertion that gives significance to
the problem of free choice, the central topic in Mr. Blan-
shard's discussion. For the question at issue is whether all
events, including human acts of choice, are caused by ante-
cedent events — not whether they in turn have certain neces-
sary consequents. And indeed a deterministic thesis to the
effect that every event causes some later events would be
quite compatible with the "uncaused" occurrence of various
phenomena, such as, perhaps, the spontaneous creation of
(hydrogen atoms envisaged by some contemporary cosmolo-
\ gists. Mr. Blanshard's thesis of determinism will therefore be
construed here as asserting that for every event B there exists
'an antecedent event A that is a sufficient condition for the
ipccurrence of B.

Mr. Blanshard uses the phrase "given A, B must occur"
as a general characterization of the relation that obtains be-
tween an event A and a later event B ii A is a sufficient con-
dition, or cause, of B. As to the meaning of that "must," he
is willing to let us make our own choice among a logical
"must," a physical or metaphysical one, and a "must" that
simply means universal factual association. In the brief space

170

Some Reflections on "The Case for Determinism" / 171

of his paper Mr. Blanshard certainly could not be expected
to give a detailed analysis of the concept in question — this
would have made it necessary, for example, to tackle the
hornets' nest of problems surrounding the counterfactual
conditional. But I think there is reason to feel uneasy at the
sweeping claims of a determinism that is said to hold true
equally for all the different interpretations of "must" that Mr.
Blanshard allows us here.

Choosing the logical "must," for example, we obtain the
thesis that for any event B occurring at some time t^, there is
an event A at an earlier time i^, such that the occurrence of
A at t^ is a logically sufficient condition for the occurrence of
B at tj. But this is trivially true in all cases. For example, let
the occurrence of B at t,^ consist in the onset of rain on the
campus of Yale University at noon of a certain day, and let t^
be 11:00 a.m. of the same day. Then, to show that the deter-
ministic thesis with the logical "must" is satisfied, it suffices
to choose for A the state of affairs that consists in the condi-
tion of the Yale campus exactly one hour before the onset of
a rainfall. To be sure, we may not be able to ascertain that A
prevailed at /„ until the occurrence of B zi t^, but this is irrele-
vant to the point to be proved, namely that there is in fact a
state of affairs A at t^ that is logically sufficient for the
occurrence of B at t^.

The preceding argument shows, I think, that the determin-
istic thesis with its "must" construed as expressing logical
necessity is of no significance for the problem of free choice:
the existence of an earlier state of affairs that is a logically
sufficient condition for a given act of choice surely does not
mean that the act is determined in any sense that would cast
doubt on the freedom of choice.

I am not clear what a metaphysical construction of "must"
might come to, but I should like to add a few words con-
cerning the interpretation of "must" by reference to physical
— or, more generally, empirical — laws. On this interpretation
Mr. Blanshard's deterministic thesis asserts that there exists
a set of laws such that every event 5 is a consequent, accord-
ing to those laws, of some preceding event A. But this state-
ment, again, is always true in a trivial manner. For no matter
what the sequence of events in the universe may be, it can
always be represented as a mathematical function of time;
and this function provides the requisite law. Suppose, for
example, that we are interested only in the changes of tem-
perature and color that occurred in the course of the history

172 / Determinism and Freedom

of the universe at a given place. The temperature at any time
may be measured on the Kelvin scale and represented by a
real number; the color can be similarly expressed by means
of a number indicating a location in the color scale. The tem-
poral succession of temperatures and colors at the given place
now corresponds to an assignment to each value of the time
variable of one temperature nimiber and one color number.
This determines two mathematical functions, which trivially
furnish laws governing the chages of temperature and color at
the given place. The argument can now be extended to apply
to the changes anj^where in the universe of temperature,
color, and any other characteristic. Thus the course of the
universe is governed by functional laws (which may, however,
be so complex as to be beyond the reach of scientific dis-
covery) , and the deterministic thesis with its "must" construed
in terms of empirical laws is true.^ In fact, the laws here
referred to are so strong that every event is determined by
the laws alone, without the need of recourse to antecedent
occurrences.

The argument here outlined shows that determinism in the
form under discussion is trivially true; thus it can have no
greater significance for the freedom of choice than the truth
of the deterministic thesis with its "must" construed as repre-
senting logical necessity.

Nor can the truth of determinism on the basis of this con-
struction of "must" cheer the empirical scientist, for he is
interested in the possibility of prediction; and this requires
the determination by law of events by eariler ones in a
stronger sense than that so far considered: the laws in ques-
tion have to be of a sufficiently simple kind to permit of
discovery and subsequent predictive application by himian
beings. (Causal laws in the technical sense of physics are

^ The basic idea of this argument is set forth by Bertrand Russell in his
essay "On the Notion of Cause," which is reprinted in his book Mysticism
and Logic (London, 1921). A similar consideration is presented in Moritz
Schlick's essay "Causality in Everyday Life and in Recent Science" (reprinted
in H. Feigl and W. Sellars, eds.. Readings in Philosophical Analysis [New
York, 1949], pp. 515£E.) Russell's idea is discussed also in Philip Frank's
Philosophy of Science (New York: Prentice-Hall, 1957), which, in chaps,
xi and xii, presents an illuminating analysis of the concepts of causality and
determinism.

The argument presented above calls for one qualifying remark, however.
Not aU functions of time can be represented by analytic mathematical expres-
sions, nor even by symbolic expressions of finite length in a language using a
finite or denumerably infinite set of different basic symbols. Thus, not all
possible laws can be expressed in the language of science, just as not all real
numbers can be represented by numerical expressions.

Some Reflertions on "The Case for Determinism" / 173

required to satisfy certain conditions as to mathematical
form, 2 but these need not be considered here.)

On this interpretation of "must" Mr. Blanshard's thesis is
by no means trivial. In fact, despite the vagueness introduced
by the requirement of formal simplicity for the laws, it makes
an extremely strong and sweeping assertion. And I would sug-
gest that the objections to determinism that Mr. Blanshard
examines in his paper be construed as concerning this strong
thesis. (In fact, the objections based on the character of
quantum physics refer to the even stronger thesis of universal
causality in the technical sense of physics.)

2. On the relevance of introspective evidence to determin-
ism. As for the first objection to determinism, which refers to
a stubborn feeling of freedom of choice, I fully agree with
Mr. Blanshard that it cannot count as evidence against deter-
minism, for this kind of feeling can surely be deceptive. In-
deed I think that the feeling is irrelevant to the question of
causal determination. For in order to decide whether a given
act of choice is causally determined we have to judge whether
there is an antecedent event with which the choice is con-
nected by a general law of simple form. And surely the data
obtainable by introspection, especially the "stubborn feeling
of freedom," have no bearing on this question. The timid man
in a hypnotist's audience, for example, who gets up to make
a speech, may truthfully protest a feeling of complete free-
dom in choosing to do so: this is quite compatible with the
possibility that his choice was causally determined (via gen-
eral laws concerning the effects of hypnosis) by the instruc-
tions he received earlier under hypnosis.

But it seems to me that in his defense of determinism Mr.
Blanshard makes use of the very same kind of argument that
he rightly rejects when it is employed by indeterminists. The
use of introspection illustrated by Galton's case surely is no
more suited to establishing the existence of antecedent con-
straints determining choices according to general laws than
the feeling of freedom invoked by the indeterminists is suited
to establishing the opposite. A person's introspective reflection
on the motives that might have determined a certain action
of his can yield quite deceptive results; surely it is not suited
to establishing the existence of general laws linking the ad-
duced motives to the action under scrutiny.

' For a discussion of these conditions, see, for example, H. Margenau,
The Nature of Physical Reality (New York, 1950), chap, xix, and Frank,
op. clt., chaps, xi and xil.

174 / Determinism and Freedom

3. On levels of causation. In his outline of a nonmechani-
cal kind of determinism for psychological occurrences, Mr.
Blanshard refers to aesthetic, logical, and moral constraints as
determining factors. Now, it would be incompatible with his
view of causation to construe these constraints as abstract
aesthetic ideals, logical truths, and moral principles; for these
lack the temporal character that Mr. Blanshard's definition of
determinism requires of all causal factors. To meet this re-
quirement we shall have to construe the constraints in ques-
tion as awareness and espousal by a given human agent of
the ideas or ideals in question. In explaining the work of a
given painter by reference to certain aesthetic ideals, for
example, we have to construe the constraining antecedent
factor as the disposition of the painter to conform to those
ideals.

There is another consideration that lends support to this
way of looking at the matter. The attribution of a causal role
to timeless ideals and relations not only violates Mr. Blan-
shard's initial definition of causal determinism: it seem to me
an inherently obscure idea. Surely Mr. Blanshard must be
using a metaphor — and I think a misleading one — when he
speaks of an abstract, nontemporal ideal as getting hold of
an artist and molding his work, or of a timeless relation serv-
ing as the condition of a temporal passage. Those timeless
entities cannot be held to exert a xmiversal causal influence,
or else there would be no logical or mathematical errors, no
immoral acts. And indeed Mr. Blanshard notes explicitly that
man is not always guided by the proper ideals. Thus how a
given person is going to act will be determined, not by an
ideal standard pertinent to the action, but rather by whether
or not the agent has a certain disposition, namely that of
acting in accordance with the standard. For example, so far
as honesty can be said to have been a causal determinant of
George Washington's confessing about the cherry tree, it
suffices to describe this factor as Washington's disposition to
act honestly, and it would be an unnecesary, and indeed very
risky, hypostatization to attribute his confession to the super-
venience, from a higher causal level, of the timeless moral
ideal of honesty.

Yet this avoidance of an appeal to causal determinants of
a higher order involves no "mechanistic" or "materialistic"
assumptions: in particular, it does not presuppose the reduci-
bility of psychology to physics and chemistry; that is, the
possibility of fully describing all psychological events in

Some Reflections on "The Case for Determinism" / 175

physicochemical terms or the possibility of explaining them
by means of physicochemical laws.

Nor, of course, does the viewpoint here suggested imply
that all psychological phenomena are causally determined
either by dispositions of the kind just mentioned or by other
factors. In fact, as was pointed out at the end of Section 1,
above, the thesis of universal determinism is inherently vague;
it is not as clear and precise an assertion as, say, a law of
physics. At the same time, the thesis makes a tremendously
stronger claim than a physical law, for it asserts the existence
of a comprehensive set of laws suflScient to determine every
event in the world of our experience. And the extent to which
this claim is correct cannot be discovered by philosophical
reasoning alone, nor by reference to the crude data provided
by everyday experience or introspection: it has to be ascer-
tained by means of rigorous and extensive scientific research.

Chapter 7

Some Further Reflections on
Moral Responsibility

Howard W. Hintz, Brooklyn College

Professor Hospers' teiesis (like Professor Edwards') denies
the existence of moral responsibility in any sense in which an
/ individual can be held finally accountable or answerable for
his acts. Thus, by the analysis of the term and concept of
responsibility attempted by Edwards and Hospers, "responsi-
bility" is relegated to the status of a pragmatic device by
which a person may by some sort of persuasion, admonition,
threat, punishment, or other social pressure be induced to
7^ pursue a course of behavior consonant with the norms ap-
proved and established by a given culture group.

What I want particularly to do in this brief criticism is not
categorically to deny the validity of the Edwards-Hospers
thesis but to point out what I believe its necessary corollaries
and implications as far as ethical considerations are con-
cerned. Thus I do not see how we can escape the conclusion
that this thesis destroys the foundations of all prescriptive
ethics except on the arbitrary-power level. If, as suggested
above, the possibility of establishing moral values and stand-
ards is removed, then the basis of meaningful and purposeful
living, of human dignity, and ultimately of civilized society
itself is undermined. It should be further noted that the thesis
is not only deterministic (despite Hospers' denial of the rele-
vance of the determinism issue) but fatalistic, no matter how
much the fact may be disguised.

Neither the Edwards-Hospers' analysis nor any other deter-
ministic explanation has yet succeded in refuting the fact that
aU individuals above the idiot level possess in some sense,
however limited by internal or external compulsions, the
power of choice, the power to decide and to select among a
given set of alternatives, or, to follow G. E. Moore, the
ability to say meaningfully, after certain types of choices have
been made, "I could have chosen or acted differently." The
inescapable truth of this is repeatedly demonstrated by the
fact that on reflection and a subsequent awareness of the con-
sequences of a choice, all of us, when a similar set of altema-

176

Some Further Reflections on Moral Responsibility / 177

tives is presented to us a second or third time, choose dififer-
ently because we have been dissatisfied with the consequences
of the former choice.

The attempt to invalidate freedom and power of choice on
the grounds that choices are attributable to underlying desires,
to previous conditioning, and to basic qualities of character
is, it seems to me, highly factitious. It is saying nothing more
than that a cause and effect relationship prevails in the phe-
nomena of human behavior as it does in other natural
phenomena; it is simply to assert the principle of causality.
But rather than vitiating the meaningfulness of discriminating
choices and acts of will, causality itself is a primary condition
of meaningful choice.

Is there any possible way — within the framework of a
causal determinism rooted in the obvious psychological and
naturalistic facts of human experience as outlined by Edwards
and Hospers — in which we can retrieve a modicum of indi-
vidual moral responsibility and hence of human dignity? I
believe it is not only possible but essential that we do so by
recourse to the familiar principle (constantly stressed by
John Dewey) that each man is responsible for making the
best choice available to hun within the scope of his limitations
and his powers. That these limitations and powers differ
widely among men no one will deny. But to the extent that an
individual acts or fails to act responsibly within the range of
his capacities, whatever they may be, to that extent he is
praiseworthy or blameworthy. It is only because this principle
is universally recognized in everyday experience and practice
that any tj^e of social order is even possible.
Two additional points should be emphasized:

1. Hospers' thesis not only destroys moral responsibility: it'
also destroys all rationality in human experience. If we ac-
cept the proposition that a man acts only as previous deter-
mining and conditioning factors compel him to act, then I
submit that the function and office of reasoning have been
utterly destroyed, for the very question whether a man rea-
sons or not is then completely dependent on the allegedly for-
tuitous and contingent factors that shaped his nature and
character. His reasoning or nonreasoning then becomes pure ;
accident and is therefore, by definition, removed from the
area of rational choice. I do not see how Hospers can escape
the logical necessity of this conclusion. ^

2. However emphatically Hospers may deny the relevance
to his argument of the determinism-indeterminism issue, his

178 / Determinism and Freedom

thesis is not only deterministic: it is wholly fatalistic. My quar-
rel is not with determinism — at least in some of its forms —
but with Hospers' denial of the relevance of his thesis to the
determinism issue. How can the logic of Hospers' argument
possibly escape the ultimate assertion of fatalism? If every-
thing that a man chooses or decides or does — including the
decisions he makes after reflection and deliberation — is en-
tirely the result of the conditioning factors of his heredity and
his environment, then what conceivable area of what we have
traditionally called moral choice has not been predetermined?
Wherein, then, does Hospers' position differ fundamentally
from that of Jonathan Edwards, who maintained that a man
may choose what he wills to do but that he has no choice
about what he wills to choose? In his own explication of pre-
determinism and predestination Jonathan Edwards was, as has
long been recognized, strictly logical, and accepted, however
reluctantly, all the implications of his logical consistency. Is
Hospers, I wonder, willing to do the same? Jonathan Edwards
finally was driven by his own rigorous logic to concede that
free will was an illusion. Is Hospers, with equal logical con-
sistency, willing to admit the same?

Neither can I recognize the validity of the distinction Hos-
pers makes between the so-called upper and lower levels of
responsibility — which presumably is a distinction between de-
sires on the one hand and specific choices and actions on the
other. On the upper level, says Hospers, will, choice, and re-
sponsibihty are meaningful; on the lower level of desires they
are not. But if my underlying desires are the wellspring of my
volitions and my actions, how can the upper level of respon-
sibility be disengaged from the lower level? In point of fact the
two levels are constantly interacting — so that the choices I
make and the acts I perform today on an upper-level volitional
basis affect in large measure the desires I shall have tomorrow.
At this juncture the two levels would seem to merge indistin-
guishably. The individual choices of the present are not only
causative agents of future events and future acts: they are
also generating and modifying agents of new desires and al-
tered character structures.

It is now widely recognized among moral philosophers that,
no matter how irrefutable the logic of absolute psychological
or naturalistic determinism may be, sane and rational human
beings in order to retain their sanity, their rationality, and
their purposefulness in living still stubbornly insist on decid-
ing, choosing, and acting as though they were autonomous,

Some Further Reflections on Moral Responsibility / 179

dignified, and free individuals. This is the phenomenon, above
all, that still needs to be explained, and — therein lies the crux
of my argument — the explanation is not to be found in the
Edwards-Hospers thesis. I suspect that the most disturbing
and serious weakness of this thesis lies in the fact that it is
based on a groundwork of completely unproved and possibly
unprovable assumptions. If it represents a form of determin-
ism, as I have insisted, it is not a scientific determinism, but a
rationalistic determinism differing in no essential respects
from Calvanistic fatalism as far as its hypothetical assumptions
and practical implications are concerned. Hospers is right in
divorcing his thesis from the conclusion of empiricists like
Hume, Mill, and Schlick. There is indeed nothing empirical
about his position. -^

For the fact is that scientifically and empirically we are at
the present moment far from having a complete knowledge of
the meaning and dimensions of human character structure, of
the complexity of the forces that create, modify, or alter this
structure or, most significantly, of the degree to which the
human organism contains within itself the autonomous power
to alter, to originate, to create, and thus to overcome previous
conditioning. It has yet to be demonstrated, empirically and
scientifically, that acts of will, choices and decisions, and new
conditioning forces may not radically alter character structure
itself.

We might well be reminded at this point of Whitehead's
conception of origination and creativity as the distinguishing
marks of an "actual entity" or of the life principle itself.

Chapter 8

Sentlmentfiiism

Sidney Hook, New York University

If the criterion of a necessary statement is that its denial is
self-contradictory, then none but logical statements or those
declared true in virtue of their form alone are strictly neces-
sary. "Every effect has a cause" is a necessary statement.
"Every event has a cause" and "Every event is a cause of an-
other event" are not necessary statements. If determinism is
^ ' the belief that all events have causes and are themselves causes
■p^ of other events, its denial is not self-contradictory. The only
, evidence we can have for belief in the validity of determinism
ij^ empirical: the success of our predictions.

Nonetheless, "predictability" and "determinism" are not by
any means interchangeable terms although they are related.
Indeterminism entails unpredictability in respect to a char-
acter or event assumed to be undetermined. In a chance world
God might be able to foresee any or all specific, fortuitous
events; but if successful guesses are ruled out as predictions,
men in such a world could make no genuine predictions. Un-
predictability, however, does not entail indeterminism, since
it is compatible with the existence of a theoretically deter-
mined system of such vast complexity that it is beyond human
power to make correct predictions. This raises a problem that
I attempted in vain to get the speakers and participants in the
Institute to address themselves to: viz.. What is the pragmatic
' difference between asserting that a system or state of affairs
^: is undetermined and asserting that the system is so complexly
^l^etermined that no predictions can be reliably made?

If one must choose between these two assertions, it is reason-
able to defend the assertion that the apparently undetermined
system is actually a complexly determined one on heuristic
grounds. If we act on the assumption that a system is deter-
mined, then it is more likely that we shall discover laws and
make successful predictions about the future than if we as-
sume that chance reigns. But can anything more than this be

180

Necessity, Indeterminism, and Sentimentalism / 181

said for the belief in determinism? It does not carry us beyond
Peirce's observation that determinism is a postulate, and a
postulate is something we hope is true. Obviously, to say that
all events are caused, that determinism holds not only in this
or that area, but universally, is always to say more than we
definitely know at any actual time, even if we have a right to
say it — a right that is contested by scientists who assert that
the advance of scientific knowledge no longer depends on it in
every domain. We can definitely reject indeterminism as false
if it asserts that nothing is determined, because we know that
some things are; but if it asserts merely that not everything is
determined, it cannot be rejected out of hand. The issue must
be joined in the field where the sway of determinism is dis-
puted. Operationally, however, the only evidence one can
have for the belief that a state of affairs is determined, as dis-
tinct from the reasonableness of the hope and faith that it is,
is measured by the degree to which we can control, predict,
construct, and, to use Professor Black's phrase, "make things
happen."

If I understand Professor Blanshard, he overstates the case
for determinism by interpreting causality in the realm of cog-
nitive mental events, with some overtones that in the end this
is true for aU events, as a relation of necessity. He assimilates,
although he does not completely identify, the necessities of
aesthetic and moral thinking to logical thinking, which itself is
partly under the constraint of objective logical necessity in
the subject matter thought about. I see no valid reason for in-
terpreting determinism in this way when applied to thinking
in any or all of its forms. According to his conception of
determinism, an event, "thinking," whether it is thinking
about music or mathematics or anything at all, is said to be
determined if some other event, or set of events, is a sufficient
condition of its occurrence. The same is true whether the
event in question is "dreaming" or "hoping," "creating" or
"deducing." It still remains true whether the event is "think-
ing correctly" or "thinking incorrectly." If determinism is
valid as applied to the realm of mind, an incorrect answer to
any question is just as much determined, just as "necessary"
in Professor Blanshard's sense, as a correct answer. This does
not wipe out or call into question in any way the difference
between a correct and an incorrect answer. It indicates only
that what makes the difference between the objectively correct
and objectively incorrect answer, especially if interpreted as
the enduring presence of objective, timeless truths, is irrele-

182 / Determinism and Freedom

vant in answering the question whether, given the mental
event B, whatever its character, true or false etc., there exists
another event, or set of events. A, such that y4 is a sufficient
condition of B. If over and above ^4, it is necessary to invoke
the compulsion of timeless truths to explain B when B con-
sists in the thinking of a correct answer, why is it not neces-
sary to invoke the compulsion of timeless errors or falsities to
explain B when B consists in the thinking of an incorrect
answer?

Empirically I am not at all convinced that the musician,
for example, of whom Professor Blanshard speaks, who adds
one bar to another in the process of composition always does
so because he feels that it is an aesthetically required neces-
sity, and that his feeling so is either a necessary or sufficient
condition for the mental event that consists in the composing
of the bar. He must sound the bar before he can judge or
feel whether it is aesthetically required. If it is determined it
must be determined by a previous event. A musician might
have the feeling of aesthetic necessity for any one or even
a number of quite different ways of completing a musical
phrase, and a critic might say that the musician was mistaken
in his judgment or feeling about each variant. In general,
whether anything is required in a creation or not cannot be
ascertained independently of a purpose, aim, or goal —
whether of artist, critic, or spectator. Once this is given, the
question of what is required permits of an objective and rcla-
j tive solution. We sometimes say that a goal, plan, or purpose
! determines the occurrence of an event. But this is an ellipsis.
It is the thought or desire of the goal, a psychic or physico-
psychic event, that determines, if determinism operates in the
realm of mind.

There is one curious feature in Professor Blanshard's
brilliant essay. He begins by denying that the content of con-
sciousness, "the stubborn feeling" of being free and undeter-
mined, is sufficient to disprove the fact of determinism. But
before he concludes he cites with approval the plain man's
stubborn feeling of absurdity or incredulity at the idea that
the determinism to which he has no objections "as applied by
physicists to atoms, by himself to machines, or by his doctor
to his body," can be applied by anyone "to his reflection and
will." No matter how strong the feeling, it seems to me no
more decisive in assessing the validity of the claim that such
a determinism can be applied to man than in appraising the
claims of indeterminism. As evidence tWs feeling is not to be

Necessity, Indeterminism, and Sentimentalism / 183

ignored or dismissed; but it is less weighty than the "medical
evidence" (I use the expression as a summary term for all
the scientific evidence) that willing and reflecting, like all
other mental processes, are in manifold ways, too numerous to
mention here, dependent on the brain and other conditions
of the human body and human society. Undoubtedly it goes
beyond the evidence to say that all willing and reflecting are
dependent on earlier events in the history of the acculturated
organism that is man. But it is less unreasonable to believe it
than any other alternative. We have more evidence for it to-
day than a hundred years ago. If we have still more evidence
for it a hundred years from now, it will be an even more
reasonable belief, unless there is better evidence for some
other alternative.

n

Normally I should be reluctant to speak of the principle
of indeterminacy in the presence of physicists, but having
recently read what eminent physicists have written about
philosophy I feel absolutely shameless. In addition, the inter-
pretation of the principle of indeterminacy presented by both
Professors Bridgman and Land6 obviously raises no specific
questions in physics but treats of matters, as Professor Munitz
points out, traditionally considered in the philosophy of sci-
ence. I have no "emotional commitment" to the view that
every individual event in the area investigated by quantum
theor^ as well as in the macroscopic world is causally deter-
mined. Nor do I regard the principle of causality as a logical
principle applicable of necessity to everything. Even if it is
considered a presupposition to understanding anything, it
does not follow that everything can or will be understood.
Since we did not create the world we ought to be able to
recover from our surprise at its ways. So far as I can under-
stand, no one challenges the truth of the experimental findings
of the quantum physicist but only his way of talking about
them. When he says that there is a pattern of statistical regu-
larity for the swarm of haphazardly moving photons that
enables us to make predictions about their aggregate behavior,
there is no problem. When he says that the individual photon
is not causally determined and that, by the very nature of
our attempt to investigate and describe it, it cannot be, the
bewilderment begins. In scientific inquiry the language of
common sense is notoriously misleading and must often be
abandoned in the interest of clarity, precision, and fruitful-

184 / Determinism and Freedom

ness. But sometimes there is a needless paradox-and-puzzle-
making quality in the talk of physicists reporting or inter-
preting their findings in common-sense terms. This was very
fashionable in the early years of relativity theory. Assuming
that the idiom of determinism is the language in which com-
mon sense understands the behavior of things, must it neces-
sarily be abandoned at some point in the study of micro-
physical phenomena? I am quite willing to leave this question
to the community of physicists, especially since new scientific
knowledge is being won independently of the language habits
and thought ways of plain men and professional philosophers.
My doubts arise only when the quantum physicist offers an
explanation and, far from insisting on the uniqueness of the
microphysical situation, claims that it is familiar in ordinary
situations in which until now there seemed no need to aban-
don the concepts of causality and determinism.

"It is to be expected," says Professor Bridgman, "that the
roots of the difficulties revealed to us by quantum theory are
already present in the sphere of ordinary life and should be
discoverable by acute enough analysis." With this. Professor
Land6 cordially agrees. The two classes of phenomena in
ordinary life that presumably illustrate in an analogical way
the principle of indeterminacy are (1) games of chance (of
which Land6's game of balls dropped through a chute onto
a knife edge may be considered an ideal case) or games of
insurance in which, although statistical frequencies of mor-
talities are predictable, individual deaths are not, and (2)
measurement that "interferes" with the state of the physical
situation investigated.

1. I venture to suggest that if ordinary games of chance or
insurance are the analogues of the principle of indeter-
minacy, then no special language is required to describe the
behavior of photons; nor is it necessary to abandon the prin-
ciple of causality. In a clear and legitimate sense it is possible
for the insurance actuary to say that, although his knowledge
of statistical regularities does not enable him to predict on
the basis of this kind of data whether and when any indi-
vidual, X, wiU die, it would be the sheerest dogmatism to assert
that no matter what other data were available to a physician
or biologist such a prediction could not be made. For the fact
is that sometimes such predictions are made with remarkable
accuracy, and there is every reason to believe that their ac-
curacy will increase. Similarly, if the values of the different
variables that aflfect the fall of a particular ball were known,

Necessity, Indetenninisin, and Sentimentalism / 185

the laws of classical mechanics could supply a reliable an-
swer as to where a particular ball would fall. It may be
difficult to discover the values of some of these variables, but
the difficulty is not of the kind that makes it impossible to
specify simultaneously the position and velocity of a sub-
atomic particle. For in learning the values of some of the
variables affecting the fall of the ball we are not thereby of
necessity precluded from learning the values of the other
variables, as is allegedly the case with subatomic particles.

Professor Land6 retorts that as far as his illustration of the
falling balls is concerned this type of criticism is only a back-
handed way of recognizing randomness and the absence of
causality by pushing it farther back. Once "irreducible ran-
dom" is accepted, he tells us, it is a comparatively minor
matter whether "(c) each new experiment constitutes a new
game of chance (as quantum theory maintains), or (b)
random was set up once, a long or infinite time ago, and
random distributions observed at present are but the deter-
ministic effects of that one initial 'shuffling of the cards' (as
classical statistical mechanics maintains)."

This seems to me to assume that in a deterministic system
everything is to be explained. But a deterministic system is
one whose state at any future time we can predict if the initial
conditions and its laws are known. It is not the less deter-
mined because we cannot derive the initial conditions and
laws from some other system. Suppose we could: the same
thing would hold for that system. Any "basic and irreducible"
set of initial conditions set up "a long or infinite time ago"
would be a "random" distribution. What would be an "un-
random" distribution? Something that could be derived from
a previous distribution? If so, then the term "random" is
actually being used synonymously with "underivcd" by Pro-
fessor Land6. Unless I radically misunderstand him, all he is
really calling attention to is the fact that every determined
system must start from some initial conditions of material
distribution before our laws can be used to make specific
predictions. This is a logical truth about the nature of any
material system: some data must at some p>oint be given as
basic and irreducible. It is not a discovery but a tautology. If
it constituted an objection to the possibility of "a deterministic
system," the expression would be meaningless. Even a mathe-
matical system has some undefined terms and undemonstrable
propositions.

2. That the subject investigated is affected by the instru-

186 / Determinism and Freedom

ments and methods of investigation is an important point. I
am in wholehearted agreement with Professor Bridgman
about it. But its implication is — so it seems to me — to call
into question not the principle of causality but the "spectator
theory" of knowledge. Sometimes the interaction produced
by the use of instruments and techniques makes it difficult
to predict accurately the outcome of an investigation. Some-
times, however, these reactions are foreseeable. A surgeon can
allow for the effect of his probe or knife on the organism; a
psychologist for the resistance his question arouses in his
subject; a public opinion expert, for the effect of the publica-
tion of his poll on behavior at the election poUs. If the extent
to which the immersion of a thermometer in a solution raises
the temperature of the solution is not detectable at present,
this by itself is not a sufficient reason to assert that it wiU
always remain undetectable.

I am prepared to grant that every act of inquiry into
matters of fact involves the use of our body and its organs as
instruments, and that the process of inquiry is one in which
some actual change in the structure of the situation to be
known takes place. To the extent that this is true, it is true
whether the objects investigated are large or small. I can
see that sometimes it makes sense to say that "a small instru-
ment can find out more about a large object than a large
instrument can find out about a small object." But I can see
that sometimes the contrary makes sense too: for example,
when we bring to bear on a microcellular organism a high-
powered microscope that magnifies it many times over. What
could we learn about such an organism with an instrument
smaller than itself? Perhaps it is true that a large instrument
produces greater changes in a small object than a small in-
strument in a large object, but I should have imagined that
not the size but the relevance and significance of the change is
the point at issue where prediction is concerned. I therefore
cannot see why it should make more sense to say, merely be-
cause of the facts of instrumentation, that causality should
fail for small objects than that it should fail for large, espe-
cially since a large object can sometimes be considered an
organized system of small objects. I could believe the contrary
just as well. My only difficulty is that I cannot see any reason
why we must believe one or the other, or change the idiom
of our talk as we go from one to the other. I conclude, there-
fore, with the observation that although "the Heisenberg un-
certainty relation does not outrage my feeling of what makes

Necessity, Indeterminism, and Sentimentalism / 187

sense" I have not been convinced, by anything Professors
Bridgman and Land6 have said about ordinary experience,
that it is necessary to interpret the uncertainty relation as
entailing a denial of the relation of causality.

ni

The fatal error in the papers of Professors Hospers and
Edwards, as read, is that they alternate between two concep-
tions of "moral responsibility" — one, a conception of moral
responsibility as empty but meaningful, and the other as
vacuous and meaningless. On the first conception, although it
may be true in fact that no one is morally responsible, we can
state the conditions under which one might be. We can differ-
entiate between the two states. On the second, there are no
possible conditions under which anyone can be declared
"morally responsible." The expression has no intelligible op-
posite and thus makes no sense.

The force of most of their arguments, which gives them an
air of high moral concern, is based on the assumption that
under certain circumstances individuals are being improperly
considered responsible. Hospers actually says that "frequently
persons we think responsible are not properly to be called so,"
and Edwards implies the same thing. They explicitly appeal
against the injustice of improperly blaming the morally in-
nocent who, because their desires are determined, are the
victims, not the agents, of misfortune. We eagerly await the
description of the set of conditions under which an individual
is properly held responsible, under which he is not a victim
of circumstances. It then turns out that even if his desires
were undetermined, even if circumstances were completely
different, he would still not be responsible, would still be a
morally innocent victim. The empty conception of moral re-
sponsibility becomes completely vacuous. This makes the
whole procedure of Professors Hospers and Edwards method-
ologically self-defeating, and particularly their expressions
of concern about the injustice of blaming the morally in-
nocent. For to be morally innocent of having committed an
evil deed entails that one is not responsible for its commission,
and to be morally guilty entails that one is. // moral responsi-
bility is a vacuous expression, then moral innocence and guilt
are too. Were Hospers and Edwards consistent they could not
plead for the innocent or condemn the guilty. Edwards in
places suggests that a person would be responsible if he could
ultimately and completely shape or choose his own character.

188 / Determinism and Freedom

But this is explaining an obscure notion by a still obscurer
one. Since every decision to shape or choose one's character,
to be responsible, must be one's own, and therefore already
an indication of the kind of person one is, the notion that one
can ultimately and completely shape or choose one's character
is unintelligible. C. A. Campbell, to be sure, tries to distinguish
between a choice that is the expression of a formed character,
and therefore determined, and a choice of a self. But on
Hospers' and Edwards' argument what is true of character
must be true of self. Either the self has the power to mold
character or it has not. In either case it cannot be held re-
sponsible for having or not having such a native power. And
the same is true if we bring in a Self to explain the powers
of the self and a Great Self to explain the powers of the Self,
etc.

It is true that the notion of moral responsibility is often
ambiguous and not clearly defined in ordinary experience. But
if we follow Professor Hart's illuminating procedure, we can
recognize certain actions in which we clearly admit the pres-
ence of excusing conditions — infancy, insanity, paralysis,
duress, coercion, etc. — and actions in which we do not. We
then try to formulate the principle we recognize in this dis-
tinction and apply it to more complicated and borderline
cases. We find that we tend to hold individuals responsible
for their voluntary or uncoerced actions, for knowingly doing
or not doing what it was in their power to do or leave undone.
All these terms are vague and need further specification.
There are difficulties in ascertaining in particular instances
what it was in one's power to do or leave undone. Nonethe-
less, no one can live in human society without learning to
recognize the distinction between the actions he holds others
and himself responsible for and the actions he does not.

For all its vagueness there is more agreement about how
the distinction is to be applied than about the grounds of the
distinction. No one blames a crawling infant who overturns
a kerosene stove that starts a fire. Almost everybody would
blame a man who, normal in every other way and by all
known tests, insures a house beyond its value and then sets
fire to it without even giving its occupants a chance to escape.
If we make a list of the circumstances behind actions for
which we hold individuals responsible and those for which
we do not, we shall find that as a ruje the first class consists
of those in which evidence exists that praise and reward,
blame and punishment, tend to influence the future conduct

Necessity, Indetermlnism, and Sentimentalism / 189

of those involved and/ or those tempted. This is not the whole
story. Campbell objects ^ that animals are not held responsi-
ble for their actions even though we can re-educate their
desires and impulses by punishment. This is true, but it is
also true that the higher the animal in the scale of intelligence,
the more likely we are to blame it. If we believed that an
axumal could think like a man we would blame it like a man.
The behavior of infants, too, is modifiable by appropriate
reward and punishment even though we do not hold them
morally responsible. But as the age of rationality approaches
we gradually do. This suggests that in addition to suscepti-
bility to reward and punishment, we attribute responsibiUty
where there is a tendency to respond to valid reasons, to
behave rationally, to respond to human emotions in a human
way. Perhaps a third element involved in the attribution of
moral responsibility to voluntary action is the assumption
that volvmtary action is approved action. A man is morally
responsible for an action he commits to the extent that he
approves of it. If he sincerely disapproves of his action, re-
gards it as wrong and condemns it as wrong, but still commits
it we tend to regard him as ill, as acting under "compulsion."
It is some such consideration as this that lies behind our
extenuation of certain kinds of apparently voluntary action
(as when we say: "He didn't mean to do it"), especially
where igorance is present.

There may be other elements involved in the complex
notion of moral responsibility, but the foregoing explains an
interesting phenomenological fact. Sickness, accident, or in-
capacity aside, one feels lessened as a human being if one's
actions are always excused or explained away on the ground
that despite appearances one is really not responsible for
them. It means being treated like an object, an infant, or
someone out of his mind. Our dignity as rational human
beings sometimes leads us to protest, when a zealous friend
seeks to extenuate our conduct on the ground that we were
not responsible (we didn't know or intend what we were
doing, etc.), that we really are responsible and that we are
prepared to take the consequences of our responsibility. As
bad as the priggishness of the self-righteous is the whine of
the self-pitying.

The so-called "hard" determinism professed by Professors
Hospers and Edwards, especially in the popular form de-

^Mind, 1951

190 / Determinism and Freedom

fended by Darrow, whom Edwards so extravagantly praised,
often leads to sentimentality, to so much pity for the criminal
as a victim not of a special set of particular circumstances but
of any circumstances in general (referred to as heredity and
environment, the sway of the law of causality) that there is
not sufficient pity or concern left for the criminal's victims —
not only for his past victims but his future ones and the
victims of others his actions may inspire. To blame and to
punish, of course, are two distinct things logically (except
where blame is considered a form of punishment), but psy-
chologically there is a great reluctance to punish if one be-
lieves there is no blame. Darrow as a "hard" determinist
argued on a priori grounds that everyone was blameless and
often won acquittals not on the evidence but despite it. If
needless pain and cruelty are evils, then punishment that
prevents or deters human beings from committing actions
likely to result in much greater pain and cruelty than it im-
poses is sometimes the lesser evil.

It is argued by Professor Edwards that "hard" determin-
ism, which, according to him, entails the belief that no one is
morally responsible because no one ultimately shapes his own
character, leads to the abandonment of retributive punish-
ment. Even if this were so, it would not make the doctrine of
"hard" determinism any more intelligible. But historically it
is not so. From Augustine to Calvin to Barth the torment of
eternal damnation is assigned and approved independently of
moral responsibility. It is not related of the oft-quoted Puritan
who piously observed to his son when they saw a man being
led to the gallows, "There but for the grace of God go I,"
that he opposed retributive punishment. Nor can Edwards con-
sistently with his own theory assert that "hard" determinists
should repudiate retributive punishment, or morally blame
them or anyone else, as he freely does, for approving of retri-
butive punishment. For has he not told us that a man can't
help having the character he has, no matter what kind of a
character it is? Further, if retributive punishment is the
enemy, there seems to me to be no necessary logical connec-
tion between a belief in moral responsibility and approval of
retributive punishment. Certainly, "soft" determinists who
assign responsibility to actions only when there is reason to
believe that blame or punishment will modify future conduct
are hardly likely to defend retributive punishment.

Why, after all, is retributive punishment evil? Not because
the wrongdoer "ultimately did not shape his own character"—

Necessity, Indeterminlsm, and Sentimentalism / 191

whatever that may mean — but simply because the pain in-
flicted on him gratuitously adds to the sum total of suffering
in the world without any compensating alleviation of anybody
else's sufferings. Even if an individual were considered able
"ultimately to shape his own character" and were held mor-
ally responsible for an evU act, punishment that would be
purely retributive and that did not contribute to deterring him
or others from evil doing, or did nothing toward rehabilitating
him, would still be morally wrong. This is quite evident in
situations in which the "hard" determinist who is not a fatal-
ist, if there be any such, admits that a man is to some extent,
not ultimately but proximately, responsible for some change
in his character — for example, when his desire to gamble
leads him to steal a beggar's portion. In such situations retri-
butive punishment as such would be regarded as morally
wrong. Directed only to the past, it would not give the beggar
back his portion or wipe out his pain, and therefore the new
sufferings it inflicts are futile and needlessly cruel. If one can
oppose retributive punishment when one believes a person
is proximately responsible for his action, one can oppose it
even when one believes a person is ultimately responsible,
whatever the cognitive content of that belief turns out to be.
If retributive punishment is the target of their criticism,
Hospers and Edwards are training their guns in the wrong
direction.

Far from diminishing the amount of needless cruelty and
suffering in the world, I am firmly convinced that the belief
that nobody is ever morally responsible, in addition to being
false, is quite certain to have a mischievous effect and to in-
crease the amount of needless cruelty and suffering. For it
justifies Smerdyakov's formula in The Brothers Karamazov:
"All things are permissible." One of the commonest experi-
ences is to meet someone whose belief that he can't help doing
what he is doing (or faihng to do) is often an excuse for not
doing as well as he can or at least better than he is at present
doing. What often passes as irremediable evil in this world,
or inevitable suffering, is a consequence of our failure to act
in time. We are responsible, whether we admit it or not, for
what it is in our power to do; and most of the time we can't
be sure what it is in our power to do until we attempt it. In
spite of the alleged inevitabilities in personal life and history
human effort can redetermine the direction of events, even
though it cannot determine the conditions that make human
effort fK>ssibIe. It is time enough to reconcile oneself to a

192 / Determimsm and Freedom

secret shame or a public tyranny after one has done one's best
to overcome it, and even then it isn't necessary.

To say, as Professor Hospers does, that "It's all a matter of
luck" is no more sensible tiian saying: "Nothing is a matter
of luck" — assuming "luck" has a meaning in a world of
hard determinism. It is true that we did not choose to be
bom. It is also true that we choose, most of us, to keep on
living. It is not true that everything that happens to us is like
"being struck down by a dread disease." The treatment and
cure of disease — ^to use an illustration that can serve as a
moral paradigm for the whole human situation — would never
have begun unless we believed that some things that were
did not have to be, that they could be different, and that we
could make them different. And what we can make different
we are responsible for.

Chapter 9

PurBBshmenf as Jystiee and @s Friee;

Abba Lerner, Roosevelt College

I

Regarding the problem of personal responsibility, it seems
to me that Professor Edwards' argument, with which I am
in close agreement, does not really depend on a demonstra-
tion that nobody makes his own character any more than it
depends on the arguments and examples brought forth by Pro-
fessor Hospers on the importance of subconscious drives and
compulsions in criminal activity. Professor Edwards' position
really seems to rest on the rejection of the axiom that it is
desirable or just that a person who has committed a crime
should be made to suffer. This axiom does not depend at all
on who, if anyone, is responsible for the bad character of the
bad man who performs the bad act. There is thus no sug-
gestion that he should be punished for having a bad character
if he did not commit the crime or even for having chosen to
have the bad character or to make it for himself. If making
a bad character is a crime in itself that is another crime. It
may raise the question whether whoever was responsible for
making the bad character should be punished for it, but that
is a different question. The rejection of the axiom that crime
must be atoned by the suffering of punishment leaves punish-
ment without justification unless some other justification is
found, such as the protection of society through the provision
of a deterrent to antisocial behavior. The history of how the
criminal became a bad man is then irrelevant, even though it
may induce jurors to direct acquittals. The relating of such
history does not constitute a logical argument against carrying
out any particular punishment, but it may lead to the reieclion
of the axiom that justice is done when crime is balanced by
the suffering of punishment.

II

I should like to comment on Professor Hart's treatment of
the conditions for excusing punishment on the same footing

193

194 / Determinism and Freedom

as conditions for the invalidation of contracts, as if punish-
ment were a price. His argument is that it is good to be able
to measure the punishment against the satisfaction one may
get from committing a crime. This increases freedom by per-
mitting the exercise of rational choice in committing the crime
only where the satisfaction is greater than the punishment-
price. Any contract-crime not freely and responsibly entered
into is then invalidated and the price-punishment is canceled.

I was particularly interested in this because I have been
doing almost exactly the opposite in economics — not treating
punishment as a price, but treating price as a punishment.
The price of a commodity is a deterrent to the consumer; it
deters him from consuming the product. Ideally, the magni-
tude of the price should correspond to the amount of damage
that the consumer does to the rest of society by consuming
the product and thereby making it unavailable for others to
consume. The individual will then choose to consume if the
satisfaction he gets is greater than the price and therefore
greater than the loss to the rest of society. And since by the
payment of the price he completely compensates the rest of
society, society as a whole (including our consumer) benefits
from his decision to commit the consumption. Furthermore,
the complete compensation of the rest of society (by the price
that he pays) leaves it without reason for wishing to interfere
with his freedom of choice. The proper price, therefore, leads
to the maximization of welfare and of freedom.

This kind of maximization does not, however, apply to the
punishment of crime because the suffering of the punished
criminal is not balanced by a benefit to the rest of society
(except where the punishment takes the form of a money
fine and does operate just like a price). The committing of a
crime does not mean that society as a whole (including our
criminal) benefits. Even if the criminal enjoys a net benefit
after punishment, his punishment does not compensate the
rest of society for the damage done them by the crime. On
the contrary, the rest of society suffers further losses from the
costs involved in providing justice and keeping the man in
prison. In the case of the price mechanism, if the consumer
decides to consume the commodity and suffer the purchase
price, the price system is performing its function of ensuring
maximum welfare and freedom perfectly. But if the criminal
decides to commit the crime and then pay the punishment, the
punishment system has failed in its function of preventing the
crime.

Punishment as Justice and as Price; On Randomness / 195

III

A point that intrigued me particularly was Professor
Lande's treatment of randomness. His argument seemed based
on the idea that a random distribution is a special kind of
order that in a determinate universe must have been planned
by some demon. The question that remains is only whether
there is a demon operating at every point where an apparently
random distribution exists — or perhaps in every subatomic
particle — or whether there was some superdemon who ar-
ranged all these random distributions at the creation. I don't
find myself tempted by either form of demonology because I
have always thought of random distribution as precisely the
opposite of order, on the assumption that randomness is the
mark of the absence of order and that as long as any distribu-
tion is random no amount of study or observation of it will
give any information as to the probabilities of the next throw.
But I am not very happy about my position. I think I should
look into the philosophical and perhaps mathematical treat-
ments of randomness to see if what I am saying does not
involve an elementary blimder.

Chapter 10

Some Notes on Determinism

Ernest Nagel, Columbia University

Whether the occurrence of every discriminable event is
determined, whether for every event there is a unique set of

,; conditions without whose presence the event would not take
7 tplace, and whether if conditions of a specified kind are given
an event of a certain type will invariably happen, are variant
forms of a question that cannot be settled by a priori argu-
ments. Nor do I think the questions can be answered defini-

■tively and finally, even on the basis of factual evidence; for,

f as I shall presently suggest, the question is best construed as
dealing with a rule of procedure for the conduct of cognitive

^ inquiry, rather than with a thesis concerning the constitution
of the world. I am therefore not convinced by Professor
Blanshard's acute argument attempting to show that an an-
swer to the question other than an affirmative one is indefen-
sible, if not unintelligible. Moreover, his assertion that even in
deductive thinking and artistic invention each step is neces-
sitated by the logical and the aesthetic relations that exercise a
power over the mind seems to me untenable — if it is admitted
as relevant to his major contention. For his assumption that
logical and aesthetic relations (as distinct from apprehensions
of such relations) may be said to engage in causal action at-
tributes causal efficacy to something that, in no recognizable
and identifiable sense of the phrase, can exercise such agency.
Nevertheless, the belief in determinism is not unfounded;
and it would be just silly to maintain that in no area of experi-
ence can we rightly affirm that anything is caused or deter-
mined by anything else. It seems to me the special merit of
Professor Black's paper that it indicated at least one identifi-
able class of contexts in which the words "caused," "deter-
mined," and their derivates have an unquestionable and im-
portant use. In these contexts involving human action it is
simply nonsense to deny that events have causes or effects, in
senses of these words appropriate to these contexts. Black has
also made plain that the conditions under which "caused" and
"determined" have significant uses, in those situations where
men initiate actions and are responsible for the occurrence of

196

Some Notes on Determinism / 197

events, require the presence of identifiable contingencies and
the absence of just such "necessities" as those for which
Blanshard argues.

But it does not follow from Black's analysis that the only
sense that can be attached to "cause" and "determined" is the
sense they manifestly do have in the indicated contexts — any
more than it follows that, because the word "number" is un-
doubtedly used in situations involving the counting of objects,
the meaning of the word in statements about such irrational
magnitudes as the area of a circle with a unit radius must also
involve reference to counting. There are, to be sure, historical
continuities and important analogies between the use of these
words in contexts of human action and their use in discussions
about, say, the "indeterminism ' of electrons. But it is patently
a mistake to construe the meanings of those words in this
latter context in terms of the "paradigm"' for their use in situa-
tions where men are correctly identified as causal agents. Al-
though Black does not explicitly guard himself against the
suspicion that he does take his paradigm as basic for all uses
of "caused," it is unlikely that he would commit himself to
such as position. On the other hand, it does seem to me that
Professor Bridgman (and perhaps even Professor Lande)
commits a somewhat similar mistake when he suggests that
the "indeterminacy" of quantum theory can be explicated in
terms of familiar facts "in the sphere of ordinary life."

In the voluminous literature on the "indeterminism" of
microphysics, one point stands out clearly: whatever the issue
may be, it is generated by the theoretical interpretations that
are placed on the acknowledged data rather than by any dis-
agreement as to what those data are. Thus no one disputes
that when a beam of light passes through appropriately ar-
ranged slits and strikes a zinc sulphide screen, scintillations oc-
cur that fall into a definite pattern; or that quantum theory
accounts for the occurrence of each individual scintillation.
Problems arise, however, when the structure of quantum
theory is analyzed with a view to showing why it is that this
theory cannot account for individual scintillations. But the
problems are generated because answers to the questions are
proposed in terms of familiar facts "in the sphere of ordinary
life" rather than in terms of the structure of the theory itself.

It is a commonplace that quantum theory employs a dis-
tinctive way of "describing" the state of a physical system
with which the theory can deal. This state description (the psi
function) is such that, given its value for some initial time, and

198 / Determinism and Freedom

assuming an appropriate set of boundary conditions for the
application of the theory, the theory makes it possible to cal-
culate the value of the function for any other time. In this
respect quantum theory is as "deterministic" as are the dy-
namical theories of classical physics. It differs from these in
that, while the state description of the latter can be construed
as representing magnitudes associated with certain individual
elements that constitute the physical systems in question, its
state function can be construed as representing only a statisti-
cal property of the individual elements making up the physical
system. In short, the state description of quantum theory is a
statistical parameter So far nothing could be more straight-
forward or less puzzling. The puzzle begms when reasons are
offered why the state function of quantum mechanics is a
statistical parameter.

The reason Professor Land6 appears to give is that any
given value assigned to the psi function constitutes the initial
conditions for the application of the theory to a concrete situ-
ation, and that since initial conditions constitute a brute and
underived fact they represent an inherently chance or random
feature of the world I doubt very much whether I have under-
stood Professor Land^'s presentation of his views, and my
comments may be entirely irrelevant to his real intent. But as
I understand him, he has not made clear what he set out to
clarify. For every theory — not only quantum mechanics —
requires initial conditions that at some point or other in an
investigation into concrete subject matter must be accepted as
underived and therefore as representative of a "random" fea-
ture of the world, as are the initial conditions for quantum
theory. This attempt to assimilate the "indeterminacy" of
microphysics to facts "in the sphere of ordinary life" is not a
successful one.

Professor Bridgman seeks to explain the statistical aspect
of the psi function by invoking the general principle that
whenever measurements are made the instruments employed
interact with the things measured and thereby introduce
changes into the latter. His contention is that, although such
alterations are practically negligible when we measure things
that are sufficiently large, the changes cannot be ignored when
the minute "particles" of microphysics are measured with the
relatively large instruments at our disposal, so that the psi
function inevitably represents only statistically significant
magnitudes associated with the elementary particles of quan
turn physics. Now, the general principle Professor Bridgman

Some Notes on Determinism / 199

invokes is undoubtedly sound. The difficulty in his explana-
tion, however, is that though the principle is sound it does not,
in other areas of inquiry, prevent us from calculating the
effects of measuring instruments on the things measured and
so making corresponding allowances in assigning magnitudes
to the objects under investigation. Why should the situation be
inherently different in quantum physics? I find it difficult to
escape the impression that Professor Bridgman has put the
cart before the horse. For it seems to me that the alleged effect
of measurement on microphysical "particles" must be as-
sumed as at best a consequence that follows from the accept-
ance of quantum theory, rather than that the theory is based
on independently ascertained facts concerning the alterations
made by instruments of measurement on microphysical "par-
ticles." At any rate, I do not think Professor Bridgman has
convincingly shown that the "indeterminism" inherent in the
structure of quantum theory is but another illustration of a
familiar feature "in the sphere of ordinary life."

Although quantum theory is not deterministic in the precise
sense in which the dynamical theories of classical physics are
deterministic, and although quantum theory (or, for that
matter, any other available theory of contemporary physics)
does not account in detail for such occurrences as the individ-
ual scintillations mentioned previously, it of course does not
follow that there really are no precise conditions for the oc-
currence of those events that quantum mechanics does not
explain, or that a theory that can account for these things is
impossible. The assumption that there always are such precise
conditions for every event, even if we continue to remain
permanently ignorant of them, is the assumption of a uni-
versal determinism. As I have already indicated, determinism
so understood is capable neither of decisive proof nor dis-
proof. I think, nevertheless, that determinism can be regarded
as a fruitful maxim or regulative principle for inquiry. It
does not express a necessity of thought, for it can be aban-
doned. But if it is abandoned, then inquiry in certain direc-
tions is, at least temporarily, brought to a halt. In an impor-
tant sense, therefore, the deterministic maxim is explicative of
what is generally understood to be a goal of the scientific
enterprise.

I want to conclude with a brief comment on the contention
of Professor Edwards and Hospers that, if determinism sup-
plies a true account of the nature of things, it does not make
sense to hold anyone morally responsible for his actions or to

200 / Determinism and Freedom

offer moral praise and blame. Under what conditions do we
hold a person morally responsible for an action? Consider an
example. I engage the services of a student as baby sitter on
the assumption that she is capable of doing certain things. Her
ability to do them depends on a number of conditions, includ-
ing the state of her body, her education, and her previous
experience with children of a certain age. If she does indeed
satisfy these conditions and also agrees to perform certain
tasks that are compatible with her abilities, she is morally
responsible for performing them. The fact that she did not
create her own body, or that she did not choose the education
she received, are not relevant considerations for judging
whether she is morally responsible for some event that may
take place during my absence from home. On the other hand,
if during my absence burglars enter and tie up the student, or
if she becomes unconscious for causes not within her control,
she is not morally responsible for what may befall my chil-
dren. The point is obvious. Moral responsibility is correctly
ascribed to individuals who possess certain capacities; and it is
correct to make the ascription for the sufficient reason that
this is just the way the phrase "morally responsible" is used.
The fact that possessing these capacities is contingent on a
variety of conditions, most of which are perhaps beyond the
control of an individual, is irrelevant to the analysis of what
we do mean by the phrase as well as to the grounds on which
the ascription is rightly made. To maintain the contrary is in
effect to maintain that no property can be correctly predicated
of an object if the property is causally dependent on anything
either in the composition or in the evironment of the object.
Such a view makes all predication impossible. But in any
event Professor Edwards and Hospers can sustain their thesis
only by radically altering the customary conception of what
it means for anyone to be morally responsible.

Chapter 1 1

Causation, Def©rminism, and

the ''G9@d''

F. S. C. Northrop, School of Law, Yale University

Professor Bridgman has suggested that there is much more
agreement among physicists concerning the status of the con-
cept of causality in quantum mechanics than appears super-
ficially from quotations pulled out of context. There are rea-
sons for believing that this agreement is more unanimous even
than Professor Bridgman has indicated.

One difficulty is that mechanical causation in modem
physics has a stronger and a weaker meaning, and as yet there
is no convention among physicists and informed philosophers
of science as to which of these two meanings the expression
"mechanical causation" is to have. Some physicists use these
words in the stronger sense, others in the weaker sense. Since
with respect to quantxmi mechanics what holds for the weaker
meaning does not hold for the stronger meaning, physicists
often appear to differ about the status of causality in quantum
mechanics, when in fact there is complete agreement. Fur-
thermore, there was no difference of opinion between Einstein
and quantum physicists concerning the status of mechanical
causality and determinism in quantum mechanics. It is pre-
cisely because Einstein agreed with them about what the
theory requires that he objected to quantum theory. It can be
shown, however, that according to Einstein's own theory of
the relation between theoretical concepts and experimental
evidence, his objection is untenable. The situation becomes
clear when the weaker and stronger meanings of mechanical
causality are precisely defined. As understood by physicists,
determinism is equivalent to the stronger of the two types of
mechanical causation. This is probably also what a man with
common-sense knowledge means by determinism, even though
he is rarely, if ever, clear about what this meaning is.

In the paradigm that marked Professor Black's approach to
the problem of determinism, by the method of contemporary
British analytic philosophy, common sense was again fol-

201

202 / Determinism and Freedom

lowed. In order to relate his conclusions to the concept of
determinism as understood in mathematical physics, two
things must be noted: (1) the ambiguities in his paradigm
and in the method of British analytic philosophy generally;
(2) the difference between the definition of cause in his para-
digm and the concept of cause in mathematical physics.

The ambiguities of Professor Black's paradigm, as of Brit-
ish philosophical analysis generally, arise from the fact that
such philosophical analysis, so-called, never makes explicit
which one of several possible epistemological meanings the
words convey in the common-sense contexts. The result is a
surreptitious shift from assertions true for only one possible
epistemological meaning to conclusions valid, if at all, only
for different epistemological meanings. As a result, British
analytic philosophy confuses more than it clarifies, as the fol-
lowing analysis of Professor Black's paradigm shows.

In the sense in which this paradigm "stands on its own
feet," as Professor Black maintained it did, it must be taken
in its radical empirical, nominalistic meaning. But in this
sense, as Hume showed, there is merely temporal succession
and no relation of necessary connection, or causality, of any
kind. Professor Goodman's remark that causality is a tautol-
ogy is interesting in this connection. It is the inevitable con-
sequence of any attempt, after the manner of Professors
Quine and Goodman, to rear mathematics and mathematical
physics on nothing but nominalistic, radical empirical epis-
temological meanings. Then not merely mechanical causality
in either the weaker or the stronger form — to be defined in the
sequel — but most of the other concepts of mathematical
physics become similarly either meaningless or trivial.

When one takes Professor Black's paradigm in its naive
realistic epistemological meaning, as he suggests in using the
word "object" and conceiving of causality as a relation be-
tween objects, then, to be sure, perhaps a bit more than
Hume's mere temporal succession is obtained, but in this case,
as Professor Weiss points out, the paradigm does not stand on
its own feet, and the conclusion Professor Black draws about
the compatibility of determinism and freedom fails to fol-
low. Certainly, interpreted as naive realistic epistemological
material objects, the things described by the paradigm may
well be subject to mechanically antecedent causes.

The latter consideration suggests that the paradigm becomes
relevant to the problem of causality — to say nothing about its
relation to mechanically causal determinism — only if a third

Causation, Determinism, and the "Good" / 203

surreptitious epistemological shift occurs in the interpretation
of it, in which the person who "causes something to happen"
is not viewed from without but is instead viewed as purpose-
fully acting from within. But again, if this is the meaning, the
causality is teleological, and the difficulties raised by the
mechanically causal determinism of modern physics are not
touched. Furthermore, if the causality is purposeful and teleo-
logical, may not the individual's choice of this purpose have
been mechanically determined? Again we see that so far as
the paradigm has any relevance to the problem of determin-
ism it does not stand on its own feet. Like British philosophi-
cal analysis generally, the so-called clarification of the prob-
lem of determinism that Professor Black's common-sense
paradigm is supposed to give is so shot through with epistemo-
logical ambiguities and surreptitious shifts in the epistemo-
logical meanings of the words used that it is, like this method
generally, philosophically useless.

In any event. Professor Black doesn't touch the problem of
determinism, which arises only when one thinks of causality
as modern physicists do: not, after the manner of Professor
Black, as a relation between objects but as the relation be-
tween the states of the same object, or the same system of ob-
jects, at different times. Mechanical causation affirms, in either
its weaker or its stronger meaning, that this temporal relation,
between the states of a system at different times is a relation/
of necessary connection.

Since, as Hume showed, we do not observe any relations of
necessary connection, two things foUow: (1) The concept of
mechanical causation in modern physics cannot be attained
merely by direct inspection of a common-sense example or
by so-called "analysis" of the grammar of an Englishman's
description of such an example; only temporal succession wUl
be found by such a method. (2) Physical systems obeying
mechanical causation can therefore be known only by deduc-
tively formulated, axiomatically constructed, indirectly veri-
fied theory. To determine, therefore, what the status of
mechanical causality is in modem physics, one must do more
than examine the inductive, radically empirical sensible opera-
tions— important as these are — ^by means of which the specu-
latively proposed, axiomatically constructed theory is verified
indirectly. This type of examination will give one merely
operational, theory-of-errors probability, which is present in
any experimental testing of any theory and hence is quite in-
sufficient to distinguish quantum mechanics from previous

204 / Determinism and Freedom

modern physics with respect to the status of mechanical cau-
sation and, more particularly, its stronger form, determinism.
Operational probability enters into all scientific theory. This
type of probability and chance is frequently referred to as
epistemological probability and chance, since it has its basis
in the finiteness and errors that accompany the scientist's
attempt to relate himself in knowledge to the object. To under-
stand the unique conclusion of quantum mechanics with
respect to mechanical causation and its stronger form, deter-
minism, one must concentrate attention on the speculatively
introduced, axiomatically constructed postulates of the modern
theories of mathematical physics, with particular reference to
the presence or absence of the concept of theoretical probabil-
ity— i.e., probability introduced in principle — in the postu-
lates. Since such probability, if it occurs, refers to the object
\ of scientific knowledge, it is appropriately called ontological
probability or chance.

Two factors in any deductively formulated theory of
modern physics must command the focus of one's attention.
These two factors are (1) the definition of the state of the
system at any given moment of time and (2) the definition of
the time relation between states of the system at different
times.

Let us concentrate on the latter relation first. There are
three major possibilities. One is instanced by any merely in-
ductive empirical observations of the changes of the system
through time without any speculatively introduced, axiomati-
cally constructed theory. By plotting these changes from the
earlier to the later state of the system, a curve can be ob-
tained. For every curve, as Professor Philip Frank and others
have noted, there is a formula or law. Such lawfully defined
relation between the earlier and later states of a system is not,
however, regarded by modem physicists as constituting a
truly mechanically causal system. The reason for this conclu-
sion is that such a law or formula cannot be given until after
one has seen the system pass from its earlier to its later state.
Thus it has no predictive power. As Professor Frank noted, it
merely affirms that what happens happens and in this sense,
again, is little more than a trivial tautology so far as causality
is concerned.

There is a second possibility with respect to the nature of
the time relation between states of an isolated system. The
relation may be one of necessary connection characterized
by a repetition of constant time relations that hold true not

Causation, Determinism, and the "Good" / 205

merely for past and present but also for future cases; but it
is a relation such that in any first observed instance, given the
initial state of the system, one cannot predict the future state
until one has observed the present state passing into the future
or final state in at least one instance. The physical system in
which the initial state is that of the acorn and the final state
that of the oak constitutes an example. Such a relation of
necessary connection, since a knowledge for prediction of
future states of the system depends on a knowledge of what
the future state is, is called teleological causality.

There is a third possible form that the time relation between
the states of a scientifically determined system may possess.
In this third type, as in the second, the relation is one of
necessary connection, but the relation of necessary connection
has the following properties: (1) There exists a speculatively
introduced, axiomatically constructed set of postulates of an
indirectly and experimentally confirmed, deductively formu-
lated theory. (2) The postulates of this theory specify a very
small number of independent variables necessary and suffici-
ent to define completely the state of a system at any specific
moment of time /^. (3) Given the operationally determined,
concept-by-intuition values of the independent variables of
this definition of the state of the system at an earlier time /^,
(a) all the other properties of the system at that time ?j can
be deduced, and (b), without any observation in a present or
past instance of the future state of a similar system, the em-
pirical values of the independent variables of the physical
system at the later time t^ can be deduced by solving a second-
order differential equation that the postulates and theorems
of the theory provide. When the foregoing conditions are satis-
fied, the causality of the system, i.e., the relation of necessary
connection between the states of the system at different times,
is mechanical causality in the forementioned weaker meaning
of this term. More concretely, this means that the deductively
formulated theory provides a time equation such that, by feed-
ing the operationally determined empirical values of the con-
cept-by-postulation, theoretically introduced independent vari-
ables of the state function into the equation, the values of
these variables for a specific later time t^ are completely deter-
mined by solving the equation for that time t^.

Note that causality in this sense is not a trivial tautology.
It is, instead, a relation of necessary connection such that,
given the postulates of the theoryand given the determination
of the independent variables in any present state of the sys-

206 / Determinism and Freedom

tern, a completely novel future state, never before observed in
any past instance, can be deduced. Furthermore, a large num-
ber of other properties of the system in the initial state and
in the final state can also be deduced.

What is the role of the foregoing type of mechanical caus-
ality in the weaker meaning of these two words in any theory
jof modern physics in which such a relation of necessary con-
nection between the earlier and later states of a system is pres-
jent? The answer to this question is unequivocal. In the postu-
i.lates of the theory the concept of probability does not enter
into the definition of the time relation between the states of
any system. The first important thing to note about quantum
mechanics is that, as for Newton's mechanics, Maxwell's
electromagnetics, and Einstein's special and general theories
of relativity, mechanical causality in the weaker meaning
holds. This is the case for two reasons: First, the concept of
theoretical probability does not enter into its definition of the
time relation between the states of a subatomic physical sys-
tem. Second, it provides the Schrodinger time equation, which
has the aforementioned property of enabling one, given the
operationally determined empirical values of the present state
of the system as defined by the theory, to deduce the future
state of the system, as so defined, at any specified later time t^
merely by solving this equation.

But if mechanical causation in the weaker meaning of the
words holds for quantum mechanics just as it did for New-
ton's mechanics, Maxwell's electromagentics, and Einstein's
special and general theories of relativity, why, then, has it
been affirmed by physicists that quantum mechanics alters the
status of causality — and more particularly of determinism —
in modern physics? The answer to this question will become
clear if one shifts attention from the definition of the time
relation between states to the definition of state in the mod-
ern theories. The novelty of quantum mechanics consists in
this, and solely in this, that theoretically, and hence in prin-
ciple, it has found it necessary in order to account for the
experimental data to introduce the concept of theoretical
probability into the definition of state of a subatomic physical
system. In Newton's mechanics, Maxwell's electromagnetics,
and Einstein's special and general theories of relativity, the
concept of theoretical probability was not introduced either
into the definition of state or into the definition of the time
relation between states. Hence in these four theories mechani-
cal causality in the stronger sense of absolute determinism

Causation, Determinism, and the "Good" / 207

held. There was, in other words, no ontological chance or
probability in the object of empirically verified, theoretically
designated scientific knowledge. Thus, according to these four
theories, causality is not merely mechanical, but it is also
unqualifiedly deterministic. Whatever probability existed be-
longed merely to the experimental, operational side and hence
was merely epistemological. It was precisely because Einstein's
own theory required, and because he consequently believed,
that probability in physical science should have a merely epis-
temological, theory-of-errors, operational status that he ob-
jected to quantum mechanics. Thus there was no disagree-
ment between Einstein and quantum physicists about what
quantum mechanics afl5rms with respect to the existence of
ontological chance and probability. The answer, therefore, to
Professor Quine's question whether probability in quantum
mechanics is merely epistemological is unequivocally "No."

For this reason also Professor Blanshard's and Sir David j
Ross's contention that quantum mechanics provides no evi-/
dence against causal determinism is erroneous. Professor
Blanshard was quite right when, quoting Sir David, he noted
that epistemological probability and chance do not imply
ontological probability and chance. But the novelty of quan-
tum mechanics consists precisely in the fact that, unlike pre-
vious modern scientific theories in which probability and
chance are merely epistemological, quantum physicists have
found it impossible, when faced with experimental findings on
black-body radiation, to accept mere epistemological or opera- j
tional, theory-of-errors probability and have been forced to 1
introduce instead the concept of probability in principle at the
theoretical level into the specification of the type of causality /
governing the object of scientific knowledge itself. This is the
case because in quantum mechanics' definition of state the
postulates in principle prescribe not merely the specification
of the two independent variables, position and momentum, for
each object of the system, but also a packet of such numbers
with each one accompanied by its respective probability num-
ber. Furthermore, because the probability is introduced in
principle, the determinism is not, as Professor Williams has
suggested, restored by including the experimental apparatus
within the object of scientific knowledge. Also, the Compton
effect provides experimental evidence to the same conclusion.

Nevertheless, given a state of a subatomic system defined
in terms of a packet of position and momentum numbers for
each object in the system, accompanied by their correspond-

208 / Determinism and Freedom

ing probability numbers, the postulates of the theory are such
that future state of the system defined in terms of position-
momentum numbers and probability numbers can be deduced
by recourse to the Schrodinger time equation. Consequently,
in quantum mechanics causality in the weaker sense still holds,
but mechanical causality in the stronger sense of determinism
does not hold.
f It remains to be shown why {a) Einstein's objections to

/quantum mechanics and {b) the interpretation of it suggested
in this conference by Professor Williams are untenable. When

- Einstein came to the conclusion that it is possible to reconcile
the results of the Michelson-Morley experiment with the basic
theoretical assumptions of modern mechanics and electro-
magnetics only by modifying a basic postulate introduced by
Newton and assumed by Maxwell concerning space and time
and the simultaneity of spatially separated events, Einstein
himself tells us that he was faced with a methodological and
epistemological difficulty. Newton, having asserted that he in-
troduced no hypotheses and deduced the concepts of his
theory from the experimental data, left the impression — like
Professor Williams in his attempt to derive the concepts of
modern physics from the experimental data — that the experi-
mental data entailed the concepts. Einstein saw that if this
were so he could not meet the difficulty raised by the Michel-
son-Morley experiment by modifying Newton's theoretical
assumptions. For the experiments performed by Newton can
certainly be repeated today with the same results. Hence, if
the experiments entail Newton's theory, Newton's theory must
still be retained. Consequently, Einstein conducted a fresh
examination of the relation between the experiments of the
experimental physicist and the theoretical concepts of his
science. In this essentially epistemological investigation Ein-
stein had no difficulty showing that the simultaneity of spati-
ally separated events, which is the same for all observers on
the same Galilean frame of reference, is not directly sensed
but is instead a speculatively introduced, indirectly verified
hypothesis. Hence it followed that it was not necessary to
specify the relation between space and time in the manner
introduced by Newton. This opened the way for Einstein to
account for all the experiments in accord with Newton's
theory and the Michelson-Morley experiment without contra-
diction in the theoretically novel manner of Einstein's special
theory of relativity. This means that there is no a priori rea-
son knowable ahead of time why the causality of a physical

Causation, Determinism, and the "Good" / 209

system must be of the stronger mechanical type. Hence there
is no epistemological justification, on Einstein's own analysis
of the epistemology of modern physical knowledge, for his
objection to the introduction of the theoretical concept of
probability into the definition of the state of a subatomic
physical system in quantum mechanics.

Furthermore, there are experimental reasons, exactly ana-
logous to the Michelson-Morley experiment, demonstrating
that the concept of probability cannot be kept out of the
definition of state for subatomic systems. Before Planck intro-
duced his quantum concept the traditional experimentally veri-
fied, speculatively proposed theories prescribed in principle
that the positions and momenta of subatomic, as of molar,
scientific objects were determinable and behaved without
reference to chance or probability. In short, the theory of
subatomic systems was strictly deterministic, introducing the
concept of probability into neither the definition of the time
relation between states nor the definition of state. When, how-
ever, the latter theory of the type of causality governing sub-
atomic scientific objects was pursued mathematically to its
deductive consequences, certain experiments on black-body
radiation, exactly analogous to the Michelson-Morley experi-
ment with respect to motion in electromagnetic systems, were
performed, and these experiments did not give the result that
the assumption of no probability in the definition of the state
of subatomic systems requires. It was to meet this difficulty
that Planck introduced his constant. Later Heisenberg showed
that Planck's constant entails the uncertainty principle. It is
this entailment, necessary to reconcile the experimental find-
ings on black-body radiation with physical theory, that forces
the introduction of the concept of probability in principle into
the definition of state in quantum theory. Thus the very same
epistemology and logic of the experimental situation that
drove Einstein, following the Michelson-Morley experiment,
to the modification of Newton's assumption concerning space
and time, drives the subatomic physicists, on the basis of the
experiments on black-body radiation, to the modification of
the definition of state and the strict determinism of previous
physical theories. Hence, on his own theory of the episte-
mology of mathematical physics and its methodology, Ein-
stein's objections to quantum mechanics are invalid. We have
no alternative but to hold that for subatomic systems, mechani-
cal causality in the weaker sense of the words still holds, but
determinism or mechanical causality in the stronger sense does

210 / Determinism and Freedom

not hold. In short, quantum mechanics does not introduce the
concept of theoretical probability into the definition of the
time relation between states but does introduce it into the
definition of state. In other words, quantum mechanics, as
Professor Bridgman's paper suggested, introduces ontological
and not merely epistemological chance and probability.

By way of contrast it may be relevant to note a theory of
modern physics — namely, thermodynamics in its statistical
interpretation — in which the concept of probability is not in-
troduced into the definition of state but is introduced into the
definition of the time relation between states. Thus, in deter-
mining the state of such a thermodynamic system at any given
time, merely the values of the independent variables — tem-
perature, energy, etc. — need be specified and no probability
numbers need be attached. But given these operationally
determined values of the independent variables in the state
function, the future state of the system, again described com-
pletely without probability numbers, is predicted only with a
high, specified degree of probability. The time relation be-
tween states is one, therefore, not of necessary connection, but
merely of highly probable connection. This is why such scien-
tific theories are called statistical and noncausal theories.

II

Two questions may be asked of Professors Edwards and
Hospers. The question to be put to the former is, "What is the
meaning of the word 'good' when you say that it would be
better if people did not pass moral judgments of praise and
blame on what others do, but accepted the 'objective' truth
that they cannot help themselves? If everything is absolutely
determined and there is no moral responsibility, what is the
meaning of 'better' and 'worse' as the words are used at the
end of your paper?" The question addressed to Professor
Hospers is, "If aU theories obtained by rigorous scientific
methods are rationalizations in the vicious sense of this word,
how can you be sure that the Freudian theory in which you
have so much confidence — whose scientific methods still re-
main misty and unclarified — is not also a vicious ration-
alization?"

May I suggest a possible answer to the questions addressed
to Professor Edwards. Implicit in his use of the word "better"
in his assumption that "good conduct" is conduct in accord
with empirically verified scientific theory concerning the
status of human deeds with respect to causal determinism.

Causation, Determinism, and the "Good" / 211

This implicit assumption is equivalent to saying that "good" is
not an undefinable concept, as G. E. Moore maintained, but
is instead to be defined in terms of theory that is scientifically
verified as true. There are other reasons for supporting this
analysis of the meaning of the word "good" and of the words
"morally better" and "morally worse." Clearly, facts merely
are; they are not true or false, good or bad. This being so, it
seems difficult to escape the conclusion that goodness and bad-
ness, like truth and falsity, must be a function of propositions
about fact. But if this analysis of the meaning of the word
"good" is to be significant, then human behavior and human
choice must be, in part at least, a function not merely of fact,
but of the theory, true or false, that people hold about facts
and on which they act.

The following question takes one, then, to the heart of the
problem of moral responsibility: Do theories matter in human
acts? Empirical cultural anthropology has shown that what
people do in a given culture is a function of the mentality they
share, i.e., the propositions they explicitly and implicitly be-
lieve. There is experimental evidence that the theory of
"trapped universals" of McCulIoch and Pitts makes it mean-
ingful to say that the motor response of the nervous system
(i.e., what a person does) is a function not merely of the sen-
sory impulses (i.e., the purely inductively given facts affecting
the nervous system) but also of the symbolic trapped univer-
sals in the cortex that integrate and interpret the input stimuli,
thereby to a significant extent specifying the form of the
motor response.

Hence, quite apart from the scientific evidence that deter-
minism does not hold even for subatomic inorganic systems in
quantum mechanics, there is evidence that scientific knowl-
edge is quite compatible with moral responsibility and the re-
jection of the reductionism of judgments of right and wrong
to causally deterministic antecedent factors. This conclusion is
confirmed even by Professor Edwards, since at the end of his
paper, notwithstanding his scientifically questionable affirma-
tion of absolute determinism, he does not avoid passing a
judgment of "morally worse" on those who do not accept, and
of "morally better" on those who do accept, his thesis in judg-
ing human conduct.

Chapter 12

Determinism, Freedom, Moral .
Res^nsibility/ and Causal Talk

Arthur Pap, Yale University

It is the contention of the philosophers whom William James,
as reported by Professor Edwards, caUed "soft determinists"
(a most inappropriate name, I think, in its suggestion of "soft-
headed") that determinism is compatible with the occurrence
of free actions, since what distinguishes free from unfree ac-
tions is the mode of causation, not the absence of causes.
Some of them further explain the appearance of incompati-
bility as the result of an equivocation, a confusion of the
meanings of "compelled" and "caused." Now, I still think that
this diagnosis is correct and that the "problem" of freedom
and determinism can be resolved by careful analysis of such
slippery key expressions as "unavoidable," "having a choice,"
"could have acted differently," etc.

Professor Blanshard tries to explain how we can feel free
although our actions and the sUent decisions preceding them
are uniquely determined by antecedent events and laws. Now,
on whatever grounds a philosopher feels sure that our de-
cisions, like all events, are "necessitated," I fail to see why our
"feeling free" calls for explanation if one believes in deter-
minism. Innumerable times in the past I have verified that my
decisions were followed by the actions I decided to perform;
hence I ascribe to myself the power "to do as I please." That
I feel free just means that I believe on good grounds that —
within limits, of course — I can do as I please. But this does
not even seem incompatible with a causal determination of
one's decisions unless it is fallaciously assumed that a deter-
mined (caused) decision eo ipso ceases to be a decision; in
other words, that "caused decision" is a contradiction in terms.
It seems to me that what requires explanation is not the feel-
ing of freedom but the fallacious belief that a caused decision
is not really a decision. And I think I can supply this
explanation.

Suppose I were an omniscient psychologist engaged in pre-
dicting scientifically what I was going to decide to do a short
time later — say, that after five minutes I would decide to
scratch my forehead. Could I really make a decision while

212

Determinism, Freedom, Moral Responsibility, Causal Talk / 213

predicting it silently; that is, while thinking that I was going
to make it? I doubt it. Making a decision and predicting that
decision are mental states that exclude each other in the same
mind, since making a decision implies, by the very meaning of
the term, uncertainty as to what one is going to do. But is it
fallacious thence to conclude that my decision is not predict-
able by another mind. From the fact that "predicting my own
decision while making the decision" is meaningless it does not
follow that "predicting a decision" is meaningless or that "pre-
dicting my own decision before making the decision (and at
the time of deciding oblivious of my prediction)" is meaning-
less. We may also note that, if to a set of antecedents that
make the occurrence of a decision D highly probable we add
the antecedent expressed by "He predicts D on the basis of
those antecedents" or "He knows that someone else has pre-
dicted D," we get a set that may well reduce the probability
of D to improbability or even impossibility. But this is no
contradiction. In particular, knowledge of some determinist's
pretensions to be able to predict my decision may well pro-
duce in me the desire to falsify his prediction, but of course if
I falsify the prediction deliberately I have not thereby refuted
determinism: "same cause, same effect" is compatible with
"different cause, different effect!"

Professors Blanshard and Edwards seem to be agreed —
whatever their differences on other points — that the definition
of a free act as one that is caused by the agent's own choice
(i.e., as an act that would not occur unless the agent willed it)
leaves the real problem untouched; which is, in Professor
Blanshard's words, "not whether we can do what we choose to
do, but whether we can choose our own choice, whether the
choice itself issues in accordance with law from some anteced-
ent." Professor Edwards seems to mean roughly the same by
the words "choosing his own character — the character that
now displays itself in his choices and desires and efforts." But
I cannot attach any meaning to the expression "choosing my
^3wn choice." I understand what is meant by "I choose to read
this book" but not what is meant by "I choose to choose to
read this book." And what could be meant by "I chose my
own character"? One gathers that by "character" Professor
Edwards means a set of dispositions, including dispositions to
make such-and-such choices in such-and-such situations. Now,
surely it would sound queer to say, "I chose to acquire the
habit of smiling broadly at my boss whenever I should meet
him." It does make sense to say that what was at first a de-

214 / Determinism and Freedom

liberate, conscious effort eventually became an automatic
response. But then the object of one's choice, in the usual
sense of the word, is still a specific act ("I chose to put on a
smUe though I can't stand him"), not a disposition to act or
to decide to act in a specific way in such-and-such a situation.

Indeed I stUl maintain that when people, be they philoso-
phers or not, ask whether one's choice itself is free or con-
strained they confuse determination with constraint (or com-
pulsion). If by the assertion that the choice itself is not free —
though the overt act is free in the "soft" determinists' sense —
one means, in Professor Blanshard's words, that "the choice
itself issues in accordance with law from some antecedent,"
then one means simply that the choice is determined; but from
its being determined it does not follow that it is compelled in
the ordinary sense of "compelled." "I was compelled to choose
between death and poverty (by the gangster)" means that
either alternative, surrendering the money or not surrendering
it, had unpleasant consequences. On the other hand, we call
a choice free if the agent has no reason to believe that both
alternatives, doing A or not doing A, are intrinsically un-
pleasant or have unpleasant consequences; in other words, if
it is empirically possible to choose in accordance with one's
desire. The usual counterargument is that, nevertheless, if all
events are strictly determined, then / could not have made
any other choice than the one made; therefore the freedom of
choice is a mere illusion. But what does "could" mean? "I
went to the movies, but I could have gone to the concert in-
stead" clearly means that if I had preferred to go to the con-
cert I should have gone (contrast this familiar statement with
the equally familiar one: "Even if I had preferred to go to the
concert, I should not have gone, because it was sold out").

Now, certain philosophers and psychologists are not satis-
fied with the common-sense distinction between avoidable and
unavoidable acts; they ask whether the mental preference it-
self is avoidable. But if "avoidable" has the same meaning in
this new context, then "My preference for the movies that
night was avoidable" means "If I had preferred not to prefer
the movies to the concert, I should not have preferred them";
and this, I submit, is the same meaningless iteration as "choos-
ing to choose to read this book." If, on the other hand, "I
could not have preferred anything else" means that the ante-
cedents uniquely determined that preference, then those who
deny freedom of choice on the ground of determinism are
simply asserting the tautology that determinism is incompat-

Determinism, Freedom, Moral Responsibility, Causal Talk / 215

ible with indeterminism, not the interesting but false proposi-
tion that determinism is incompatible with freedom of action
and freedom of choice in any usual sense of these expressions.

Implicit in the above is my reaction to the conclusion of
Professor Edwards' paper that there is a sense of "moral
responsibility" — indeed, this is alleged to be the sense in which
the expression is used reflectively — in which determinism en-
tails that one is never morally responsible for anything. Pro-
fessor Edwards, following Professor C. A. Campbell, tells us
that for some people the statement "X is morally responsible
for doing A" entails, "X made (or chose) his own character."
Now, if "choosing my own character" is meaningless, as it is
relative to the ordinary use of "choose" (just as much as "eat-
ing my own character"), then Professor Edwards has at-
tempted to deduce a meaningful and startling conclusion from
a meaningless definition of the expression "morally respons-
ible." Suppose, on the other hand, that "I did not choose my
own character" means "My character is the product of en-
vironmental and hereditary factors that come into existence
quite independently of my will." I think Professor Edwards is
right in saying that many who believe in determinism refuse
to hold people "ultimately" responsible for their actions on
the ground that they did not choose their own character in this
sense. But I think the reason why such determinists take this
attitude is primarily that they fail to distinguish clearly be-
tween determination and compulsion; that is, they think it
follows from the premise that one's actions are causally deter-
mined that they are neither right nor wrong because they fal-
laciously and unconsciously substitute for this premise the far
different proposition that we are always compelled to act the
way we act, that all our actions are unavoidable.

Professor Blanshard agrees with the "soft" determinists that
determinism is compatible with moral criticism. But he sides
with the indeterminists in holding that moral actions, actions
inspired by a sense of moral duty, for example, are not "neces-
sitated" in the same way as habitual and impulsive acts. This
may be so, but I am not clear what the distinction between
causation (or necessitation) on different levels comes to
exactly. Professor Blanshard discusses deductive inference to
bring out the difference. My deducing "All A are C" from "All
A are 5" and "All B are C," he seems to maintain, is not
simply an example of association; the implication itself neces-
sitates the transition of thought from premises to conclusion.
Now, it seems to me that if implications have such causative

216 / Determinism and Freedom

force — directing the movement of thought the way a river bed
directs the flow of water — they must have it whether the
thinker sees them or not. Yet it is notorious that we may think
of a set of premises and not see what follows from them or
draw the wrong conclusion from them. One may counter that
one would see the entailed conclusion if one clearly grasped the
meanings of the premised statements. But this is either em-
pirically false or else an unexciting tautology — the latter if the
stipulated definition of "grasping the meaning of p" is that
one know all the necessary consequences of p. At any rate,
this is not a definition of "knowing the meaning of p" that
Professor Blanshard could accept consistently with his well-
known view that there are synthetic entailments. I admit that
where the entailment is fairly simple — syllogistic, for example
— there is a considerable probability that thinking of the prem-
ises will make one think of the valid conclusion. Suppose that
"If p and q, then ;-" expresses such an entailment, where r is
not a component of the premises (it must not be a modus pon-
ens entailment, for example). Then we are justified, I presume,
in making the corresponding causal statement: other things
being equal, thinking of p and q will cause a mind to thing of
r (the "other things" are such factors as being "minded" to
draw a conclusion from the premises). But I fail to see why
"cause" in this context should have a different meaning from
the one it has in the context "If a man has frequently witnessed
that the striking of a match is followed by a flame, then
a perception of the first event wUl cause a thought (an expec-
tation) of the second event" — which I suppose is an example
of what Professor Blanshard would call mere association. I
suspect that from the difference in the kind of implication cor-
responding to the causal sequence of mental events — neces-
sary implication in one instance, contingent regularity in the
other — Professor Blanshard erroneously deduces a difference
in the kind of causation. But it is svu-ely not logically necessary
that, where p and q entail r, the thought of p and q should be
followed by the thought of r. Otherwise we should all be per-
fect logicians.

In its semantic insight Professor Black's paper impressed
me as the highlight of the New York University Institute. I
agree completely with nearly everything in it, especially with
the argument that to deduce from the proposition that an
act is caused that it is not free is to be guilty of what amounts
to self-contradiction; that free acts like lifting a glass of beer
to one's lips are jxist the sort of events with reference to which

Determinism, Freedom, Moral Responsibility, Causal Talk / 217

"cause" is ostensively defined in the first place. There is only
one point Professor Black made that leaves me unconvinced.
Regularity of sequence, he says, is presupposed rather than as-
serted by a singular counterfactual like "If he had not lifted
the glass, it would have remained motionless." Whatever the
exact definition of the assertion-presupposition distinction may
be, I am sure that Professor Black, like Stravi'son, so uses
"presupposition" that the falsity of a presupposition does not
entail the falsity of the assertion that presupposes it. For ex-
ample, if while pointing at Mr. X, I say, "That American is
very rich," I presuppose — do not assert — that Mr. X is an
American; for if it turned out that he was not an American
but very rich, the proposition I asserted would not be refuted;
it is just that a false assumption about Mr. X led me to de-
scribe him incorrectly.

Now, let the relevant generalization that according to Pro-
fessor Black is presupposed, not asserted, by a person utter-
ing the counterfactual "If he had not lifted the glass, it would
have remained motionless" be: in situations like that one (or
perhaps in all situations whatever) a glass does not move to-
ward a man's lips unless someone — usually that very man —
makes it move that way. Is is really true that no amount of
disconfirmation of this generalization would lead us to retract
the counterfactual as false as long as we saw the man's clasp-
ing of the glass followed by the approach of the glass to his
lips? I very much doubt it. If I experience repeatedly that after
I clasped a glass an invisible force pushed the glass to my
mouth though I made no effort whatever, and other people
reported similar experiences, I should say to myself: Perhaps
it was the same way in that case! Perhaps he made no effort at
all; and even if he did make an effort, perhaps the same effect
would have taken place if he had made no effort. Professor
Black may reply that, if the generalization is indeed implicitly
asserted when we utter the counterfactual, it is inexplicable
how we can be as certain of the truth of the counterfactual as
we normally are. But I think that this high degree of certainty
can be accounted for without difficulty by one who stands by
the regularity analysis of singular causal statements. We usu-
ally apply the terms "cause" and "effect" to changes that
somehow catch our attention. And we use these terms in such
a way that only a change can be said to cause a change. Hence
it is analytic to say that only a, preceding or simultaneous,
change can have caused the observed movement of the glass;
for example, the drinker's passage from a state of indifference

218 / Detenninisni and Freedom

toward the beer to a state of desiring it. At once, then, factors
that remained constant are ruled out: the movement of the
glass could not, for example, have been caused by anybody's
perception of it, for it was perceived all along before it moved,
etc. True, many other changes preceded the effect: thus, some-
one near-by laughed shortly before it moved. But as we know
of many instances when similar changes were not followed by
a similar effect, they are likewise ruled out as possible causes.
My point is that the very regularity analysis of the meaning
of "cause" justifies an instantaneously and unconsciously per-
formed eliminative induction that bestows a high probability
on the singular causal statement.

Chapter 13

Some Eqyivo€al-ions of the

Noflof^ of Responsibility

Alfred Schutz, New School for Social Research

Our discussion of the problem of responsibility was mainly
concerned with the question on what grounds a person might
be held answerable or accountable by law or from a moral
point of view for something he did or omitted to do. The con-
sequence of responsibility, in this sense, is the infliction of
punishment, if we take this term in a sense broad enough to
include reprehension, criticism, and censure. But even in this
sense the notion "to be responsible" may mean two different
things: on the one hand, a man is responsible for what he did;
on the other hand, he is responsible to someone — the person,
the group, or the authority who makes him answerable.

This distinction between "being responsible for" and "being
responsible to" becomes of particular importance if another
equivocation of the notion of responsibility is taken into ac-
count, namely that between its use in terms of the third (or
second) person and in terms of the first person. It is submitted
that the notion of "responsible" is an entirely different one if
used in a proposition of the type "This person is responsible
for this and that" and in a proposition of the type "I feel
responsible for this and that (e.g., for the proper education of
my children)." It is further submitted that these two notions
of responsibility cannot fully coincide and that any philo-
sophical analysis of the problem of responsibility must remain
incomplete without taking into account its subjective aspect.

In using the expression "the subjective aspect" about the
notion "feeling responsible" in terms of the first person, we
adopt an unfortunate, but by now generally accepted, termi-
nology of the social sciences, which distinguishes between the
subjective and the objective meaning of human actions, hu-
man relations, and human situations. It was Max Weber who
made this distinction the cornerstone of his methodology.
Subjective meaning, in this sense, is the meaning that an action
has for the actor or that a relation or situation has for the
person or persons involved therein; objective meaning is the
meaning the same action, relation, or situation has for any-

219

220 / Determinism and Freedom

body else — a partner or observer in everyday life, the social
scientist, or the philosopher. The terminology is unfortunate
because the term "objective meaning" is obviously a misnomer
in that the so-called "objective" interpretations are in turn rela-
tive to the particular attitudes of the interpreters and there-
fore, in a certain sense, "subjective."

To elaborate on the difference between the subjective and
the objective meaning of responsibility would require a rather
lengthy analysis. We have to restrict ourselves to some scanty
remarks. If I feel merely subjectively responsible for what I
did or omitted to do without being held answerable by another
person, the consequence of my misdeed will not be reprehen-
sion, criticism, censure, or another form of punishment in-
flicted on me by someone else, but regret, remorse, or repent-
ance— or, in theological terms, contrition and not attrition.
The resulting states of grief, anguish, or distress are marks of
the true sense of guilt that is phenomenologically something
entirely different from the "guilt feeling" in psychoanalytic
terminology. It is the outcome of the feeling of being respon-
sible for something done or left undone and of the impossibil-
ity of restoring the past. Orestes in Aeschylus' Eumenides was
not redeemed before the goddess had reconciled the Furies,
although the judges of the Areopagus had placed an equal
number of white and black balls into the urn. In our times
we find certain eminent scientists suffering under a deep-rooted
sense of responsibility for having co-operated in the produc-
tion of atomic weapons, in spite of the honours bestowed on
them by a grateful government. On the other hand, the law
might hold me answerable for an act that my personal sense
of responsibility motivated me in performing (Antigone's
conflict is an example). And here the distinction between be-
ing responsible for something and being responsible to some-
one appears in a new light. I may agree with the other's ver-
dict that I am responsible for a particular state of affairs but
maintain that I feel accountable for my deed merely to God
or my conscience but not to my government.

These are merely examples for the complicated underlying
dialectic of t'ne subjective and the objective meaning of re-
sponsibility. But the same dialectic underlies the meaning of a
norm for the norm-giver and the norm-addressee. Any law
means something different to the legislator, the person subject
to the law (the law-abiding citizen and the lawbreaker), the
law-interpreting court, and the agent who enforces it. Duty has
a different meaning as defined by me autonomously and as im-

Some Equivocations of the Notion of Responsibility / 221

posed on me from outside. The whole question of determinism
in law and ethics will have to be answered in a different way if .
formulated in subjective or objective terms.

The distinction between the subjective and the objective
meaning of laws, values, morals, and responsibility merely
from the point of view of the individual can also be made on
the level of group relations. Adopting Sumner's classical dis-
tinction between in-group and out-group, it can be said that
responsibility, for example, has a different meaning if an in-
group acknowledges responsibility for its acts and holds some
of its members responsible, or if an out-group makes the in-
group and its members responsible for misdeeds. It is one
thing if, in the Nuremberg trials, the Nazi leaders were held
responsible by the Allied Powers, and quite another thing if
they were held answerable by the German people.

Chapter 14

Dennis Sciama, Trinity College,
Cambridge University

Professor Blanshard raised the question of our feeling that
we have free will. He suggested that we are under an illusion
and mentioned some psychological reasons for our mistake.
I agree that we may be under an illusion, but I should like to
suggest that the reasons may be logical rather than psychologi-
cal I have in mind the fact that a computing machine, for
instance, cannot know all about itself for the usual self-
referential reasons. Similarly a brain cannot know all about
itself and may therefore be unaware of some of the factors
determining its behavior. This might then lead to the illusion
that we have free will — an illusion that no amount of intro-
spection could dispel. (Professor Feigl told me that this idea
has been elaborated by Karl Popper in an article in the
British Journal of the Philosophy of Science ^ — an article I
have not seen.)

n

Professor Munitz raised the question: What is knowledge?
Perhaps a scientific comment on this question might be useful.
If you walk into a laboratory and find a bar of metal that is
much hotter at one end than the other, you infer that the bar
must have been heated at one end. (The problem of justifying
this inference is interesting, but I cannot go into it here.) In
other words, the bar carries some information about the past.
In a specialized sense the bar can be said to "know" some-
thing about the past. To put it more abstractly: a macroscopic
system that is substantially far from equilibrium carries infor-
mation about the past and can be said, in a specialized sense,
to "know" something about the past.

Of course we usually want to mean something more than
this when we use the word "knowledge." We want the system

1 "Indetermmism ia Quantum Physics and Classical Physics" (1950), I.
222

Observations / 223

to know that it knows. This can be achieved with the human
brain if we follow those theorists who suppose that inside the
brain there is a neural circuit that partially maps the state of
the brain as a whole. The brain then carries information by
being far from equilibrium, and this information is mapped
onto the special circuit, whose deviation from equilbrium en-
ables the brain to know that it knows.

m

Professor Lande raised the question how we are to under-
stand the existence in nature of random or nearly random
processes. Professor Frank suggested that one need only sup-
pose that fields of force are sufficiently complicated. This is
the conventional answer, but I agree with Lande that there is
more to it than that. I do not know the answer, but I should
like to mention some of the relevant considerations.

As we trace back in time the world line of each atom par-
taking of the random motion, we seek some moment at which
we can say: the randomizing element is introduced here. This
forces us to consider the cosmological problem concerned
with the state of the universe a long time ago. Two possibili-
ties are usually discussed in this connection. The first is that
there was a singular moment about five billion years ago when
all the material in the universe was crowded together at an
infinitely high density. One must then assume that there was
the required amount of randomness present at that time, a
procedure that is perhaps somewhat ad hoc. The second pos-
sibility is that the universe has the same large-scale appear-
ance at all times, the dilution arising as the expansion of the
universe is compensated by the continual creation of new
matter. On this view, when one traces back the world line of
an atom one comes to the point where it first appeared. Have
we the right to say that this appearance is a random process?
We cannot answer this question in the absence of a detailed
theory of the process of creation, but it seems to me that this
is the question that must be answered (or the corresponding
one for the high-density singularity) before Lande's problem
will be solved.

Chapter 15

Determinism and the
Theory of Agency

Richard Taylor, Brown University

I SHALL NEITHER provc nor disprove determinism. Instead, I
shall ( 1 ) give a precise statement of it, as I think Edwards
and Hospers understand it, (2) show that it does, as they
maintain, entail that men have no moral responsibilities, (3)
elicit the defects of the usual answers to this claim, (4) indi-
cate how a simple indeterminism supplies no better basis for
responsibility, and (5) sketch a theory of agency that I think
anyone insisting on moral responsibility must be driven to.

Determinism. Determinism is the thesis that whatever oc-
curs occurs under conditions given which nothing else could
occur. Indeterminism is simply the minimum denial of this,
viz., that at least some things occur under conditions given
which something else could occur instead. But these state-
ments need to be made precise.

The modal term "could" expresses causal contingency,
which I shall define in terms of causal or nomical necessity.
There is, that is to say, a clear and common sense in which,
for examole, a man who has been decapitated necessarily dies,
or can not go on living; or, for another example, water heated
to a certain point under certain conditions has to boil, though
no logical necessities are here involved.

Given, then, this sense of necessitation, we can define the
other modal words in terms of it and reformulate determinism
and indeterminism accordingly. Thus:

e is impossible = — e is necessary,

e is possible = — ( — e is necessary),

e is contingent = — (e is necessary) • — ( — e is necessary),
where e stands for any event whatever, and — e for any
event incompatible with e. These definitions show that the
possible and the contingent are not coextensive.

Determinism, then, is the thesis that in the case of any true
statement of the form "e occurs" the event whose name or
description replaces "e" i" ca'""'!!" nccessitat«*d, nf ■t con-
tingent. Indeterminism is the thesis that in the case of some

224

Determinism and the Theory of Agency / 225

true statement of that form, the event named or described is
contingent.

Responsibility. Edwards and Hospers believe that determin-
ism is incompatible with responsibility and obligation on the
basis, I believe, of the following argument. It is assumed (a)
that responsibility and obligation, in their strictly ethical sense,
if they have any application at all, figure only in the context
of human conduct, not in that of the behavior of animals, and
(b) that a necessary (not sufficient) condition for ascribing
this responsibility to a man for what he has done, or obliga-
tion for what he has yet to do, is that he could have done, or
could do, something else; that is, that the occurrence for
which he is responsible or obligated is contingent. But (c)
this condition is never fulfilled. Hence (d) no man has ever
been morally responsible for anything he has ever done, or
wiU ever be morally obligated to do anything else. A coroUary
of this is that the notions of "ought" and "ought not" have
no application to human conduct in any sense in which they
do not equally apply to the behavior of animals.

"Soft determinism." Determinists unwilling to accept this
conclusion have tried several rejoinders, of which I shall cite,
and reject, the four most common.

1. It is said that, even assuming determinism, the necessary
condition for responsibility is often fulfilled, for to say that
an agent could have done otherwise means only that he would
have done otherwise had he chosen to.

But this neglects the fact that, if determinism is true, he
could not have chosen otherwise. Indeed, by this kind of
argument, one could say that, though a man has died of
decapitation, he did not have to die, that he could have lived
on — meaning only that he would have lived had he somehow
kept his head on! And this is hardly the sort of contingency
we want.

2. Again, it is said that a sufficient condition for ascribing
moral responsibility is that an agent act from deliberation
with knowledge of the consequences, and that this condition
is often fulfilled.

But this presupposes the necessary condition for such re-
sponsibility that according to determinism is never fulfilled,
for, as Hart reminded us (citing Aristotle), it makes sense to
deliberate only about things that are, or are believed to be,
contingent. This point becomes clear if we remind ourselves
that, according to determinism, not only is a man's behavior
causally necessitated (among other things, by the course of

226 / Determinism and Freedom

his deliberations), but so also is every step and detail of his
dehberations, and so also are his beliefs (true or false)
about the future, and hence his beliefs concerning the effects
of his actions. Under these conditions one can no more
ascribe moral responsibility to a man than to a robot who
"deliberates," and then "acts," in response to our pushing
various buttons (labeled "deliberate" etc.), every step of the
chain then following by causal necessity.

3. Determinists sometimes say that we are, after all, re-
sponsible only for our acts, not our intentions, choices, or
decisions. A man is not punished or rewarded for deciding to
do something unless he then goes ahead and does it, and since
even advocates of "free will" concede that our acts are caus-
ally determined (by our choices or "wills," for instance),
there is evidently no absurdity in being held responsible for
what is determined.

But an indeterminist is not likely to concede that all our
acts are causally determined. Moreover, this view conceives
of responsibility only in terms of reward and punishment, con-
fusing moral responsibility with corrigibility. What I am re-
ferring to as moral responsibility comes out more clearly if
we consider cases in which no questions of law, no questions
of benefiting or harming others, and no questions of reward,
punishment, or retribution are at all involved. If, for instance,
an intelligent man studies what is in fact a valid philosophical
argument, understands it, accepts the premises as true and the
reasoning as valid, and yet refuses to accept the conclusion,
there is no philosopher, save those who deny obligation al-
together, who would not say that he ought to accept the con-
clusion. But here no overt act is involved, but only a decision,
and no question of reward, punishment, or retribution comes
into the picture, much less one of legality. It cannot be true,
therefore, that men have no responsibilities or obligations
with respect to their decisions, unless it should turn out that
they have no responsibilities or obligations whatever.

4. Finally, many determinists have said that moral respon-
sibility consists only in amenability to a change of behavior
through the force of real or anticipated rewards or punish-
ments; in other words, that responsibility simply consists in
corrigibility — a view that not only is compatible with deter-
minism but presupposes it.

This definition of responsibility, however, violates what was
assumed at the outset — namely, that lower animals have nei-
ther moral obligations nor responsibilities. The behavior of

Determinism and the Theory of Agency / 227

almost any sentient thing — rodents and fish, for instance — is
alterable by the stimulus of reward or punishment. Moreover,
this queer conception of moral responsibility and obligation,
in addition to applying to situations to which moral concepts
do not apply, fails to cover cases of the sort we just con-
sidered. For when one says that a man ought to accept a con-
clusion, in the light of probative evidence known to him, he
does not mean merely that he can be induced to do so by
threat or reward; indeed, the obligation might hold when this
condition does not.

I regard it as reasonable, then, that if determinism is true
no man has ever been morally responsible for anything he has
ever done, and no man ever will be under any obligation to
do anything. This is a painful conclusion to accept, particu-
larly in view of the fact that, if one does accept it, one can
try to persuade others to do so only by threats, blows, or
arguments, but can never say that tiicy ought to accept it,
even if it is proved. But the conclusion may well be true none-
theless, for it seems to be entailed by what most philosophers
regard as obviously true, viz., determinism.

Indeterminism. The denial of determinism, however, seems
no more compatible with moral judgment than determinism,
for it would seem to rob human actions, in Liebnitz's phrase,
of any "rhyme or reason." Since this thought is fairly familiar
I shall not elaborate it but only illustrate it.

Suppose an agent so constituted that his actions are deter-
mined by the numbers that turn up on a roulette wheel, and
suppose, further, that the wheel obeys no causal laws what-
ever, so that its behavior is unpredictable in principle. Now,
it would be plainly irrational to consider an agent so consti-
tuted morally responsible for those acts or obligated to per-
form others, since they are obviously utterly beyond his con-
trol; yet this situation corresponds exactly, so far as moral
judgment is concerned, to that of an agent whose acts are
quite undetermined. -^

Agency. To salvage moral responsibility one must resort to
certain odd metaphysical notions that have long since been
out of fashion and that are admittedly most difficult to com-
prehend clearly. What is needed, that is, is a view according
to which (a) there is a reason for everything that happens,
but {b) some such happenings — viz., some human acts — are
contingent. The only way of satisfying these seemingly in-
compatible requirements is to suppose that an act for which
an agent is responsible is performed by him, but that he, in

228 / Determinism and Freedom

turn, is not causally necessitated to do it. Now, this does, I
think, accord with what men take themselves to be — namely,
agents (cf. Latin agere) or beings that act rather than things
all aspects of whose behavior are the causal consequences of
the way they are acted upon. It now remains to elicit just
what this theory involves, and see whether it is compatible
with responsibility.

First, then, it involves the conception of a self or person
(i.e., an agent) that is not merely a congeries or series of
states or events, for on this view it is an agent who performs
certain acts (i.e., who acts) rather than states or events in his
history that causally determine them — these states or events
being, presumably, if not things done by himself, then simply
the causal consequences of other states or events, whether of
his own history or that of other things.

Second, it involves an extraordinary conception of causa-
tion, according to which something that is not an event can
nevertheless bring about an event — a conception, that is, ac-
cording to which a "cause" can be something other than a
sufficient condition; for if we say that a person is the "cause"
of his act, we are not saying that he is a sufficient condition
for its occurrence, since he plainly is not. We must accord-
ingly not speak of an agent as causing an act, since "being a
cause" ordinarily just means "being a sufficient condition,"
but rather of his originating it or, simply, of his performing it
— in a manner in which things in the physical world, so far
as we know, are never done or brought about. And this is
evidently the conception of Aristotle, who spoke of living
things as "self -moved," It is also what later philosophers, like
Thomas Reid, meant by "active power," viz., the power to
act without being acted upon, and it may be what Kant meant
when he obscurely spoke of a "noumenal" self that is free.
I Now, both of these conceptions — that of an agent as dis-
tinct from the states or events of his history, and that of
performing as distinct from being a sufficient condition — are
certainly odd and hard to conceive of clearly. Indeed, a phil-
osopher could not be accused of stubbornness if he preferred
to give up moral responsibility to embracing these two no-
tions. But I am sure that only by accepting them can one also
accept the notions of moral responsibility, obligation, and
what Professor Hook referred to as "dignity."

It still remains to see, however, whether this conception of
agency is compatible with responsibility. To show that it is,
it needs to be shown that on this view (a) some human acts

Determinism and the Theory of Agenqf / 229

are contingent, in the sense defined; (b) animals are not ren-
dered morally responsible; and (c) acts do not arise "without
rhyme or reason," i.e., are not capricious.

With regard to the first point: some acts are contingent on
this view, for they are not simply the causal consequences of
antecedent conditions. Now, it will be tempting to say that
there must be sufficient conditions for an agent's doing just
what he does, but this simply begs the question, being just
what the theory denies. There are certainly always conditions
under which any event occurs, but such conditions do not
in all cases necessitate just that event to the exclusion of any
other; otherwise there would be no such thing as an act, nor
would anything ever be done. We may further assume that for
j any act that is performed, there are reasons why it is per-
formed; but such reasons need not be causal conditions.
Rather, they may be, for example, motives or purposes, which
lare not sufficient conditions. To say, for instance, that an
agent acted in a certain way in order to achieve a certain
purpose is to give an explanation, but not a causal one, for his
conduct. And if it is now insisted that there must, in any case,
be conditions sufficient for an agent's having just such pur-
poses and motives as he has, this may or may not be true
(I think it is not); it is in any case irrelevant, for it would
mean that only his purposes and motives — but not thereby
his acts — are causally determined.

Secondly, as to the question whether this theory would
render animals morally responsible: it evidently would not.
If animals are "self-moving," as Aristotle thought, they do
indeed satisfy a necessary condition for responsibility; but it
does not follow that they satisfy any suflftcient condition for it,
and, in fact, they evidently do not.

Finally, as to the question whether this theory avoids capri-
ciousness in human acts: it plainly does not, // by "capricious"
nothing more is meant than "contingent." That is, it does
deny that there are conditions sufficient for the occurrence of
all events that occur. But it does not deny that there is an
explanation or reason for every event, as we have just seen,
for there are ways of explaining a man's conduct otherwise
than by a recitation of causal conditions. The concept of
agency, then, is quite unlike that of a thing whose behavior
is arranged to coincide with the random selections of a rou-
lette; for, assuming the wheel to be causally undetermined,
there is no ultimate explanation for the roulette's behavior,
whereas there is for the agent's. Moreover, saying of an agent

230 / Detenninism and Freedom

that he acts makes sense; but we cannot conceive of a wheel
— no agent at all — as "deciding" what is to be "done."

Conclusion. I do not claim to have proved a theory of
agency, but I believe I have shown that, if it is intelligible, it
renders moral responsibility possible. The conditions of moral
responsibility can thus be elicited, in terms of agency, as
follows.

Consider a situation in which some object O grasps a knife
and cuts off a man's hand. Now assume: (1) there were no
conditions sufficient for this event — i.e., it was contingent;
(2) O is an agent, e.g., a man; (3) the event described is an
act of O's and not, for example, a reflex; (4) O realized,
while contemplating the act, what it consisted in, and (5) he
knew what its consequences would be, and that they would be
evil.

A. Each of these assumptions can be true, and they can all
be true together; that is, there is no proposition known to be
true with which the conjunction of these five is causally or
logically incompatible.

B. If (1) were false, O would not be morally responsible
for the event described.

C. Hence neither (2), nor (3), (4), nor (5) is a sufficient
condition for responsibility; nor are these four together suffi-
cient, except as they may presuppose (1).

D. But (1), (2), (3), (4), and (5) are each a necessary
condition for moral responsibility for this event, and

E. Together they are sufficient.

If one denies E one must, I believe, either deny B, as
Edwards and Hospers do not, or else deny A, which would
seem arbitrary and implausible.

Chapter 16

Commera Seiise @n4 Beyond
Paul Weiss, Yale University

A GOOD DEAL of Contemporary discussion of the problem of
determinism and freedom rests on a number of unexamined
but rather dubious assumptions. No untU these are brought
to the fore, and either modified or replaced, can there, I think,
be much hope of progress toward the solution of this per-
plexing basic issue. One of the conspicuous virtues of Mr.
Blanshard's and Mr. Black's papers is that they so evidently
make these assumptions. Despite their apparent difference in
outlook and conclusion, the two are substantially in accord.
Their quarrel is an intramural one; the basic issue is blurred
and avoided by both with equal success.

A surprisingly large group of thinkers today suppose that
untutored common sense is beyond legitimate criticism. Many
of those who make this assumption deny that philosophy has
any problems of its own. There is, according to them, only the
problem of understanding how it is that philosophers should
think there are problems. Mr. Blanshard and Mr. Black are
happily free of this folly, but they, like the others, are in
accord in their acceptance of the testimony of untutored com-
mon sense. But it is surely the case that there is no single
coherent, consistent body of common-sense truth. The com-
mon-sense man, whether don or plain citizen, mixes shrewd
observation and practical wisdom with folly, superstition,
prejudice, and convention. Men of good common sense to-
day believe in the efficacy of individual prayers, in the wisdom
of having ministers of warring nations look to God for aid,
in mental telepathy, in a double sexual standard, in thrift, in
the agreement and profundity of scientists, and so on. It is
good common sense to claim that chalk is white because it
is seen to be so. But when twilight comes and the chalk ap-
pears grayish it is also good common sense to insist that the
chalk remains white, though it is seen not to be. The state-
ment "If I don't wear rubbers when it rains, I'll catch cold"

231

232 / Determinism and Freedom

is for many a robust common-sense man as root and firm as
any other and could be treated as a paradigm.

Any view that wholly abandons common sense is at best
a fiction or a fantasy. Any view that refuses to examine it is
at best uncritical and dogmatic. Reflection and reason require
one to stand somewhere between these two extremes. Taking
into account whatever else needs explanation, one should try
to forgo accounts that do maximum justice to what is daily
experienced. This maximum justice is consistent with the
abandonment of this or that phase of common sense if it is in
conflict with a larger body of coherent, clarified knowledge.
We ought to begin with some such simple fact as that a man
acts or, more specifically, that certain events ensue upon his
behaving in certain ways and, as experience shows, do not
ensue when he does not so act. We will then not know yet
whether the act was freshly initiated and freely carried out.
To achieve this knowledge we go beyond common sense in
order to obtain more abstract, consistent, and comprehensive
accounts than common sense can provide.

It is good common sense to say that whatever occurs is de-
termined, and it is good common sense to say that men act
freely. Were a systematic treatment of man in his diverse en-
terprises and an understanding of time, motion, causation,
mathematics, and science to make determinism the most co-
herent and illuminating outlook, we should have to conclude
that, though common sense often says that a man acts freely,
this would have to be understood merely as shorthand for the
remark that he is the occasion, or the most proximate source,
of an energy expended for the transformation of some objec-
tive combination of facts. On the other hand, if a satisfactory
account of man and these other matters required an acknowl-
edgement of an element of chance, novelty, or freedom in
the universe, we should have to conclude that the Sequent
common-sense assertion that all acts are determined reflects
the failure of common sense to make use of fine discrimina-
tions, its refusal to pursue reflections, and its incapacity to
reason in the precise fashion that is necessary if subtle truths
are to be grasped.

The contention that men are determined and the contention
that they are free seem opposed. They are not. Determinism
applies to what has happened when all the conditions are
already present and fulfilled. Freedom applies to what is hap-
pening and will happen; it concerns the creation of new con-

Common Sense and Beyond / 233

ditions and thus of consequences that until then have not
been necessitated.

II

Most discussions of freedom do not distinguish among ( 1 )
freedom from, (2) freedom to, (3) freedom for, and (4)
freedom with.

1. Whenever men are subject to unusual or unconventional
restraints we think of them as being compelled; when those
constraints are removed we think of them as freed, and speak
of them as free. Thus we say that the released slave is a free
man.

2. Freedom to is the power to act either inwardly as a
being of intent or outwardly as one who can publicly express
his wishes or carry out his obligations. Thus we say that he
who wishes to take drugs, and can, is free to take them; and
that he who ought to pay his debts and has the money to do
so is free to pay them.

3. Freedom for is the power to commit oneself to an end
and to work to bring it about. This end may be set by society
or the state; one's commitment to it might be a function of
heredity and training. But the capacity to put oneself in a
position to focus on this end and engage in activity to bring
it about is to have a freedom for. Thus we say that men are
at their best when they have been made free for a civilized life
of leisure.

4. Finally, we are men in a society and men in a cosmos,
and whatever freedom we may express is largely futile or
frustrated if it does not intermesh with that exercised by
others. Thus we say that we are genuinely free only when we
are in harmony with equally free fellow men.

Each of these types of freedom has a number of distinct
cases. Most interesting perhaps are those that fall under the
second and third types, for these are most germane to ques-
tions of morality and ethics. It is most helpful, I think, to
distinguish among freedom of preference, of choice, and of
will. Freedom of preference refers to the selection of some
means to an accepted goal, good or bad. Freedom of choice
concerns the formation of a single unity of means and end,
where the end promises to make good whatever loss in value
the means involve. Freedom of will, finally, is related to the
adoption of a final prescriptive good in such a way as to
require a fresh approach to whatever means there are. Free-
dom of preference allows for a determinism dictating which

234 / Determinism and Freedom

among a number of goals will be found most agreeable, and
which means most appealing; it is a freedom connected with
the joining of goals and alternative means, with a consequent
mutual alteration in their import and a possible substitution of
means. Freedom of choice allows for a determinism dictating
which means will be favored; but it also requires a freedom
through which a compensating end is adopted and the means
previously favored re-evaluated and perhaps rejected. Finally,
freedom of will allows for a determinism that dictates why,
and perhaps even when, the will should be exercised, while
leaving free the exercise of the will in this way or that, with
this or that consequence.

There are surely other meanings of freedom besides these.
But enough has been indicated, perhaps, to show that we can-
not settle the problem of determinism and freedom until we
have decided just what type of freedom, and thus what anti-
thetical sense of determinism, we wish to consider.

Ill

It is commonly assumed that prediction and determinism
are either equivalent or identical, or that one follows neces-
sarily from the other. But none of these suppositions are cor-
rect. Prediction is a prediction, a saying in advance what is to
ensue. In the ideal case it consists in the deduction, through
the agency of mathematics and logic, of the nature of the
effect that rationally follows from the given condition. Now,
all deductions are in terms of generalities, of expressions that
lack the completeness and concreteness characteristic of ac-
tual occurrences. Predictions thus relate only to a phase, an
aspect, of the world, to something generic, to be instanced by
what in fact occurs. The occurrence of the instance is not
encompassed in the prediction; it could have been determin-
istically produced, though in that case we should not have
been able to say anything about it in advance, or even per-
haps when it occurred. Conversely, predictability is compati-
ble with freedom. Predictability, since it concerns only kinds
of things, the generic features, the merely possible, allows
for the free occurrence of this or that particular instance.

Most thinkers confuse the problem of prediction with that
of determinism because they are phenomenologists in doc-
trine or in intent, and suppose that what scientists affirm to
be real, or what common sense observes, is all there is in the
universe. Yet a scientific account that requires the acknowl-
edgement of determinacies or indeterminacies in the world

Ck>inmon Sense and Beyond / 235

tells us nothing about the nature of things, but only about
that phase of existence isolated and pursued in common ways
by a community of thinkers. In the end we have no right to
say that anything completely beyond all knowledge is real.
But it does not follow that what is beyond the reach of sci-
entific instruments, scientific needs, scientific procedures, and
scientific tests is beyond the reach of aU cognition and out-
side the pale of meaningful and illuminating discourse. Poetic
insight, dialectical deduction, speculative construction, sys-
tematic explanation all make claims not to be dismissed out
of hand. The concurrence of these is required if we are to
know whether determinism or freedom holds sway in this
world.

A scientific indeterminism is compatible with an ontological
determinism; a scientific determinism is compatible with an
ontological indeterminism. The wise scientist makes no claim
that he is able to deal with or to know anything about the
universe outside the scope of his language, instruments,
experiments, or criteria. More often than not it is the philoso-
pher who hurries to convert a laudable scientific methodologi-
cal caution into an unwarranted claim that what science does
not know is either not knowable or nonexistent, and who
makes the uncritical contention that what science treats as a
desirable supposition or a genuine fact is to be urged as a
solid and unquestionable basic truth.

IV

The importance of the problem of ethical responsibility
makes most thinkers attend to the question of freedom only
as it relates to man. Almost all writers in the western world
today are clearly, and sometimes even outspokenly, anthro-
pocentric. Confident that the world of nature is thoroughly
determined (except perhaps in the region of quanta phenom-
ena), they insist that man, at least if ethics is to have any
meaning and responsibility a locus, is somehow free. But their
position involves the cutting off of man from nature. It is a
view that is incompatible with the theory of evolution, and
in the end it deals with man as though he were "a kingdom
in a kindgdom." To be sure, man is different from other
beings; he has and can exercise powers denied to them. But
the powers that they do have must fall under categories com-
prehending his, if both he and they are to be parts of the same
cosmos.

236 / Determinism and Freedom

The nature of causation has been oversimplified to make it
seem as if there were only two factors to consider, the cause
and the effect. As a consequence the discussion of causation
has been riddled with paradox.

A cause must precede its effect. That is why history can
cover an extensive stretch of time. This means that the cause
cannot necessitate the effect. If it did, the effect would exist
when the cause did. What is normally termed a cause is only
an antecedent condition. The nature of the effect is defined
by it, and can often be predicted in the light of what we
know of it. But the cause does not produce the effect that in
fact ensues. The actual effect comes about as the result of an
activity that, taking its start with the cause, ends by producing
an instance of the predictable effect. The causal situation,
then, has not two but three components; the cause, the process
of production, and the effect. Since freedom is the doing of
something beyond the determination by a cause, the process
of production is evidently free. An actual effect is freely
produced inside a frame that necessarily binds together an
antecedent cause and a predictable type of effect.

These observations, it is hoped, may suffice to indicate that
we ought not to be content with uncriticized common sense,
with the hypotheses or conclusions of the sciences, or with
the traditional philosophic formulations of the problem of
determinism. We ought to take our start with common-sense
items, examine them critically, and then move on to what
they presuppose; that is, to a systematic account encompass-
ing man and nature and whatever else there is. If we are
willing to make provision for every facet of being and knowl-
edge, we shall soon find that no untutored common sense, no
rationalism, no anthropocentrism, no simple empiricism or
positivism, no theory of a radical opposition between deter-
minism and freedom, will do justice to the facts or to the
need to get a coherent explanation of them.

Chapter 1 7

H. Van Rensselar Wilson, Brooklyn College

There are two kinds of problems connected with causation:
epistemological problems and ontological or metaphysical
problems. Although they are interrelated in various ways, it is
possible for purposes of analysis to keep them relatively dis-
tinct, and in my opinion it is important that confusion be-
tween them be avoided so far as possible. It therefore strikes
me as unfortunate to speak (as Professor Barrett does) as
though predictability (an epistemological concept) were syn-
onymous with causal necessity (an ontological concept). Lack
of causal necessity would certainly entail lack of predicta-
bility; but I see no reason to assume that lack of predictability
entails lack of causal necessity. The fact that the epistemo-
logical difficulties in sociological, psychological, and many
biological situations preclude our knowing what all the spe-
cific relevant causal factors are in a particular case does not
warrant the conclusion that there are none. Present inability
to specify the values of a variable can hardly be construed as
evidence that no such values exist.

Professor Black used a very brief illustration that has a
bearing on this point. He remarked that when a penny is
tossed into the air it has to come down, but it does not have
to come down tails. Now, I have no quarrel with this remark
if it is simply meant to assert two epistemological proposi-
tions: (1) knowing only that the penny was tossed up, but
not knowing in what precise direction, with what precise
force, with what kinds of spin, how it lay in my hand before-
hand, what kind of surface it will strike, and so on, we are
warranted in asserting (without specifying when, where, or
how) that it has to come down; but (2) in the absence of
fuller information neither the statement "It has to come down
heads" nor the statement "It has to come down tails" is a
warranted assertion. But in the context of a discussion of
causal determinism one tends to construe Professor Black's
remark as denying that there is causal necessity in a particu-

237

238 / Determinism and Freedom

lar case of the penny's coming down tails, which is a very
different matter. The difficulties of measuring even approxi-
mately the momentum and spin of a tossed penny — to say
nothing of doing so precisely enough for safe prediction — are
obvious. But surely this does not alter the fact that, irrespec-
tive of anyone's knowledge or lack of it, the penny has a
specific momentum and spin, and that if all the data were
precisely known it would be clear that the penny "has to'.'
come down the way it does.

The fact that five hundred years ago no one could have
computed the penny's trajectory even if precise data had been
available, whereas today we can, is beside the point, of course.
If the trajectory is admittedly necessitated by computable
causal factors today, then similar trajectories were always
necessitated. The unavoidable analogy between (1) a fif-
teenth-century physicist in relation to today's physicist and
(2) today's neurologist (let us say) in relation to an imagin-
able twenty-fifth-century neurologist needs no elaboration.
The analogy proves nothing, to be sure; but it should give us
pause, restraining our tendency to equate absence of knowl-
edge of precise data and formulas with absence of existence
of precise states of affairs and objective correlations in fields
where they are not currently specifiable.

Causal necessity in a particular case, I would contend, is
holistic with respect to the relevant controlling factors rather
than fragmental with respect to those factors; that is, causal
necessity is a concept that is meaningful only as a function of
all the relevant causal factors (material, formal, and efficient),
whether known or unknown, taken collectively. Fragmentally
(i.e., when not even one relevant factor is taken account of)
neither assertion nor denial of causal necessity in a given case
would be warranted, although one could certainly form
hypotheses, for conceivable future testing, about what sorts
of currently unavailable data would need to become available
in order to complete the picture. A basic heuristic postulate
for determinism is that it is worth while without limit to
fonnulate and (as it becomes possible) to test experimentally
such hypotheses regarding currently missing causal factors,
since (the determinist assumes) every instance of apparent
lack of causal necessity is an instance of fragmental knowl-
edge of the relevant causal factors. The pragmatic effects on
the future development of science of proceeding (without
proof) as if this assumption were true and of proceeding as if
it were false are quite evident.

On Causation / 239

Professor Blanshard's suggestion that there are levels of
causation is one with which I agree, if I may construe it as
referring primarily to the levels of complexity of the relevant
causal factors in various types of situations, and secondarily
to the levels of concomitant epistemological difficulty in ascer-
taining without error or omission (1) what all the relevant
factors are about which specific information is needed before
we can specify the holistic causal necessity in such situations,
(2) the specific formulas that would adequately formulate
the relevant causal laws, and (3) the precise value of the
variables to be substituted in the relevant formulas in a par-
ticular case. For the tossed penny most of the data that
would be relevant at a more specific level are irrelevant when
we only wish to assert nonspecifically that it must come down
sometime, somewhere, somehow. We may therefore ignore
such data and still assert causal necessity without using the
concept (at this level) fragmentally.

When we assert that there is holistic causal necessity in the
penny's coming down tails in a specific case, the domain of
relevance is much vaster and the epistemological difficulties
far greater, but the relevant formulas and data are still theoret-
ically within the scope of current physics and current observa-
bility. In asking, however, whether the tosser "had to" decide
to toss the penny one enters a level of complexity and diffi-
culty where only fragmental consideration of the relevant
factors is currently possible, and from fragmental considera-
tion no specifiable necessity can emerge, any more than it
could from a fragmental consideration of the factors in the
penny's trajectory. This being the case, the pertinent questions
would seem to be not those about current knowledge and
predictability but rather the frankly speculative, as yet un-
tested, hypotheses and assumptions about the factors that
still elude us, entailing a projected program of further inquiry
that is believed (without proof, but equally without disproof)
to be worth pursuing in the hope of gradually reducing our
ignorance. In these terms I agree with Professor Blanshard (as
I understand him) that to assume, for animate and human
and societal cases as for inanimate cases, that holistic causal
necessity exists, as a function not merely of some or most but
collectively of all the revelant causal factors involved (regard-
less of their complexity and our current inability to specify
them exhaustively), makes more sense than it does to assume,
as indeterminists do, that no such causal necessity exists even
as a function of all the relevant causal factors, known and

240 / Detenninism and Freedom

unknown, or to assume (as Professor Barrett appears to do)
that absence of specifiable causal necessity in terms of a
fragmental consideration of those factors that happen to be
currently known is to be equated with absence of causal
necessity altogether.

II

We would all agree, I suppose, that the verb "to determine"
has two independent meanings which are not to be confused:
(1) to ascertain, or obtain information about; and (2) to
necessitate, or to be related to a dependent variable in such
a way as to render all but one of its values impossible. "Cur-
rently ascertainable" is surely not synonymous with "existing
in a definite state," nor is "epistemologically determinable"
synonymous with "ontologically determinate." And yet we
seem to find ourselves speaking at times almost as if "having
no currently measurable difference" were synonymous with
"having absolutely no difference," and even as if "present
human inability to determine (ascertain) the state of affairs"
were synonymous with "ontological indeterminacy of the state
of affairs."

Possibly the trouble arises from an unfortunate semantic
assumption that meaningful terms include only those for
which operational tests are currently performable and exclude
those for which such tests are conceived only and not cur-
rently feasible to perform. But so to restrict meaning is dis-
astrous to theoretical speculation. To forbid, by arbitrary
semantic fiat, hypothetical discussion of situations that cur-
rently preclude our ascertaining precise values for the rele-
vant variables, or for which we have not yet been able to
formulate any applicable differential equations, would seem
to exhibit a defeatist attitude that is utterly foreign to the
history of science.

When Professor Lande speaks about dropping ball bearings
from a chute onto a knife edge, he seems to make the tacit
assumption that the several occurrences are mathematically
identical, except for the random falling to left and right. He
would undoubtedly agree, however, that the most that could
be validly asserted about the objective state of affairs is that
the actual differences (if any) among the occurrences are too
slight to be detected by available measuring devices. But does
this justify the assumption that there are no differences?
Someone has suggested performing the experiment in a vac-
uum to avoid stray currents of air, I agree. But that is not

On Causation / 241

sufficient. Slight but definite deviations from mathematical
sphericity or from absolutely homogeneous density, even if
undetectable by available instruments, would presumably
affect the results, and even if such factors were completely
controllable, the knife edge itself would lose a few molecules
of material each time a ball struck it and the next ball would
therefore strike a knife edge of slightly different shape. Unless
one insists on equating "below the threshold of detectability
with currently available instruments" with "mathematically
infinitesimal," there is no relevance to the objection that in-
finitesimal differences in the conditions could not account for
finite differences in the outcome. I see no logical inconsistency
in positing the hypothesis that in such an experiment there is
no randomness in the outcome of really identical events, but
varying behavior under varjdng conditions in accordance with
perfectly definite formulas — a hypothesis that will become
progressively testable as measuring techniques become pro-
gressively more refined. No new or "hidden" variables are
here involved, but merely slight, uncontrolled differences in
the values of the familiar variables.

These slight, imcontroUed differences in the conditions are,
I suppose, the "phantom knife edge suspended above the
actual knife edge," to which Professor Land6 alluded. But I
see no other way out without a quite needless retreat from
essential scientific attitudes. A basic programmatic decision
is involved. One alternative is to accept "inherent randomness
of nature" as an explanation of the observed facts, in which
case no program of further experimentation and theoretical
refinement is called for, since one already has the "explana-
tion," such as it is. The other alternative is to act upon a heu-
ristic postulate that construes the phenomenon of apparent
randomness, wherever observed and quite unconditionally, as
a challenge to the experimenter to refine his measurements or
his theory or both imtil determinate orderliness replaces the
randomness — assuming that eventually it will, if he and his
successors persist. Instead of interpreting this assumption (as
Professor Munitz seems to do) as implying a belief that the
universe is like a humanly constructed "puzzle" that one as-
sumes to have a "solution" merely because puzzles are always
constructed that way, I would interpret it as inextricably
involved in the basic attitude of scientific curiosity itself,
without which there would never have been any science and
in the absence of which further scientific advancement would
stop cold. Without celestial mechanics the places where sue-

242 / Determinism and Freedom

cessive solar eclipses are visible give every appearance of
randomness. Without meteorology the weather has every
appearance of randomness. Is it not scientific defeatism to
assume that contemporary types of randomness are more in-
vincible than these?

At the risk of appearing utterly stupid to the experts in
quantum mechanics, I must still ask, How does anyone know,
beyond all doubt, that every photon is identical with every
other photon? Not long ago we thought that every hydrogen
atom was identical with every other, but then we discovered
deuterium and tritium. Is there any logical absurdity in the
hypothesis that when and if we reach a higher degree of pre-
cision we shall find (as often in the past) our current instances
of apparent randomness exhibiting determinate order after
aU?

Of course we might then discover new instances of appar-
ent randomness at the new level, and thus be faced with the
same choice again — either admit defeat and accept "inherent
randomness of nature" as the answer or act again on the
former heuristic postulate. But this should not surpise us. The
"erratic" orbits of Uranus and then of Neptune were not
treated as instances of randomness but as challenges. If some-
time, with more precise measurements, we find that Pluto too
is behaving in random fashion, would we quit or proceed as
before? Determinists and indeterminists might appear to
sanction opposite answers to such a question.

Ill

I make only two very brief comments on the papers of
Edwards and Hospers.

1. There may be some danger of mistakenly supposing
that an actual cause of an effect is somehow less a cause if it
can be shown to have prior causes of its own. But this would
be like supposing that I am somehow less the father of my
children because I in turn have a father. If I am really the
cause of my decisions, I am no less so after it is shown that,
in addition to being the cause of these and other effects, I am
also the product of earlier causes. Whatever moral responsi-
bility I have for my decisions, I have as their proximate cause,
and the responsibility is not diminished by showing that re-
sponsible selves (as well as irresponsible selves) have causes.

2. The relation of one's "self" or character to one's de-
cisions and actions is sometimes such that (a) a modification
of the self would result in a modification of future behavior

On Causation / 243

in similar sitiiations, and the self is modifiable by praise,
blame, and other manifestations of approval and disapproval
from other persons (including courts of law) and from one's
own conscience; while at other times it is such that (b) either
the self is not thus modifiable, or such modification would
not result in modified future behavior. In ordinary language
we call the person "responsible" in case a and "not respon-
sible" in case b. If Professors Edwards and Hospers are as-
serting that all human actions come under category b, then
I think the empirical facts are clearly against them. On the
other hand, if they agree that there are actions that come
under category a but simply refuse to call such actions "re-
sponsible," then I think linguistic usage is against them. Also,
I think they would then be imder obligation to introduce some
new terms of their own as a translation of the quite essential
distinction between a and b, which otherwise seems to dis-
appear.

INDEX

Index

Agreement, method of, 165

Ambiguity, 202

Analysis, 96, 160, 169

Analytic: expressions, 172n,
217; implication, 166, 217;
philosophy, 201

Anthropology, 211

Aristotle, 30, 128, 225, 228-29

Art, 51-2

Association of ideas, 26, 28

Augustine, St., 190

Austin, J. L., 105

Avoidability: as criterion of re-
sponsibility, 157ff, 214-16;
conditional, 158

Axiom: of retribution, 193; of
scientific method, 69, 70

Ayer, A. J., 118, 140-41

Barrett, W., 46, 237, 240
Barth, K., 190
Beardsley, E. L., 145
Benfey, O. T„ 67n
Bentham, J., 105, 106
Bergler, E., 132n
Bergson, H., 46
Black, M., 31, 46, 181, 196-97,

201ff, 216-17, 231, 237
Blanshard, B., 19, 47ff, 158,

170ff, 181-82, 196-97, 207,

212ff, 222, 231, 239
Bohm, D., 57
Bohr, N., 57flF, 89
Bom, M., 22, 57ff, 89
Boundary and initial conditions,

90-1, 185, 198, 204ff
Bradley, F. H., 118
Brandt, R., 149
Bridgman, P. W., 57, 76ff, 155,

183ff, 197-98, 201, 210

Bunge, M., 57, 69ff

Calvin, 190

Campbell. C. A., 122-23, 146-
47, 188-89, 215

Carnap, R., 26

Cassirer, E., 67-8

Causality: and anthropomorph-
ism, 37ff, 139-40; and con-
stant conjunction, 26ff, 40-1,
163ff, 202ff; and "making
something happen," 3 Iff; def-
nition of, 36ff, 130, 163ff,
197, 202ff, 215-16, 236ff;
language of, 43-4; levels of,
26ff, 146ff, 173-74, 239; me-
chanical, 202-3; mental, 26,
27, 118ff, 126ff, 167-68, 177,
226; metaphysical, 39, 227ff;
physical, 22ff, 37ff, 58ff, 81,
83ff, 163ff, 210ff; proximate,
164; related to choice, 37,
110, llSflF, 147-48, 177; re-
lated to punishment, 107-8,
156; tcleological, 202ff

Causal laws, 26, 58ff, 77, 95-6,
122, 164, 171-72, 228, 239

Chance: in history, 53; in phys-
ics, 65-6, 83-4, 185, 197-98,
204

Change, 43; as criterion of re-
sponsibility, 131; in definition
of "cause," 164-65, 217; in
definition of "event," 19

Chisholm, R. W., 157

Choice: and disposition, 213;
eflScacy of, as maximized by
law, 109ff; freedom of, 20-1,
95n, 119ff, 126ff, 157-58,
167ff. 172, 176flf, 187ff, 212ff,

247

248 / Determinism and Freedom

232ff; meaning of, 21; system
of, in criminal law, 109ff. See
also Freedom

Coleridge, S. T., 52

Collins, A., 120

Common sense, 183; criticism
of, 23 Iff

Complementarity, principle of,
61, 89

Completeness and incomplete-
ness: in mathematics, 49; of
physical knowledge, 160ff

Compton effect, 207

Consumer, 194

Continuity: postulate of, 88;
continuous creation of mat-
ter, 223

Contradiction, in mathematics,
49

Counterf actual conditionals, 39,
107, 171, 217

Creativity, 179; in art, 51-2; in
mathematics, 49, 50

Darrow, C, 105, 123, 138, 190
DeBroglie, L., 57, 66, 71
Deduction, 70, 205, 215, 234
Definition: and undefinability,
211; by stipulation, 33, 216;
formal and informal, 34; in
quantum theory, 81, 90,
204fT; ostensive, 217; paradig-
matic, 34ff, 201-2; pragmatic
criteria of, 166
Denotation, 161
Determinism: meaning of, 20,
44, 48-9, 81, 90-1, 160ff.
nOff, 204ff, 224ff; "soft" and
"hard," 117ff, 126ff, 189-90,
212ff, 224ff. See also Neces-
sity; Causality; Predictability
Dewey, J., 46, 177
Difference, method of, 165
Dignity, as basis of moral re-

sponsibility, 189, 228
Dilemma, of moral responsi-
bility, 158
Dispositions, as causes of moral

actions, 174
Ducasse, C. J., 160

Economy of threats, doctrine
of, 105ff

EddLugton, A., 23, 24

Edwards, Jonathan, 117, 178

Edwards, P., 117, 126, 131,
146fT, 152, 176ff, 187ff, 193,
199, 200, 211, 212ff, 224-25,
230, 242-43

Einstein, A., 23, 57, 64, 77ff,
201, 206ff

Electron: definition of, 73; posi-
tion and velocity of, 23-4,
73-4, 80, 86, 161, 184

Emergence, 166

Entailment: analytic and syn-
thetic, 216; in defining cau-
sality, 164, 216; of a theory,
208

Epistemology, 202ff, 237flf

Erewhon, 115-16

Event: definition of, 19; mental,
26, 161-62, 181; physical, 22,
161, 183, 196ff, 241; proba-
bility of, 66; uniqueness of,
161-62

Evidence: for determinism, 181-
83; for fatalism, absence of,
163; introspective, 173-75;
of causal efficacy, 31; of vol-
untary action, 113, 141

Excusing conditions : distin-
guished from invalidating
conditions, 100; in law, 95ff,
145ff, 149ff, 188ff, 193

Fatalism, 47, 120, 162, 178-79
Feigl, H., 172n, 222

Frank, P., 172n, 173n, 204. 223
Freedom: and indeterminism,
19-20, 160ff, 232ff; as ab-
sence of constraint, 37, 118-
19, 146, 167-68, 212; as qual-
ity of action, 38; as rational
control, 29, 30, 189; feeling
or sense of, 20-2, 29, 173,
182, 212, 222; of will, 24,
85, 95ff, 155-56, 162, 167-68,
178, 226, 233; types of, 233ff.
See also Choice
Freud, S., 20, 117, 128, 210

Galileo, 79
Galton, F., 21
Generalization: and causal

laws, 165; by abstraction, 165
God: and divine law, 122, 140;

and foreknowledge, 51, 180;

and prayer, 231
Godel, K., 49-50
Goodman, N., 202
Green, T. H., 118

Habit, related to moral respon-
sibility, 127-28, 136-37

Hall, A. J. B., 99n, 102

Hart, H. L. A., 95, 145-46, 188,
193, 225

Heisenberg, W., 22, 51, 57flf,
72, 80, 89, 186, 209

Hempel, C. G., 170

Hintz, H. W., 176

History, unpredictability of, 54

Hobbes, T., 118

Holbach, P., 117, 129

Hook. S., ix, 53, 180, 228

Hospers, L, 126, 148, 181ff,
193, 199, 200, 210, 224-25,
230, 242-43

Housman, A, E., 123-24

Hume, D., 118ff, 140-41, 163-
64, 179, 202-3

Index / 249

Indeterminism, see Determin-
ism; Freedom; Choice;
Chance

Induction: eliminative, 218;
role in causal inference, 40

Initial conditions, see Boundary
conditions

Insanity, as excusing condition
in law, 105-6

Interference patterns, 59

Introspection, 173-75

James, W., 46, 117ff, 212
Jennings, H. S., 161n, 166
Jones, E., 128

Justice, as basis of punishment,
102ff; cost of. 194

Kant, I., 29, 228

Khayydm, Omar, 117

Knowledge: distinguished from
information, 80-1; limits of,
74-5; meaning of, 222-23; re-
lated to determinism, 163,
205, 238ff; related to legal
responsibility, lOOflf; related
to virtue, 98ff; spectator
theory of, 186

Land6, A., 83, 89n, 183ff. 194,
197-98, 223, 240-41

Language: causal, 3 Iff, 43ff; of
morals, 42, 129, 141; ordi-
nary and scientific, 44, 156,
183; rules of, 33ff

Laplace, P. S., 50-1, 161

Law: as mechanism for maxi-
mizing efficacy of choice,
llOff; causal, see Causal laws;
criminal, 95ff

Leibnitz, G. W., 87, 227

Lemer, A., 193

Levin, M., 133n, 138n

250 / Determinism and Freedom

Liability, 98ff, 145ff

Logic, distinguished from ontol-
ogy, 78. See also Deduction;
Induction; Truth, logical

Lowes, i. L., 52

Mach, E., 90
Marxism, 53, 114
Materialism : and mechanism,

174; criticism of, 25
Mathematics: application of, to

physics, 59-60, 173, 202ff,

240-4 1 ; incompleteness of,

48-9
Matson, W. L, viii, 139
Maxwell, C, 206, 208
McCulloch and Pitts, theory of

trapped universals, 211
Meaning: analysis of, 32, 96,

160, 167; borderline cases of,

33; criteria of, 36-7, 240;

epistemological, 202ff; related

to culture, 68
Meaninglessness, 24, 140ff, 187,

223
Measurement, in physics, 61,

198
Mechanics: Newtonian, 62, 70,

76, 90, 161-62, 198; quantum,

57ff, 75flf, 83ff, 90flf, 183ff,

197ff, 201flf
Mens rea, in criminal law, lOlflf
Metaphysics, 39, 158, 171, 237
Michelson-Morley experiment,

208-9
Mill, J. S., 118ff, 140-1, 179
Modality, 225ff
Moore, G. E., 176, 211
Moral criticism, social utility

of, 149fif
Morality, 25; as basis of law,

lOlff; as sense of duty, 29;

feelings of, 124, 152; judg-
ments of, 122ff, 146ff; lan-

guage of, 42. See also Re-
sponsibility, moral
Munitz, M., 76, 183, 222, 241

Nagel, E., 196

Necessity: aesthetic, 182, 196;
and sufficiency, 39, 40, 164,
165, 182, 225; as compulsion,
127ff, 214; causal, 224, 225,
237; conditional, 39, 40; eco-
nomic, 114; etiological, 165;
historical, 53; logical, 27,
120, 167, 168, 172, 180, 181,
196, 216; mathematical, 50;
moral, 28; nominal, 224;
physical, 23, 24, 50, 203ff;
problem of meaning of, 48,
139, 170ff, 180

Neumann, L von, 63, 86-87

Neurath, O., 26

Neurosis: definition of, 131; re-
lated to moral responsibility,
126ff, 146

Newsom, C. V., 15

Newton, L, 206ff. See also Me-
chanics, Newtonian

Noel-Smith, P. H., 118

Nominalism, 202

Northrop, F. S. C, 201

Oedipal sources of motivation,

126-27
Orwell, G., viii
Owen, R., 117

Pap, A., 212

Parameter: "concealed," 62;

statistical, 198
Peirce, C. S., 46, 181
Philosophy: and analysis, 203;

distinguished from science,

58, 76-7
Photons, 59ff, 184, 242
Physicalism, criticism of, 25

Index / 251

Planck, M., 22, 57, 63, 70, 73,
91, 209

Plato, 115

Poincare, H., 51, 63

Popper, K., 78, 84n, 222

Possibility, causal and logical,
157-58

Pragmatism, in scientific
method, 78, 166, 180, 181,
200, 238-39

Praise and blame, 149ff, 168,
177, 187ff

Predestination, 178. See also
Fatalism

Predictability and impredicta-
bility: in art, 51; in definition
of determinism, 150-54, 162-
66, 180, 234, 237; in every-
day life, 54; in history, 53; in
mathematics, 48, 49; in phys-
ics, 24, 50, 77ff, 85ff, 90-91,
163, 172, 184ff; in psy-
chology, 24, 51, 162; of psi
function, 65

Prediction, disadvantages of, 47;
related to choice, 95n, 110,
213; role in science, 81, 180,
204

Pre-established harmony, 84

Presuppositions, of causal state-
ments, 41ff, 217

Price, related to punishment,
194

Priestley, J., 117, 120

Probability, 218; epistemologi-
cal, operational, ontological,

. and theoretical, 203-204; in
physics, 60ff, 90ff, 203fT; sub-
jective and objective, 91

Process: biological, 161; of
creation, 223; of production,
236; physical, 24, 231; psy-
chical, 25

Psi function, 60, 66, 198

Psychoanalysis and psychiatry,
22, 38, 51-52, 121, 126fif, 137,
155, 220

Punishment: deterrent theory
of, 98fif, 190, 193-94; eco-
nomic theory of, 193-94;
Erewhon program of, 116;
retributive theory of, 115-16,
190-91; social utility of,
149ff, 190-91, 194

Purposive action, 161, 181, 229

Quantum theory, see Mechan-
ics, quantum
Quine, W. V., 202, 207

Randomness, 84ff, 183ff, 195,
198, 223, 240-41

Rationalism, in philosophy of
science, 79

Rationalization, 210

Realism, in philosophy of sci-
ence, 65-66, 79

Reality: change in meaning of,
61; ambiguity of, 66

Reflex action, bearing on moral
responsibility, 127

Reid, T., 163, 228

Relations: dyadic and monadic,
161; triadic, 164; in defini-
tion of "cause," 164-65

Relativity, theory of, 78fr, 206ff

Responsibility : criteria of,
1191T, 127ff, 188ff, 219ff;
dilemma of, 158; etiological,
168; legal and criminal, 95ff,
132ff. 168-69, 194ff, 220;
meanings of, 141, 219-20,
225-26; mental elements in^
99ff; moral, 25, 97ff, 117ff,
145ff, 149ff, 157ff, 168-69,-
176ff, 187ff, 193ff, 200, 210-
11, 212ff, 219ff, 224ff, 243;
reflective and unreflective

252 / Determinism and Freedom

conceptions of, \22S, 146;
related to compulsion, 129-
30; related to premeditation,
127-28; subjective and objec-
tive, 219-20

Roberts, K. V., 90

Ross, W. D., 23, 207

Russell, B., 22, 73, 140, 172tt

Russell, E. S., 161n

Ryle, G., 167

Schlick, M., 118ff, 140-41,
172n, 179

Schopenhauer, A., 117, 120

Schrodinger, E., 57, 60, 66fif,
206-8

Schutz, A., 219

Sciama, D. W., 90, 222

Self: as moral agent, 227ff, 242;
problem of, 188; metaphysi-
cal theory of, 158-59, 228

Sellars, W., 172n

Sentimentalism, and determin-
ism, 180ff

Sequence, regular and causal,
163ff, 216

Sidgwick, H., 20

Simultaneity, 208; of cause and
effect, 38

Sin, related to law, 102-4

Skinner, B. F., viii

Spinoza, B., 138

Statistical mechanics, see Me-
chanics, quantum; Probabil-
ity, in physics

Stephen, J. F., 102

Stevenson, C. L., 118

Strawson, P. F., 217

Sumner, W. G., 221

Superego, 134

Tautology, 202, 205, 216
Taylor, R., 224

Thucydides, 54 "*

Time, 204fif

Trotsky, L., 53

Truth: and coimnon sense, 231;
as timeless, 182; logical, 157,
166, 180; of principle of uni-
versal causality, 163ff; of sci-
entific theories, 79-80

Uncertainty, principal of, 60,
79-80, 184ff, 209

Unconscious, role in motiva-
tion of, 52, 126ff

Utilitarianism, as ground of
punishment, 149fif, 189ff; in
legal theory, 105flf

Vagueness, 33flf, 75, 99; of

moral concepts, 188; of

philosophical concepts, 58; of

physical concepts, 23
Value: as secured by law, 115;

of a variable, 48, 185, 239ff;

of excusing conditions, 99ff;

of psi function, 197
Variables, 48, 70, 91, 184-85,

205ff, 239fif
Verification: of counterf actuals,

217; of quantum theory, 91;

of standards, 34; of theories,

79ff
Voltaire, 120
Voluntary action, see Freedom

of will; Responsibility, moral;

Responsibility, legal; Choice

Weber, M., 219
Weiss, P., 231
Whitehead, A. N., 46, 179
Williams, D., 207-8
Williams, G., 102n, 106
Wilson, H. V. R., 237

40*^ 10

m
0

COLLEGE LlBRARyate Due

UNIVERSITY OF FLORIDA

3 1262 04824 7851