Beruflich Dokumente
Kultur Dokumente
COSMOLOOICAL
ARGUMENT
A Reassessment
6
THE
COSMOLOGICAL
ARGUMENT
A Reassessment
By
CHARLES C THOMAS • P U B L I S H E R
Springfield • Illinois • U.S.A.
Published and Distributed Throughout the World by
C H A R L E S G T H O M A S • PUBLISHER
BANNERSTONE HOUSE
301-327 East Lawrence Avenue, Springfield, Illinois, U.S.A.
N A T C H E Z PLANTATION HOUSE
735 North Atlantic Boulevard, Fort Lauderdale, Florida, U.S.A.
/
PREFACE
w ITHIN THE PAST decade or so there has been a gradual awak-
ening of interest in philosophical theology in general and in
the theistic arguments in particular. One of the first indications
of this was the response which the publication of Flew and
Maclntyre's collection, New Essays in Philosophical Theology,
evoked. A major result of this volume was to bring to the at-
tention of the philosophical world questions which had long
been passed off as meaningless. It pointed to the fact that we
can once again critically and cautiously investigate an area of
philosophy which for half a century or more was deemed barren.
It is true that the majority of articles contained therein were
largely negative in the judgment which they passed on a variety
of issues, including the truth and validity of the theistic argu-
ments. And considering the philosophical background out of
which this volume came, this is not surprising. But the lasting
value and importance of this volume are not so much the con-
clusions arrived at, but rather the indication of a new direction
in philosophical theology which it manifests; the endeavor to
do philosophical theology is again made respectable.
In the developments which have succeeded the publication
of this volume, it was the ontological argument which initially
stimulated the greatest interest and debate. The criticisms which
Kant raised had long been accepted by the overwhelming ma-
jority of philosophers as ultimately telling against Anselm's argu-
ment. But Norman Malcolm's reformulation (or discovery of a
second argument in Anselm) ignited a flurry of response, a
reaction which continues to the present. Interest was again
sparked in one of the oldest theses of natural theology. More
recently, interest has also been generated in the cosmological
argument. Has the argument been properly understood? Are
the traditional criticisms really damaging to the argument? Can
the argument be reformulated in order to overcome certain
weaknesses which have been discerned? These are some of the
vii
viii The Cosmological Argument
questions which have been raised in recent books and philosophi-
cal journals. Here again, what is important, to me at least, is not
so much the conclusions which these writers come to (though
they in no wise can be passed over lightly), as their willingness
to take another hard look at the cosmological argument, some-
thing which would have been generally considered unthinkable
not so long ago. Theirs is an unmistakable signal that it is time
to think through the multifarious and difficult issues which sur-
round the cosmological argument.
It is this challenge to analyze the positions which the cos-
mological argument propounds and presupposes which we have
undertaken to meet. The time has arrived for a reassessment
of both the truth and validity of what is to me the most inter-
esting and exciting of the theistic arguments. Such a reassessment,
such an investigation, is precisely the task of this book. It will
be the reevaluation of both the argument itself and the criticisms
which have been raised against it, both historically and in con-
temporary philosophy, which will occupy our attention through-
out.
Our program will be to establish in the beginning, what
initially appears to be a plausible cosmological argument, while
leaving to succeeding chapters the task of defending this against
its critics. In the first chapter we will endeavor to set forth argu-
ments which fully support the various steps to be found in our
cosmological argument. Accordingly, the complete development
of what we believe to be a true and valid cosmological argument
will be given here. Out of this presentation will arise the fact
of the argument's basic reliance on the principles of causation
and sufficient reason. The second, third, and fourth chapters will
be devoted to a detailed defense of the truth and necessity of
both principles. The final chapters will consider the various
criticisms which have been or are now raised against the argu-
ment. We will devote our attention to dealing with each in de-
tail. Consequently, if we can establish a plausible cosmological
argument, and if we can defend this initial argument against
the criticisms which usually have been thought to invalidate it,
then we have given new life to an argument which traditionally
has had an important place in philosophical theology.
Preface IX
•Reprinted from The Monist (Vol. 54, No. 3, 1970), LaSalle, Illinois, with per-
mission of the publisher.
xi
CONTENTS
Page
Preface vii
Acknowledgments xi
Chapter
1. A COSMOLOGICAL ARGUMENT 3
The Cosmological Argument as an Argument Form 4
The Argument 5
Conclusion 19
Notes 20
2. CAUSATION 23
Causation as Constant Conjunction 25
Causation as a Covering Law 33
Causation as Production 39
The Cause as the Necessary and Sufficient Conditions 47
Notes 49
xiii
xiv The Cosmological Argument
6. T H E PROBLEM OF N E C E S S I T Y IN T H E C O N C L U S I O N 107
7. I s T H E COSMOLOGICAL A R G U M E N T D E P E N D E N T U P O N
T H E ONTOLOGICAL A R G U M E N T ? 123
Kant's Analysis of the Argument 124
Conclusion 135
Notes 135
Index 149
THE
COSMOLOGICAL
ARGUMENT
A Reassessment
Chapter 1
A COSMOLOGICAL ARGUMENT
3
4 The Cosmological Argument
the cosmological argument. It is imperative, therefore, that we
assemble this new data into a coherent whole and show its
application to the argument. Secondly, I believe that it is time
that the cosmological argument be analyzed in fresh and con-
temporary terms and that precise arguments be given for the
various steps involved in the argument. Too often the admittedly
skeletal arguments given by St. Thomas Aquinas appear as the
primary representative of the proposed argument. Too seldom
have positive contemporary developments and alterations of
the basic argument appeared in those selections of writings
which purport to give both sides of the issues surrounding
the theistic arguments. What results is a presentation of a
theistic argument which appears inherently weak and horrify-
ingly outmoded, a relic of the thirteenth century.
Thus, it is incumbent upon us to prepare, by an in-depth
study of the cosmological argument, a viable alternative to the
usual fare. This analysis must involve not a mere sketch of some
general pattern or argument form, not a brief recollection and
restatement of some past sketches, but a rigorous development,
as best we can, of a detailed and specific cosmological argu-
ment. To make the strongest possible case we must develop an
argument which, at least initially, appears to be true and valid.
New and old ideas must be brought together; they must be
molded into a coherent, defensible argument. As such we will
attempt to approach the cosmological argument in our own
way. Hopefully, by starting out afresh, though in no way losing
sight of our historical heritage of past cosmological arguments,
we can thoughtfully and positively develop a revised and philo-
sophically important verison of the argument.
THE ARGUMENT
The cosmological argument commences with an appeal to
the world of fact. Its first premise or step is the following ex-
istential assertion:
(Sx) A contingent being exists.
That is, there exists in the real world a being which, if it exists,
could conceivably not have existed.
At first it might seem strange to apply the word "contin-
6 The Cosmological Argument
gent" to beings. We are accustomed to predicating "contingent"
solely of propositions; it strains no credibility to say that a
proposition is contingent, that its opposite is logically possi-
ble. But what can we possibly mean when we say that a
being is contingent? Is it not impossible to apply such modal
terms as "contingent" and "necessary" to beings or things? Do
not these terms become meaningless when so applied?
To answer this objection, we must first determine what the
word "contingent" means. To be contingent is to be such that
it could have been other than it is. Of what, then, can we predi-
cate this adjective? For one, we can predicate it of propositions.
A proposition p is contingent if both p and not-p are possible,
either logically (in that neither p nor not-p contradict the sub-
ject) or conditionally (in that neither p nor not-p contradicts
a certain set of given conditions). For example, "the grass is
short" is a logically contingent propostion, for it is logically
possible for grass to be long as well as short. Again, "roses are
red" is a logically contingent proposition, for it is possible for
roses to be of a color other than red. Thus, "contingent" can be
meaningfully predicated of propositions.
One can also predicate "contingent" of events; an event is
contingent if it either may or may not happen. The event
"raining on Tuesday" is contingent, for on any given Tuesday it
may either rain or not rain. Or again, "meeting Mark in the
lounge" is a contingent event, for at any time I may either
meet or fail to meet Mark there. Thus, the word "contingent"
can be meaningfully predicated of events.
It can likewise be meaningfully predicated of beings. For
a being to be contingent means that it is a being which, at any
time T, either may or may not exist. Though it now exists, its
nonexistence was as possible as its existence; there was no neces-
city, logical or real, that it must now exist. It could have merely
ceased to exist in the preceding moment. Likewise, there is no
necessity that it must exist at the next moment; its nonexistence
at the next moment is as conceivable as is existence, for at any
time it could pass out of existence. Thus, for a being to be a
contingent being means that, if this being exists, it could just as
well not have existed.
A Cosmological Argument 7
Several points should be made clear here. First, we are not
saying that the nonexistence of an object which now exists is
in the future as likely or as probable as its existence. Contin-
gency is not a doctrine about the probability or improbability
of the instantiation of a particular thing. Rather, it is merely the
claim that the existence of an object in the past does not necessi-
tate its existence in the future.
Secondly, this is not to say that that which now is really
might not be, that we cannot be sure it exists. It is not a prin-
ciple of Cartesian doubt. Contingency as applied to existence
is not a statement about some certainty or uncertainty that we
have with respect to knowing whether an object exists; it is not
reporting an epistemological state of affairs. Rather, it is telling
us something about the ontology of the existent or future ex-
istent; it is informing us about the being itself, namely, that
there is no logical or real necessity that it exist now or in the
future; its nonexistence is as conceivable as its existence.
Thus, in light of the above, it seems quite evident that
"contingent," as applied to beings, is quite meaningful. Those
who, in spite of this, still tenaciously argue that this word can
only be applied to propositions and not to beings are being
nothing less than pedantic, especially in light of the fact that
throughout the history of philosophy "contingent" has been con-
sistently applied to beings as well as to propositions.
But granted that the word "contingent" can be meaningfully
applied to beings as well as to propositions, are there such be-
ings in existence? That there are contingent beings seems quite
obvious. My existence, for one, is contingent; my being is such
that at no time does it necessitate its own existence. Even the
fact that I existed in the past provides no guarantee that I must
exist now. I could, at this moment, just as well not have existed
as existed. And the same holds true with respect to the future:
at the next moment I could just as well not exist as exist. I am
not guaranteed any future existence; my instantaneous death at
the next moment is one of the possibilities open to me. Thus,
I at least am contingent. And, since there is such a being, S1; as
a true datum of experience, grounds the argument in the world
of facts.
8 The Cosmological Argument
The obvious result of this factual grounding is to dis-
tinguish this argument from any a priori argument such as, for
example, the ontological argument. According to the latter, the
fact that we are capable of forming in our intellect the
concept of a being than which nothing greater can be conceived
is sufficient to allow, indeed forces us to affirm, the existence in
reality of such a being, a being which possesses certain char-
acteristics, among which are necessary existence and a great-
ness and perfection surpassing that of any empirically known
being—in short, an ens realissimum. Thus, any a priori argument
appeals solely to the conceptual world, and from this basis
argues to the existence of a particular kind of being which in-
habits the real world.
In contrast, the cosmological argument seeks to distinguish
itself from such a priori reasoning by taking as its initial pre-
mise the fact that something exists. It commences not with an
a priori concept, but with an existential statement about the
world. Thus, the argument is not a priori, but a posteriori; it
argues from the nature of the perceived world, from the world
of real existence, to the world of real existence. This initial, fac-
tual, existential premise, then, makes invalid, at least initially,
any criticism that the cosmological argument has argued from
concept to reality.
The cosmological argument continues by arguing that this
being, as a contingent being, must be caused either by itself
or by another. But since to cause itself results in an impossibility,
we have the following:
(S 2 ) This contingent being depends on something else
for its existence.
This proposition involves two contentions. First, the argu-
ment claims that to be a contingent being implies that one is
caused. Every contingent being must have a cause of its ex-
istence, a cause which is either itself or another being. And
secondly, the argument claims that this being cannot be the
cause of its own existence. What reasons do we have for think-
ing that these two propositions are true?
Taking the latter first, since it is perhaps the least con-
troversial, we might argue that if a contingent being is caused,
A Cosmological Argument 9
it must be caused either by itself or by another; there is no
third alternative. But obviously it could not cause itself to exist,
for in order to bring itself into existence, it must already be
such that it can perform this act of creating itself. Consequently,
it would precede itself in existence, if not temporally, at least
logically. But this is impossible. Something cannot precede itself
in existence; something cannot logically be before it is. What
is nothing or nonexistent causes nothing. Thus, since something
cannot cause itself to exist, it must be caused by something
other than itself. The existent which is contingent is therefore
dependent on something outside itself for its existence.
But this only sets in greater relief the first issue, namely,
need a contingent being be caused at all? Granted that if it is
caused, it must be caused by something other than itself, yet
must every contingent being have a cause of its existence? Cannot
a contingent being exist uncaused? Since in the third chapter
we will be dealing in great detail with the argument which can
be used to support the principle that every contingent being
has a cause, it will suffice here to merely anticipate our later
argument. Our argument in that chapter will attempt to show
that for a being to be contingent in and of itself entails that it
must be caused. To be such that one either could or could not
exist means that one's existence is not intrinsically a part of one's
own nature. That is, to be a contingent being implies that one's
act of existence is not essential, but accidental, to one's nature.
If it were essential to its nature, it would be such that, if it
exists, it could not not-exist. Moreover, that which is accidental
is dependent upon something. If it were not dependent but in-
dependent, it would no longer be an accident, but a substance.
Thus, all contingent beings are dependent upon something for
their act of existence, for their instantiation. They are therefore
caused by that on which they depend, for to be dependent
upon something is to be caused by it. Thus for a being to be
contingent implies that it is caused; the fact of its being caused
is derived from the very fact of its contingency. Thus, any con-
tingent being, by the very nature of its being contingent, must
be caused, and indeed caused by another being.
To summarize, the cosmological argument contends that
10 The Cosmological Argument
there exists a contingent being, and that this being, as contin-
gent, not only must be caused, but must be caused by something
other than itself.
Since we have in S2 an existing being which is dependent
on some being outside itself to supply the reason for its ex-
istence, we must now search for that which will enable us to
account for the existence of this being, which will supply the
sufficient reason for this being. It is in this regard that step
three poses the following dichotomy:
(5 3 ) That which causes (provides the sufficient reason
for) the existence of any contingent being must be
either (3) another contingent being, or (4) a non-
contingent (necessary) being.
If what we have said above about contingent beings is correct,
these two possibihties are exhaustive; there is no tertium quid
between the appeal to another contingent being and/or the ap-
peal to a non-contingent being.
However, if we appeal to the former, we really have an
infinite series of contingent beings, for each contingent be-
ing to which we appeal must itself require a sufficient reason
for its existence (from Sx and S2) and so on to infinity. Thus
what results is an alternative between an appeal to an in-
finite series of contingent beings and an appeal to a non-con-
tingent or necessary being, in order to explain the existence
of any contingent being. Step three can therefore be reformu-
lated thus:
(5 4 ) That which causes (provides the sufficient reason
for) the existence of any contingent being must
be either (5) an infinite series of contingent be-
ings, or (4) a non-contingent (necessary) being.
The question which now arises is which of these will pro-
vide the sufficient reason for the existence of any contingent
being. Will an infinite series suffice to account for the being of
S1} or must we appeal to the activity of a necessary being?
The argument of the fifth step contends that a series of
contingent beings or explanations, even if it be infinite, will
never yield a sufficient reason for the existence of any contin-
gent being.2 Thus, we have:
A Cosmological Argument 11
(SB) An infinite series of contingent beings is incapable
of providing a sufficient reason for the existence
of any being.
Since S5 is by no means intuitively obvious, what argu-
ments can be given in support of its claim? How can we show
that S5 is true? Before we can present our argument, we must
first of all clarify exactly what we mean here by an infinite
series. Only after we clearly understand the nature of the
series (5) in S4 can we show it is incapable of yielding a suffi-
cient reason for the being in Si.
What do we mean in S4 and S5 by "an infinite series"? Ac-
cording to Aristotle, the infinite has two kinds of existence. The
infinite either exists actually (in the sense that it now "fully
is"3) or potentially.4 Which kind of infinite are we here con-
cerned with? Is the infinite causal series actual or potential?
Obviously the infinite series of contingent beings in S4 is
not potentially infinite. This can be seen from the fact that in
the series the contingent being which is the cause of another
contingent being is not temporally prior to it. Rather, the cause
and effect are simultaneous. And so it is with every cause and
effect in the series. As such, the entire series under considera-
tion is ordered apart from any consideration of time.
How is this? First, with respect to each individual cause
and effect, as we will argue in detail in the next chapter, causa-
tion does not involve the often deemed essential characteristic
of temporal priority. Indeed, the cause cannot be temporally
prior to the effect at the moment of actual causation, for if there
were a temporal lapse between the causing by the cause and
the effecting of the effect, the cause could never move the effect.
This is, of course, not to say that the cause might not have been
in existence or even in motion prior to the effect; indeed, such
might or might not have been the case. However, it is to say
that the cause, when it was performing the causal activity, could
not have been temporally prior to the effect.
Let us suppose that the cause and effect were to occur at
temporally disparate times. Then at the time at which the
cause would be performing its causal activity, viz. at time Tu
there would be yet no effect for it to cause, for the latter would
12 The Cosmological Argument
have to commence only after the actual act of causation, viz.
at Tt. And when the effect was being effected, viz. at time T2,
there could be no cause there to cause or move it, for the latter
would have to act at Tt. How, then, could the object which we
designate the cause ever move the object we designate the
effect? It obviously could not, for when the one acts, the other
would not exist and hence could not be acted upon; and even
before the one is effected, the other has ceased its causal ac-
tivity. Hence, if temporal priority were a necessary condition
of causation, causation could never occur. But causation does
occur. Therefore the cause and effect must be simultaneous.
The effect must occur at the same time as when the cause is
doing the causing; there can be no temporal lag between the
cause and effect in the span of time when causation is actually
occurring.
Take, for example, the situation in which the wind blows
a leaf, causing it to flutter. The wind can be blowing all day,
but until the leaf moves upon receiving one of its gusts, the
wind is not a cause with respect to that leaf. Or take the ex-
ample of a water skier. The tow boat can be moving away from
the dock, pulling out the line. But until the skier is actually
pulled off the dock and onto the lake the boat is not the cause
with respect to the skier. But when the gust of wind does move
the leaf and the outboard flings the skier into the water, then
causation has occurred. But note that at that precise moment,
the cause and effect are occurring simultaneously; there exists
no temporal gap between them. As long as a time gap separates
the wind and leaf or the boat and skier, no causation occurs
(with respect to the effect in question). These examples clearly
illustrate, I think, the contention that unless there be temporal
simultaneity between cause and effect, there will be no causal
efficacy.
Moreover, not only is this the case with each individual
cause -and effect, but all the causes in the series under con-
sideration in S4 are likewise simultaneous with each other, such
that the entire series may be considered apart from any con-
siderations of time.5 The causes in the series are transitively
related to each other. That is, each cause is the cause of sue-
A Cosmological Argument 13
ceeding effect only insofar as it itself is being caused. Each cause,
as itself contingent, is causally dependent upon another immedi-
ate cause for its causal efficacy. If it were not so dependent, it
in and of itself would be capable of providing the necessary and
sufficient conditions, the sufficient reason, for both itself and its
effect. But as such, it would no longer be a contingent being,
but a necessary being—a being not needing any explanation for
its existence, an uncaused cause. But the series is a series of
contingent beings. Hence, each cause in the series is transitively
related to the other causes, in that it is dependent for its im-
mediate causal efficacy upon the immediate causal activity of
the other members in the series. And since each cause is transi-
tively related to the other causes, since each cause is the "bearer"
of the causal action from the (logically) prior to causes to the
subsequent effect, the entire series must take on the nontemporal
character of the relation between each individual cause and
effect. The entire infinite series, then, is a series in which each
and every cause is simultaneous with each and every effect. The
series must be considered, then, as being ordered apart from
any consideration of time.
Perhaps an example or two might serve to clarify our con-
tention at this point. The arms move the hands, which in turn
move the baseball bat, which in turn strikes the pitched ball.
Here each of the intermediate causes—the hands and the bat—is
a transitive cause; it conveys the causal activity or efficacy from
the arms to the ball. Hence, the bat can strike the ball only
insofar as it is moved by the arms via the hands; the hands and
bat are incapable in and of themselves to hit the ball; they are
moved by and simultaneous with the moving of the arms. Thus,
here we have an example of a finite, transitively related causal
series. Another example can be taken from cooking. The gas
feeds the flame, which heats the pan, which in turn causes the
water to boil. Here again we have a finite, transitively related
series. The flame, heat, and pan are instrumental causes, de-
pendent for their causal efficacy upon the continued flowing
of the gas. They convey the energy of the gas to the water, mak-
ing it boil. Moreover, in this latter case as in the first, simul-
taneity is involved. All elements, working together and at one
14 The Cosmological Argument
time, are necessary to cause the water to boil; if the flame is lit
before or after the gas has flowed, no heat will be generated.
These two examples illustrate the kind of series which is tran-
sitively related.
The objection might be raised, however, that this causal
series is unreal. Since causation does occur in time and does
manifest duration, this series cannot account for duration. Hence,
this type of series, a transitively related series, really has no
bearing on reality. Indeed, there could be no such series as the
argument supposes, for causation does occupy time. The causal
process is durational, but a transitively ordered series cannot
account for duration.
In order to reply to this, we must first distinguish between
two types of causal sequences or series. The one type of series,
about which we have been speaking, involves causes related in
such a fashion that each cause is the cause of the succeeding
effect only insofar as it itself is being caused. Each cause de-
pends upon a prior cause precisely for its own act of causing.
Or in other words, each contingent being is merely a necessary,
but not a sufficient, condition for the effect to occur. As such,
each cause is transitively related to the other causes in the series.
In an analysis of this type of series, the temporal factor is not
an important concern, for each cause is simultaneous with each
and every other cause. However, this is not to say that causal
duration is precluded. Indeed, the actual causal activity might
be temporally extended over a long period of time; the causal
efficacy need not be such that it lasts only an instant. What we
are contending, however, is that no matter what the extent
of its temporal duration, each and every cause in the entire
series is simultaneous with each and every other cause. Thus,
when we say it is ordered apart from any consideration of time,
we do not mean that the series cannot occupy time. We only
mean that temporal duration is not a relevant factor in an
analysis of the series, since each cause in the entire series occurs
(for x length of time) simultaneously.
In the other type of series, the causes are related in such
a way that preceding causes are responsible for their particular
simultaneous effects, but are not directly involved (as necessary
A Cosmological Argument 15
conditions) in producing later effects. That is, they need not be
present in order for the effect which they originally caused to
itself cause something else. For example, my father need not
be present for me to have a son. As such, the causes are related
to each other intransitively; they are not the "bearers" of the
action from one cause to another. In this type of series, the
causal process is durational; it takes place over a span of time.
However, in each particular causal action, the causes involved
act simultaneously, such that within this intransitively ordered
sequence a transitively ordered sequence still exists.
Thus we have two kinds of sequences of causes, one ordered
transitively, the other intransitively.6 Of these two, the first is
clearly primary, for though the intransitive series at the same
time necessarily involves the transitive sequence, the reverse
does not follow. The reason that this is so is due to the nature
of what it is to be a cause. Since the cause is the totality of con-
ditions necessary and sufficient to produce an effect, and since all
these conditions must be present simultaneously or else the effect
will not occur, every causal sequence will necessarily involve a
transitive sequence, for this is the very nature of causation. Thus,
one could conclude that the transitive series is the primary
usage of "causal series," whereas the intransitive series is a de-
rivative usage.
To return to the objection, both causal series have a bearing
on reality. That the transitively ordered series is real stems
from the very nature of causation. In order for any effect to
occur, the totality of conditions necessary and sufficient to pro-
duce that effect must be present. Moreover, unless all those
conditions are present at the same time, the effect will not occur.
Hence, any analysis of what it is to be a cause will presuppose
this type of causal series.
The intransitively ordered series is likewise a real causal
series, as the objection admits. What is important to note is that
this type of series (ordered intransitively) depends ultimately
upon the first type of series (ordered transitively). For unless
sufficient conditions are present to produce each particular ef-
fect in the intransitively ordered series, there will exist no such
series.
16 The Cosmological Argument
Thus, both causal series are real. However, when one is
considering what conditions must be present for an effect to
occur, as is the case in the cosmological argument, only one
type of series is relevant, namely, a transitively ordered series of
necessary conditions. The question in this case is a logical one
concerning required conditions.
Let us now return to the question whether the infinite
series under consideration is potentially or actually infinite. For
a series to be potentially infinite, it must be such that it exists
through a period of time. A potentially infinite series involves a
sequential actualization. But as we have seen, the infinite series
(5) in S4 does not involve sequential actualization. Thus, the
infinite which we are concerned with must be an actual in-
finite, not a potential infinite; the series of contingent beings
must be ordered transitively in order to provide a sufficient
reason for the existence of the contingent being or effect.
With this clarification of the nature of the infinite series (5)
in S4 in mind, we must now inquire whether an actually infinite
series of contingent causes can yield a sufficient reason for the
existence of any contingent being. That it cannot can be seen
from the following. Suppose we have an infinite series, such
that a is caused by b, which in turn and at the same time is
caused by c, which in turn is caused by d, and so on to infinity.
This series is actually infinite, and each cause is in a transitive
causal relation to another cause. Following Brown, and sub-
stituting "cause" for "move" in his account, we now ask,
What [causes] a? Well, it has already been stated that b [causes] c;
so it may be suggested that "b [causes] a" is the desired explanation
of a's motion, the desired value of "x [causes] a." But this would be
an inadequate account of the matter. For b is itself being [caused]
by c, which — owing to the transitivity of "x [causes] y" — thus
yields the implication that a is [caused] by c, with b serving merely
as an instrument or intermediate. But in turn d [causes] c; and so on
indefinitely. Now, so long as this series continues, we have not
found the real [cause] of a; that is to say, we have not found the
explaining value of the function "x [causes] a." The regress is thus
a vicious one, in that the required explanation of a's motion is de-
ferred so long as the series continues.7
And since the series continues indefinitely, the explanation of
A Cosmological Argument 17
CONCLUSION
We may summarize the detailed argument as follows:
(Si) A contingent being exists.
a. This contingent being is caused either (1) by it-
self, or (2) by another.
b. If it were caused by itself, it would have to pre-
cede itself in existence, which is impossible.
(5 2 ) Therefore, this contingent being (2) is caused by an-
other, i.e. depends on something else for its existence.
(5 3 ) That which causes (provides the sufficient reason for)
the existence of any contingent being must be either
(3) another contingent being, or (4) a non-contin-
gent (necessary) being.
c. If 3, then this contingent cause must itself be
caused by another, and so on to infinity.
20 The Cosmological Argument
(5 4 ) Therefore, that which causes (provides the sufficient
reason for) the existence of any contingent being
must be either (5) an infinite series of contingent
beings, or (4) a necessary being.
(5 5 ) An infinite series of contingent beings (5) is incapable
of yielding a sufficient reason for the existence of any
being.
(S e ) Therefore, a necessary being (4) exists.
NOTES
1. Although even a cursory reading of the first two of Thomas's Five
Ways reveals this underlying structure, it might be objected that the third
argument (the argument from contingency) does not manifest such a pat-
tern. However, I believe that a closer analysis will reveal that the pattern
is still there. This argument seems to contain two parts. The first part
attempts to prove that there exist necessary beings. Whereas the existence
of motion and causation seem to Thomas to be obvious facts unnecessary
of proof, the existence of necessary beings does seem to require such proof,
and it is this which he undertakes to develop in the first part of the argu-
ment. The second part of the argument is the cosmological argument
per se. It is here that he attempts to show that if necessary beings do exist
(as he concludes from the first part), then there must be a being which
has its necessity from itself — an uncaused necessary being. The second
part of the argument follows precisely the same pattern as the arguments
enunciated in the first two Ways.
2. In the history of the cosmological argument there seem to have
been two different lines of argumentation, the point of divergence occur-
ring here at S5. On the one hand, St. Thomas Aquinas contends that an
infinite series of moved movers cannot fully or adequately account for the
moving or causing of any one moved object. It is impossible to proceed
to infinity in the series of moved movers, for to do so implies the absurdity
that (1) an infinite number of bodies are moved in a finite time, and (2)
there would be no motion, for where there is no first mover, there is no
A Cosmological Argument 21
subsequent mover (Summa Contra Gentiles, I, Ch. 13, Pars. 11-15; Summa
Theologica, I, Q. 2, Art. 3). Hence, alternative 5 in S4 is rejected because
it cannot provide the cause or sufficient reason for any moved thing.
The other line of argumentation is represented by Duns Scotus.
Scotus argues that an infinite series of causes ordered per se is impossible,
for this whole series of infinite effects would itself be dependent upon
some prior cause. The cause of this series cannot be part of the series, for
if it were it would be the cause of itself. If it exists apart from the series
of all caused beings, it must itself be uncaused, which is precisely what
the arguer is trying to prove (Opus Oxioniese, I, Dist. II, Q. 1). Hence,
alternative 5 in S4 is rejected as an alternative because ultimately even this
infinite series of causes related per se must itself have a cause. The differ-
ence between these two forms of the argument is significant, as we shall
point out in Chapter 5.
3. Richard McKeon (Ed.), Physics, The Basic Works of Aristotle (New
York, Random House, 1941), 206a 14.
4. It is important to note that Aristotle has two different uses of "poten-
tial existence." He writes, "But the phrase 'potential existence' is ambiguous.
When we speak of the potential existence of a statue we mean that there
will be an actual statue. It is not so with the [potential which characterizes
the] infinite. There will not be an actual infinite" (Ibid., 206a 18-20). That
is to say, there is the type of potentiality which implies that the existent
can at some instant of time actually and fully exist. In this sense, the
potential existent will be completely actual at some particular time (all
of it will be actualized at one time). The other type of potentiality will
never be completely actualized at one time; it is potentially actual through
a period of time. It can never become actual all at once, but must be
actualized only through successive stages.
It is in this latter sense that we are asking whether the infinite be
potential. In this sense, the potential infinite will never exist all at once,
but only through sequential actualization.
5. Frederick Copleston, "Commentary on the Five Ways." In John
Hick (Ed.), The Existence of God (New York, Macmfflan, 1964), p. 87f.
It is important to note that this type of causal series has been the series
present in the cosmological argument ever since its formulation by St.
Thomas. Thomas himself distinguished between a causal series which was
ordered per se and one which was ordered per accidens. He writes, "In
efficient causes it is impossible to proceed to infinity per se. Thus, there
cannot be an infinite number of causes that are per se required for a certain
effect; for instance, that a stone be moved by a stick, the stick by the hand,
and so on to infinity. But it is not impossible to proceed to infinity
accidentally as regards efficient causes. . . . [For example,] it is accidental
to this particular man as generator to be generated by another man; for
he generates as a man, and not as the son of another man. . . . Hence it
22 The Cosmological Argument
is not impossible for a man to be generated by man to infinity" (Sumrna
Theologica, I, Q. 46, Art. 2, Reply Obj. 7).
The difference between these two orderings of the causal series is
clearly brought out by Duns Scotus. "Per se or essentially ordered causes
differ from accidentally ordered causes. . . . In essentially ordered causes,
the second depends upon the first precisely in its act of causation. In
accidentally ordered causes this is not the case, although the second may
depend upon the first for its existence, or in some other way. Thus, a son
depends upon his father for existence but is not dependent upon him in
exercising his own causality [that is, in himself begetting a son], since he
can act just as well whether his father be living or dead . . . The [other]
difference is that all per se and essentially ordered causes are simultaneously
required to cause the effect, for otherwise some causality essential to the
effect would be wanting" (Duns Scotus, Opus Oxioniese, I, Dist. II, Q. 1).
Thus traditionally, the type of causal series which is integral to the
cosmological argument was recognized to be one in which each and every
cause is simultaneous with every other cause in the series, and in which
each cause is dependent upon the prior cause for its causal efficacy. A
per se ordering of causes involves no temporal sequence. The other type
of causal series, i.e. one which is ordered per accidens, involves a temporal
series: x causes y, which at a later time causes z. In this type of series,
it has been deemed quite possible to proceed to infinity without contra-
diction. But this series is in no way involved in the cosmological argument;
the argument is not trying to prove the existence of a first cause in time.
6. See Paterson Brown, "Infinite Causal Regression." Philosophical
Review, LXXV, No. 4 (October, 1966), pp. 517-519.
7. Ibid., p. 522.
8. Bertrand Russell, Our Knowledge of the External World (London,
Routledge and Kegan Paul, 1963), p. 197.
9. Brown, op. cit., p. 523; substituting "cause" for "move."
10. Russell, op. cit., p. 170.
Chapter 2
CAUSATION
w HAT CONSTITUTES an adequate explanation for the existence
of contingent beings? To what can we appeal? How far, in
terms of the order of causes, need one progress? The cosmo-
logical argument contends that an adequate explanation must
include the sum total of all the conditions individually neces-
sary and jointly sufficient for that event to occur. But is this an
appropriate analysis of the requirements for adequate explana-
tion? Is this the kind of explanation that is relevant to an analysis
of the events in our world?
William Alston, in dealing with the cosmological argument,
poses the issue this way:
Thus the crucial issue on which the cosmological argument hinges
is an issue concerning the requirements for an adequate explanation.
He who resists the argument will say something like this: So long
as we explain a mother's overprotecting her daughter by pointing
out her repressed hostility to the daughter and explain that in terms
of her unconscious sense that her daughter is a threat to her, and so
on, it is true that at each stage what we bring in to do the explaining
stands in need of explanation of exactly the same sort, and that unless
we can get back to something that is logically necessary we Can never
exhaust the task. But that is just one of the things we have to learn
to face in life, and it is unreasonable to demand that things be tidier
than this. From this standpoint Aquinas is making an unreasonable
metaphysical demand on the universe by insisting that it must be
such that a final completion of this process is possible in principle.
A Thomist will reply that this insistence is just part of the general
assumption of the rationality of the universe which everyone makes
whether or not he realizes it. But to this it will be replied that there
are various grades and modes of rationality and that we can hold the
universe to be rational in some ways, for example, ordered causally,
without making such extreme demands as this. It would seem that
at this point we have come up against one of those rock-bottom
philosophical oppositions about which nothing further can be said. 1
23
24 Tne Cosmological Argument
It seems that the core of the disagreement between pro-
ponents and critics of the cosmological argument at this junc-
ture clearly lies in the nature of explanation. When one ex-
plains the occurrence of an event, what sort of explanation is
adequate or sufficient? Is it quite sufficient to present the event
which preceded the event to be explained? Will a listing of rele-
vant conditions be sufficient to account for the event in ques-
tion? Or must explanation be in terms of the totality of conditions
necessary and sufficient to explain the effect? What is the nature
of explantion?
On the surface, it would seem that Alston is correct in
his claim that we have before us a rock-bottom issue about
which little more can be said. We have reached a methodological
impasse, with the result that the cosmological argument will
be valid only to those who choose to adopt a particular meth-
odology, invalid to those who adopt another method, and
nothing much more can be done than to compare methods.
But is the issue of the nature of explanation involved here
so rock-bottom that nothing further can be relevant? One must
be careful, I think, about putting the issue as Alston did, for it
is not simply a question of what features we must give in order
to explain the event. It is not a question about how complete
our own enumeration of necessary conditions for any particular
event must be. It is not simply an epistemological issue, as
Alston seems to claim. The problem is grounded in an ontological
issue about the nature of explanation, or more accurately, about
causation. In explaining the existence of any contingent being,
one is asking for the cause of it. What brought it into existence?
What caused it to be? Hence, the question really is, What is
it to be a cause? And this question certainly is open to philo-
sophical analysis and scrutiny. In other words, since the issue
is not merely methodological, but about the nature of what it
is to be a cause, we are not at a philosophical impasse; the door
is open to further investigation.
This same point can be seen in another way. In answer
to the question posed by Alston, "What will explain a mother's
overprotecting her daughter," both the critic and the propo-
nent of the cosmological argument will give the same reply,
Causation 25
"The cause." But this answer does not terminate dialogue and
debate, with each banking on a view of explanation about which
nothing further can be said. Rather, it opens up the possibility
for further analysis, for we can now inquire what each means
by the term "cause," and evaluate the adequacy of their analyses.
Thus, the real question is, What is it to be a cause? It is
the answer to this which will determine what does or does not
constitute an adequate explanation. To proposed answers to this
question let us now turn our attention.
Thus, the confirmation of any causal law does not simply de-
pend upon the presentation of evidence which shows that two
events are constantly conjoined, for in the very nature of this
case, such evidence can be no more than partial. However, if the
law can be deduced from other laws supported by evidence, so
that together they find their place in a total hypothetical-deduc-
tive system, the law thus based has behind it the total weight
of both the evidence and the deductive system. That is to say,
the scientific laws which explain causation are not simply
generalizations based on constant conjunction, but are also de-
ductions from higher-level scientific hypotheses. It is the fact
that the law finds its place within the total system which ac-
counts for our expressions of causation in apodictic terms.18
In this view, therefore,
. . . when a person asks for a cause of a particular event, what he is
requesting is the specification of a preceding or simultaneous event
which, in conjunction with certain unspecified cause factors of the
nature of permanent conditions, is nomically 19 sufficient to determine
the occurrence of the event to be explained in accordance with a
causal law. 20
That is to say, causation can be analyzed first into those events
which precede or are simultaneous with the explicandum-event.
36 The Cosmological Argument
These events constitute the sufficient condition for the ex-
plicandum-event. Secondly, there are those conditions in the
nature of the explicandum-event which make possible such an
event (for example, the glass in the window has a certain
fragility). Finally, the constant conjunction of these events must
be according to, or deducible from, a particular causal law.
It is this aspect which gives our causal statements their apodictic
certainty. Thus, when events fulfilling the first two conditions
are constantly conjoined so that we can form a generalization
concerning their relationship, and when this general statement
can be placed within a deductive system, then we can ascribe
causation to these events.
Causation is therefore not a productive relation between
things, but designates the set of rules whereby, when events are
constantly conjoined and this generalization can be subsumed
under a causal law,21 we can apply the term "causation." And
for Braithwaite, it is the fact that this generalization can be
placed into a scientific system, in which it can be deduced from
higher laws, that avoids the difficulties of basing causation mere-
ly on constant conjunction.
Critique
Can one properly conclude that this contemporary attempt
to rescue a constant conjunction theory of causation by intro-
ducing the notion of a covering law is successful? To answer
this, let us first evaluate the particular analysis given by Braith-
waite; then let us turn to a more general evaluation of the notion
of a covering law.
Braithwaite appeals to higher order causal laws in order
to solve two difficulties which exist in an analysis of causation
simply in terms of constant conjunction. First, as we have
argued above, constant conjunction is in and of itself neither
a necessary nor a sufficient condition for causation. One can
have instances of causation where this condition is absent; one
can have events where this condition is present but causation
fails to occur. Accordingly, constant conjunction does not con-
stitute a proper explicans of causation. Braithwaite's response
to this criticism is that constant conjunction must be supple-
Causation 37
mented by an appeal to higher order causal laws. It is never
constant conjunction alone, but constant conjunction in accord-
ance with a causal law, which provides us with a proper analysis
of the causal event.
Secondly, why is it the case that a particular event occurs
only after another particular event? Why, for example, does
the throwing of a ball through a window always result in the
glass breaking, and never in the glass singing? Since any two
events can in theory be constantly conjoined, why it is that
only certain events are followed by certain other events? Since,
as Hume argued, constant conjunction means events conjoined,
not connected, there is nothing within the relation of constant
conjunction which can account for why one event must be con-
joined with another particular event, rather than with an en-
tirely different event. Constant conjunction can account for the
prediction as to what event will probably occur after another
(i.e. that one event is conjoined to another particular event), and
this because it has always been thus in the past. But it cannot
say why this particular event must always be associated with this
particular cause.
Again, to solve this Braithwaite appeals to his causal laws.
For him, we can know why one particular event must follow
another because of the nature of causal laws. A law states a
necessary relationship between events. Hence, in any particular
case we can know that x must follow upon y because there is a
causal generalization which asserts that events of class X must
follow those of class Y. And because these causal generalizations
can be deduced from higher causal laws, which in turn express
the way the world must be, these generalizations express why
particular events are causally related in the world.
In both cases, then, he answers the objection raised against
a constant conjunction theory by appealing to the fact that this
causal generalization finds its place in a hypothetical-deductive
system, in that it can be deduced from higher (causal) laws.
But surely this answer begs the question. We cannot hope to
explain causation by introducing the notion of causality in the
explicans. To say that to ask for the cause is to ask for the
generalization which must be according to and deducible from
38 The Cosmological Argument
higher causal or nomic laws or hypotheses, is certainly circular,
for here we are explaining causal generalizations in terms of
higher causal generalizations. It is this circularity which makes
his position suspect.
Moreover, how do we account for these higher causal laws?
For Braithwaite, these higher laws are themselves inductive
generalizations deduced from still higher laws or generalizations.
And these in turn can be deduced from still higher generaliza-
tions, and so on until we reach the highest level of generaliza-
tions. But these highest ones are themselves inductive general-
izations based on constant conjunction; they cannot be deduced
from anything higher because they are the highest. Thus, the
problems raised against the lowest level of generalizations again
arise in relation to the highest causal generalizations which are
formed on the basis of constant conjunction. All that Braithwaite
has succeeded in doing is to push the problems back to an-
other level. But at the highest level the same problems arise,
and here there are no higher inductive hypotheses from which
to deduce these generalizations. Thus, not only has he begged
the question by introducing at the higher levels the notion of
causation, a notion which he was attempting to account for
on the lower levels, but he has not succeeded in solving the
original problems raised against the lowest levels of inductive
generalizations, for he has merely pushed these problems up
to the highest level, where they can be reintroduced against
the inductive generalizations established there. Thus, this analy-
sis still leaves unsolved some of the basic problems encountered
by a theory based upon constant conjunction.
Putting aside for a moment the specific difficulties which
Braithwaite encounters in his analysis, what can be said in
general about analyzing causation in terms of a covering law?
A covering law states that whenever events of such and such a
kind occur, events of such and such a kind will follow. A law
thus delineates specific conditions which must be met, and
informs us what specific consequences follow upon these con-
ditions. As such, it distinguishes relevant conditions from ir-
relevant ones.
But is not this to say that (1) certain conditions are
Causation 39
necessary for a certain event to occur, (2) that the totality of
these conditions, as the sufficient condition, is the cause of the
effect, and (3) that there must be some objective necessary
connection between the cause and the effect? The first certainly
must be the case, for it is the purpose of a law to state those,
and only those, conditions under which a specific event (as the
effect) will occur. The second likewise seems to be the case,
for in laying out the conditions sufficient for the event, the
law is designating the cause of that effect. And finally, the third
also seems to follow. A general statement cannot be considered
a law if it is merely true in a contingent sense, i.e. if it merely
states a de facto connection between certain properties as
causal conditions and a certain effect. A law states a necessary,
not a contingent relationship, between objects; it must state
not only what has happened, but what must happen under
those specific conditions. But to state what must happen is to
make that statement represent a necessary connection between
the cause and the effect. It is to say that there is something in
the cause which is not present in other conditions which are
deemed extraneous, something which makes the effect occur.
Instead of dispensing with necessary connection by invoking
covering laws, this analysis requires it. Instead of explaining
it in terms of deducibility within a deductive system, it pre-
supposes it.
Thus, covering laws do not function as an explanation of
causation. The necessity which is inherent in the law derives
from the necessary connection between the cause and the effect.
Indeed, the possibility of formulating the causal law depends
on the existence of such a necessary connection. Hence, a
covering law does not explain causation; it formalizes the causal
relation but is not an analysis of causality. The question of the
nature of causation and of the necessary connection between
cause and effect still remains.
CAUSATION AS PRODUCTION
There remains one analysis of causality upon which we
have not touched. Philosophers prior to Hume viewed causation
as an act of production; the cause was that which produced
40 The Cosmological Argument
changes in another object. For example, Aristotle writes, "[Cause
can be interpreted as] the primary source of the change or
coming to rest, . . . as generally what makes of what is made
and what causes change of what is changed."22 Or later John
Locke writes, "That which produces any simple or complex
idea we denote by the general name, cause, and that which is
produced, effect."23 The (efficient) cause, then, is that object
which by means of its activity produces changes in another
object, designated the effect.
Implied in this analysis is the contention that there is
something in the cause which makes it capable of such activity.
That by which the cause produces these changes in the state,
location, or character of the effect is its causal power or causal
efficacy; the cause has within it the power to bring about those
changes observed in the effect. It is this efficacy which gives
causation its dynamic character. Thus, causation involves a
transfer of energy or power from one object to another.
These notions might appear at first sight "unscientific," but
such need not be the case. This notion of causal efficacy or
power can, I think, be adequately translated into contemporary
scientific terms. For example, in some cases the power can be
seen as the kinetic energy of the object (cause) in motion,
while the transference of this corresponds with the mechanical
work done, to the expenditure of this kinetic energy in the
production of the effect.24 In order for there to be causation,
this energy must be expended in a certain way. If the object j
(cause) is merely expending kinetic energy, but producing no i
changes in another object, action can be observed, but certainly
not causation. Causation therefore involves the expenditure of
energy in the production of change in another object.
A
/V
/ \
I \
•> <
I ' ^ \ .
. . * "A '
Figure A
NOTES
1. William P. Alston, Religions Belief and Philosophical Thought
(New York, Harcourt, Brace and World, 1963), p. 20.
2. David Hume, Treatise on Human Nature (Oxford, The Clarendon
Press, 1888), p. 75.
3. Ibid., p. 77.
4. Ibid., p. 157.
5. Ibid., p. 166.
6. Ibid., p. 165.
7. Ibid., p. 157.
8. Ibid., p. 161.
9. Richard Taylor, Action and Purpose (Englewood Cliffs, N.J.,
Prentice-Hall, 1966), p. 36.
10. Ibid., p. 38.
11. Richard Bevan Braithwaite, Scientific Explanation (New York,
Harper and Row, 1953), p. 10.
12. Bertrand Russell, Our Knowledge of the External World, (New
York, W. W. Norton and Co., 1929), p. 239.
13. Stephen Toulmin, The Philosophy of Science (New York, Harper
and Brothers, 1960), p. 123.
14. Braithwaite, op. cit., p. 2.
50 The Cosmalogical Argument
15. Ibid., p. 293.
16. Ibid., p. 9.
17. Ibid., p. 11.
18. Ibid., p. 10.
19. "Nomic" for him means "a necessary connection because it is a
law." A nomic generalization, which can be deduced from higher laws,
is to be distinguished from mere generalization by simple enumeration
{Ibid., p. 293).
20. Ibid., p. 320.
21. For Braithwaite, causal laws are a subclass of the laws of nature;
they are natural laws which involve temporal conditions (Ibid., pp. 308-
309). Laws of nature are to be distinguished from mere generalizations by
simple enumeration in that the hypothesis expressing them "occurs in an
established scientific deductive system as a higher-level hypothesis con-
taining theoretical concepts or that it occurs in an established scientific
deductive system as a deduction from higher-level hypotheses which are
supported by empirical evidence for the hypothesis itself" (Ibid., pp. 301-
302). Therefore causal laws are inductive generalizations from instances
of constant conjunction, which find their place in the hypothetical-deductive
system, in that (1) they can be deduced from higher laws; (2) they are
indirectly supported by evidence for other laws; and (3) they involve a
temporal relation, in terms of priority, between events.
22. Richard McKeon (Ed.), Physics, The Basic Works of Aristotle
(New York, Random House, 1941), Bk. 2, Ch. 3, 194b*..
23. John Locke, An Essay Concerning Human Understanding, Bk. II,
Ch. 26, sec. 1.
24. Albert Michotte, The Perception of Causality, trans, by T. R. and
Elaine Miles (London, Methuen and Co., 1963), p. 227.
25. Hume, op. cit., p. 77.
v
26. Taylor, op. cit., p. 39.
27. Hume, op. cit., p. 161.
28. Michotte, op. cit., p. 225.
29. I am using "object" here in a phenomenal sense. Since Michotte,
in his experiments, used illusions produced by mechanical arrangements, it
was not the case that there were really any objects as such involved.
However, the experiments were set up so as to give the appearance of one
object acting on another.
30. Michotte, op. cit., p. 87.
31. Ibid., p . 15, note 20.
32. Ibid., p. 15.
33. Ibid., p. 223.
34. Ibid., pp. 221-222.
Chapter 3
51
52 The Cosmological Argument
But why are not the color and form distinguishable, whereas
the cause and effect are? Hume's reason, I submit, is that we
cannot have an impression of the color alone—it must also in-
clude the form—whereas we can have one impression of the
cause and another of the effect. Thus, for Hume, the criterion
for deciding whether two things are distinguishable or not is
whether we can have two separate impressions of them. If we
can have separate impressions, then they are distinguishable.
Distinguishability, therefore, is determined in terms of the epis-
temological conditions required for knowing.
On the other hand, what does Hume mean by "separable"?
He obviously does not use it in the sense of being conceptually
abstracted, as for example when we abstract (separate) in our
mind a person's Roman nose from his face, or the flame from
the roaring fire. That is, the separation he has in mind is not
designative of an intellectual operation or state which has no
58 The Cosmological Argument
real bearing on the nature of reality or on the dependence or
independence of things. If he meant this, his argument would
merely conclude that we could possibly abstract the cause from
the cause-effect nexus, but that this abstraction would make no
comment upon, would have no bearing on, the essential or real
relationship of cause to effect. And quite obviously this kind
of separation would say nothing about the causal principle,
which concerns itself with this very relation.
Rather, he means "separable" in a real or ontological sense.
By saying that one thing is separable from another, he means
to imply that one thing is such that it is or can be entirely in-
dependent of the other things. Separability is meant to comment
upon the essential relation which one thing bears or might bear
to another. Thus, when he denies that figure can be separable
from a body, he means that they are essentially connected such
that you cannot have one without the other.7 Accordingly, re-
turning to Hume's argument and to premise 1, whenever we
have two separate impressions, the objects (and their ideas) can
exist independently of each other—they can be separate
CONCLUSION
What then can be said to Camus' objection, quoted at the
beginning of this chapter, that though thought may require the
principle of sufficient reason, the universe does not respond in
any intelligible fashion, but rather absurdly? This profound at-
tempt to separate thought and reality must be viewed with
great suspicion. On one hand, its result is to take the intellect
out of nature. The human reason becomes an isolated element,
antithetical to, indeed at enmity with, nature. For the existential-
ist, nature is true being, mind is non-being. Absurdity, frus-
tration, angst, all are a result of the impetuous desire of the
mind to become what it can never be, i.e. being-itself. But
can the human being be so abstracted from, so cut off from,
nature? Whether one accepts the creationist or evolutionary
metaphysic, both make the same claim: man is intrinsically a
part of nature. Thus, to denaturalize the intellect is to de-
naturalize man, for part of the essence of man, his differentia,
is his ability to conceptualize. And to remove man from his
world is to preclude the possibility of his finding satisfaction.
It is to force the existentialist's absurdity on to man. Hence, far
from describing the human situation, by separating the intellect
from nature they have thrust a strong intellectual predicament
on man. Only by putting man back into his proper place in
nature can this imposed predicament be resolved.
On the other hand, Camus' position proffers two sets of
categories. There are those categories according to which the
mind operates, and there are others according to which nature
functions. These two sets are disparate. For example, Camus
The Principles Of Causation And Sufficient Reason 71
contends that the mind operates according to the principle
of sufficient reason and that nature does not. But with this
radical disjunction of principles, how can he claim to know
what the principles of nature are, especially that they are differ-
ent from those of human reason? On what basis can he make
the claim that they are not the same as the principles of the
intellect? Such a pretension is impossible, for the implication
of his position must be that we cannot know those principles
of nature. Otherwise, reasoning man would have already adopt-
ed these principles as the principles of his own reasoning. If
he knows the principles of reality, man will surely attempt to
encompass them in his own thought. Camus' position, at best,
leads to subjectivism; what we know are the principles of the
intellect through which the world is filtered. Or as Kant puts
it, the phenomena we know; the noumena we postulate about.
But is this an adequate basis on which to do philosophy, science,
or anything else? Must we not operate on the assumption that
the mind can know reality? And must we not likewise assume
that the principles of reality can become the principles of the
human intellect, that the principles evolved throughout the his-
tory of the human mind correspond with reality? Such an as-
sumption, I believe, is at the heart of every serious endeavor to
come to grips with our world.
Finally, the above arguments for the necessity of the prin-
ciples of causation and sufficient reason must be taken seriously.
If one can present a cogent defense of these principles, as we
have attempted to do, the burden of disproof rests on those
who refuse to accept them. One can, of course, even in the face
of demonstrations, embrace absurdity. But then one cannot
rationally defend such a position; one cannot simultaneously
deny and accept the dictates of reason. Absurdity must then be
accepted as an item of intuition or faith. But the cosmological
argument appeals to man as a rational being, noting that it is
on this level that philosophical man operates. On this level one
must face these demonstrations as genuine attempts to rationally
defend the principles of order and intelligibility in the universe.
Our conclusion in this chapter must be that these principles are
necessarily true.
w
73
74 The Cosmological Argument
The issue here is not that we have not shown what we
claimed to show, namely, that the causal principle is both in-
formative and necessary, but rather that such conditions, when
found in a proposition, appear to be contradictory. And since
there appears to be a contradiction in our results, there must
be an error in our analysis. Thus, the issue here concerns
whether and how a proposition can be informative and necessary
at the same time. The attempt to discover an answer to this
problem will be the subject of concern in this chapter.
LOGICAL NECESSITY
What do contemporary philosophers mean when they state
that a proposition is necessary? What are the necessary and
sufficient conditions for a proposition to be necessary? On one
hand, the necessary condition is that the opposite of the propo-
sition must imply a contradiction. If the negation of the proposi-
tion does not imply a contradiction, that is to say, if the opposite
is perfectly possible or conceivable, then the proposition is not
necessary but contingent. On the other hand, the sufficient con-
dition for a proposition to be necessary is that it be necessary
solely by virtue of the meanings of the terms which compose it
and not by virtue of any state of affairs in the world. The con-
tradiction which characterizes its opposite must be produced
solely from a consideration of the logical principles and the
definitions of the terms involved. Thus, for many philosophers
a necessary proposition is a logically necessary proposition, one
whose truth and necessity is determined solely on the basis of
an analysis of the meanings of constituent terms and the stipu-
lated logical rules for transformation of linguistic expressions.
Accordingly, we shall designate this particular analysis of neces-
sity as logical necessity. '"-y.;'.-'-
To illustrate this view, we might take A. J. Ayer's analysis
of necessity as generally typical of this position. Ayer follows
in the Kantian tradition by making necessity an essential char-
acteristic of a priori propositions. If a proposition is (logically)
necessary, it is a priori. What, then, is an a priori proposition?
For Ayer, an a priori proposition is one which concerns only
the "relations of ideas,"3 as opposed to matters of fact. That is,
Can Propositions Be Informative And Necessary Simultaneously? 75
the truth of an a priori proposition is determined simply by an
analysis of the meanings of the terms involved. It is true ex vi
terminorum. Thus, a logically necessary proposition, such as an
a priori proposition, is a proposition the contradiction of whose
opposite is derived from the analysis of the meanings of the
terms involved.
From this it becomes immediately evident that logical
necessity involves analyticity. For Ayer, an analytic proposition
is one whose "validity depends solely on the definitions of the
symbols it contains."4 The truth of an analytic proposition is
decided on the basis of the meanings of the constituent terms;
it is this which sets it apart from a synthetic proposition, whose
"validity is determined by the facts of experience." Consequent-
ly, the sufficient condition (that they are true ex vi terminorum)
for a proposition to be analytic is the same as the sufficient
condition for a proposition to be logically necessary. Thus, we
can say that a logically necessary proposition must be expressed
in the form of an analytic proposition.
Indeed, for Ayer a priori (necessary) propositions are
"necessary and certain only because they are analytic."5 To be
necessary is to be such that one's opposite implies a contradic-
tion. Accordingly the opposite of a necessary proposition must
be of the form A is non-A. If its opposite is of this form, then a
necessary proposition must be of the form A is A. That is, a
necessary proposition must be a tautology.6 But for Ayer, propo-
sitions which are tautologies are analytic. He writes, "I use the
word 'tautology' in such a way that a proposition can be said
to be a tautology if it is analytic."7 Therefore, all a priori (neces-
sary ) propositions must be analytic.
Moreover, for Ayer, neither analytic nor a priori proposi-
tions, since they are tautologies, are informative8 about matters
of fact. They neither refer to matters of fact nor can be "con-
firmed or refuted by any facts of experience." They are devoid
of any factual content and consequently say nothing whatsoever
about the world. Thus, on the view of logical necessity, being
necessary and being informative about the world are two incom-
patible properties of propositions.
But though they are uninformative, though they simply
76 The Cosmological Argument
express tautologies, "we are not suggesting that they are sense-
less. . . . For although they give us no information about any
empirical situation, they do enlighten us by illustrating the way
in which we use certain symbols. . . . They call attention to
linguistic usages, of which we might otherwise not be con-
scious."9 Though they add nothing to our knowledge of the
world, they do reveal how our conventionally established lan-
guage operates.
Thus, finally, logical necessity involves a conventionalist
view of language. It is a characteristic of logical necessity that
it "merely reflects our use of words, the arbitrary conventions
of language."10 Whether necessary propositions are true or
false does not depend on reality, but rather on the conventions
of our language, on our conventionally established symbolism.
Analytic propositions "simply record our determination to use
words in a certain way. We cannot deny them without infringing
the conventions which are presupposed by our very denial."11
The truth of an analytic proposition is determined simply through
an analysis of the meanings of the terms involved. But these
terms are defined, not with reference to the nature of the refer-
ent, but simply on the basis of the conventions of linguistic usage.
They are nominal definitions which express our decision as to
how we should use a word. C. I. Lewis, stating a position which
he later refutes, puts the issue with regard to analytic truth in
this way:
It may thus suggest itself that the ultimate ground of all analytic
truth is to be found in definitive statements, together with rules of
the transformation and derivation of linguistic expressions, such rules
being themselves resultant from equivalences of definitions and con-
ventions of synthetic usage. And to this it may be added that defini-
tions also are merely conventions of the use of language; determined
by decisions made at will concerning the equivalence of expressions.
Thus it may appear that analytic truth in general expresses nothing
beyond what is determined or determinable by conventions of linguis-
tic usage But if any conception of this general sort be enter-
tained, then analytic truth becomes viewed, commensurately, as rela-
tive to the content and structure of a system of language, erected ac-
cording to rules of usage and of manipulation having, in the last
analysis, no further basis than decisions made as we choose according
to our purposes. 12
Can Propositions Be Informative And Necessary Simultaneously? 77
Necessary truths are therefore determined to be true on the
basis of Knguistic habit; they are in no way explicative of any
necessity in the world.
If they were explicative, then the doctrine of logical neces-
sity would not stand, for we would have propositions which on
one hand were informative, and on the other were necessary,
for the proposition would not be true simply on the basis of the
meanings of the terms involved, but since these terms were
explicative of reality they would be true because of the way
^ reality was structured. But on the view of logical necessity, a
proposition cannot be both necessary and informative at the
same time; only those statements which are true in virtue of
their concepts alone can be necessary. Consequently, logical
necessity involves a conventionalist view of language; it is pre-
dicated of statements, the meanings of whose terms are con-
ventionally established without reference to the nature of things.
In conclusion, then, logical necessity is expressed by analytic
a priori propositions, the meanings of whose constituent terms
are established on the basis of a conventionalist view of lan-
guage; necessity is a function of linguistic conventions about the
meanings of words and the stipulated syntactical rules for their
significant combination in sentences. Such propositions are neces-
sary but uninformative, i.e. devoid of factual content about the
world.
A Reduction of Categories
Ayer's theory of necessity is not without its difficulties.
First of all, Ayer's way of solving the problem whether a propo-
sition can be necessary and informative is much too easy. What
he seems to have done is to define "analytic" and "a priori" in
the same way, and from this to conclude that it is impossible
(by definition) that there be any necessary proposition which
is not analytic, i.e. which would be synthetic and a priori. To
be a priori is to be necessary, i.e. to be such that one's opposite
entails a self-contradiction determinable from an analysis of
the meanings of the terms involved. To be analytic is to be true
solely in virtue of the meanings of the symbols involved—hence
the equivalence, and the corresponding impossibility of a non-
78 The Cosmological Argument
analytic necessary truth. But solving the issue of the synthetic
a priori in this fashion makes it true but quite trivial to say that
all necessary propositions must be analytic; it is at best an empty
victory.
Moreover, by defining both terms in this way, he has re-
duced the category of analytic to the category of a priori, such
that there is no difference between them. Both must then func-
ion as categories in the same way. But why then maintain two
category names? Why not just eliminate one or the other? To
utilize both is redundant. But certainly Kant, when he intro-
duced this categorization, and succeeding philosophers, when
they adopted such, have meant something different by these
two categories. Ayer's confusion of the two can be cleared up
if we remember that the one refers to the structure of proposi-
tions, while the other refers to the mode of verification of
propositions. The analytic/synthetic distinction categorizes the
form or structure of propositions. A proposition is analytic if
and only if its negation leads to the form of an unobvious
tautology, i.e. A is not-A, while the negation of a synthetic propo-
sition does not lead to the form of A is not-A. The a priori/a pos-
teriori distinction, on the other hand, characterizes the mode of
verification of propositions. "To characterize a proposition as
'a priori' is to say nothing whatever about its formal structure,
or the structure of its negation, or consequences derivable there-
from. It is, rather, to remark the mode whereby the truth of the
proposition is discovered. A proposition is a priori if its truth is
established without recourse to any possible experience."13 As
such, one could discuss what kind of form or structure the
proposition had (e.g. if it was of the form of an obvious or
unobvious tautology) without having to decide how we would
discover the truth of such.
The consequence of this is that a proposition can be in-
formative (synthetic) and necessary (a priori) at the same
time. One can learn how a proposition is structured without
directly entailing anything about its mode of justification. To
establish that a proposition is synthetic is simply to establish
that it is of the form or structure such that its opposite is not
self-contradictory, is not of the form A is not-A. But this is quite
Can Propositions Be Informative And Necessary Simultaneously? 79
another thing from establishing its mode of justification, i.e. dis-
covering on what basis it is true. To be justified a priori is
possible. In other words, to be necessary (a priori) a proposition
need not be analytic. To be informative and necessary simultane-
ously is not a contradiction. Ayer's analysis, which endeavors to
show this, is guilty of reducing to one category two very different
categories.
This, of course, is not to say that there actually are any
propositions which are synthetic and a priori. Indeed, phi-
losophers like Norwood Russell Hanson, who argue as we have
just done, note that as far as they are concerned no such prop-
ositions have been brought forth, and it is unlikely that any
genuine candidates will be forthcoming in the future. In other
words, they agree with Ayer that all necessary propositions are
analytic, though they disagree with his contention that they
must be analytic. This is to say, however, that a synthetic (in-
formative) a priori (necessary) proposition is a possibility, and
not a contradiction of terms.
Some contemporary philosophers, then, have held that prop-
ositions can be both informative and necessary, in that synthetic
a priori propositions are possible. Whether this is a viable option
for those who maintain the causal principle we shall have to
evaluate in a later section.
REAL NECESSITY
What do we mean when we say that a proposition is really
(as opposed to logically) necessary? On one hand, like logical
necessity, the necessary condition for a proposition to be really
or naturally necessary is that its opposite imply a contradiction.
Unless the opposite implies a contradiction, the proposition is
contingent, not necessary. But the difference between the two
theories lies in their sufficient condition. The sufficient condition
for a proposition to be logically necessary, as noted above, is
that it be necessary solely in virtue of the principles of logic
and the meanings of its terms, without any reference to the
real world. But the sufficient condition for a proposition to be
really or naturally necessary is that it be necessary in virtue of
the very nature or structure of the world. Statements which ex-
press the way things are in the world are necessary because
of the nature of the world, and not in virtue of a priori concepts.
Their necessity is therefore derived, or arises, from the essen-
tial structure of the world; they are necessary propositions
88 The Cosmological Argument
because of the way the world essentially is. Or put another way,
we might say that a necessary truth reflects the necessary, essen-
tial structure of the world.
Now whereas logical necessity involves a conventionalist
view of language, real necessity involves what we might term
an essentialist view of language. That is to say, the meanings of
our words are not established simply on the basis of conven-
tional usage, but rather on the basis of what we believe the
nature of that object to be. We say that % means y because we
have discovered that y seems to be essentially what x is in the
' real world. For example, we define "man" as a "rational animal,"
not because we happen to use the word "man" to mean the
same as "rational animal," but rather because from our experi-
ence with men we have discovered that rationality and ani-
mality appear to be essential characteristics of them. Quite apart
from our conceiving it, prior to our thinking about it, man is
just the sort of being which is a rational animal. All the men
with whom we are acquainted seem to exhibit these character-
istics, not accidently, but as an essential part of their makeup. 27
Thus our concept of what man is, is derived from and applicable
to the real world, for our definition of "man" has been formed
on the basis of our analysis of what he essentially is in the
world. The meaning is therefore the conceptual form of the
—essence. Or as Quine puts it when he critically refers to essen-
tialism, "Meaning is what essence becomes when it is divorced
from the object of reference and wedded to the word."28
> Moreover, the statement "man is a rational animal" is
necessary because of the way things are in the world, because
man is essentially and necessarily a rational animal. To deny
that man is such would be to contradict what man in fact truly
is. Necessary truths therefore reflect a real necessity, a necessity
which is operative prior to and independent of linguistic formu-
lation.
Statements, then, which are really necessary are both in-
formative and necessary. They are informative in that they tell
us about the world, and about the essential structure of beings
in the world and of the world itself. And they are necessary in
that a denial of them would lead to contradiction; it would be
Can Propositions Be Informative And Necessary Simultaneously? 89
saying that a thing was not what it was or that the structure of
the world was not what it was.
Someone, however, might raise the following objection. It
is an obvious fact that we can be mistaken about states of affairs
in the world. Hence, informative propositions are such that
what they assert could conceivably be different; the distant
tower might be round, not square, though we have stated that
it was square. Now if that which is informative can possibly be
other than it is, how then can it be necessary, for to be necessary
means that it cannot be other than it is? It seems that we are
affirming two contradictory claims about the proposition at the
same time, namely, that it both can and cannot assert that the
state of affairs might be other than what they were claimed to be.
The solution to this difficulty can best be brought out by
an illustration from the history of philosophy. Let us take, as
an example, the definition of an atom as put forth by the early
Atomists. "Atom" was defined as "an indivisible unit of matter."
The proposition "an atom is an indivisible unit of matter" was
accordingly a necessary proposition (in that to deny it led to
a contradiction) and an informative one (in that it informed
one what an atom really and essentially was). For the Atomists,
it would have been inconceivable to deny that an atom was an
indivisible unit of matter; such would have been self-contra-
dictory. Yet, on the other hand, they were certainly talking
about real atoms and not simply about the concept of atom.
When Lucretius claimed that all things were composed of
atoms, he was not saying that entities were structured out of
concepts; such would be absurd. Concepts of atoms do not go
together to make up entities—real atoms do. Similarly, when he
attempted to define what an atom essentially was, he was not
speaking about the concept of such, but about real atoms. Thus,
here we have a case of an informative and necessary proposition.
Today, however, modern science has discovered that what
was once considered to be indivisible was not indivisible; the
atom is composed of still smaller particles. Thus today scien-
tists define the atom, not as indivisible, but as "the smallest com-
ponent of an element having all the properties of the element."
And here again, we have a necessary and informative proposi-
90 The Cosmological Argument
tion, but one which differs significantly from the ancient defini-
tion of an atom.
What is the point here? It is simply that what were once
conceived to be necessary propositions or necessary truths about
the world might conceivably not be true at all. That is to say,
such propositions are subject to correction. We may be mistaken
as to what we conceive the essence of any given thing to be.
NOTES
1. A. J. Aver, Language, Truth and Logic (New York, Dover Pub-
lications, 1946), p. 22.
2. J. J. C. Smart, "The Existence of God." In Antony Flew and
Alasdair Maclntyre (Eds.), New Essays in Philosophical Theology (New
York, Macmillan, 1955), p. 38.
92 The Cosmological Argument
3. Ayer, op. cit., p. 31.
4. Ibid., p. 78.
5. J M . , p. 31.
6. Ibid., pp. 16, 114.
7. Zfotd., p . 16.
8. Ibid., p. 79. By "informative" we mean "devoid of factual con-
tent about the world." A proposition can be uninformative in this sense,
while still being informative in the sense that it informs us about our use
of words.
9. Ayer, op. cit., pp. 78, 80.
10. J. N. Findlay, "Can God's Existence Be Disproved?" In New
Essays in Philosophical Theology (see note 2 ) , p. 54.
11. Ayer, op. cit., p. 84. See also pp. 31, 79.
12. Clarence Irving Lewis, An Analysis of Knowledge and Valuation
(LaSalle, 111., Open Court, 1946), p. 96.
13. Norwood Russell Hanson, "The Very Idea of a Synthetic-Apriori,"
Mind, LXXI, no. 284 (October, 1962), p. 521.
14. Lewis, op. cit., pp. 72, 105.
15. Ayer, op. cit., p. 31.
16. Ibid., p. 79.
17. John Wisdom, "Metaphysics and Verification." Mind, XLVII,
no. 188 (October, 1938), p. 463n.
18. Henry B. Veatch, "St. Thomas and the Question, 'How Are Syn-
thetic Judgments A Priori Possible?'" The Modern Schoolman, XLII, no. 3
(March, 1965), pp. 256-257.
19. Arthur Pap, Semantics and Necessary Truth (New Haven, Conn.,
Yale University Press, 1958), pp. 164-168.
20. Ayer, "Introduction to the Second Edition." In Language, Truth
and Logic, p. 16.
Norman Malcolm, "Are Necessary Propositions Really Verbal?" Mind,
XLIX, no. 194 (April, 1940), pp. 199-202.
Ayer ("Truth by Convention." Analysis, IV, nos. 2 & 3, pp. 19-20)
claims that since necessary truths are prescriptive, they are not propositions
at all. But Malcolm argues that in order to be a proposition, a statement
does not have to be descriptive. That is, it is perfectly possible to have
non-descriptive propositions (Malcolm, op. cit., pp. 202-203).
21. Malcolm, op. cit., p. 199.
22. Ibid., pp. 199, 202.
23. Immanuel Kant, Critique of Pure Reason, trans, by Norman Kemp
Smith (New York, St. Martin's Press, 1965), A94.
24. Ibid., A246, with the substitution of "realen" for "transcendentalem"
made by Kant in the Nachtrage.
25. Ibid., A580.
26. Ibid., A616.
Can Propositions Be Informative And Necessary Simultaneously? 93
27. The epistemological question how we discover that this is the
essential feature of man is not the issue here. I am merely attempting to
state the basis on which real necessity is grounded, so that we may under-
stand what real necessity is.
28. Willard VanOrman Quine, From a Logical Point of View (New
York, Harper and Row, 1953), pp. 22, 155.
29. Henry B. Veatch, "On Trying to Say and to Know What's What."
Philosophy and Phenomenological Research, XXIV, No. 1 (September,
1963), p. 91.
Chapter 5
94
Causation And The Totality Of Contingent Beings 95
istence of the universe. And if we cannot do this, then we can
never arrive at the existence of a necessary being which is sup-
posed to explain the existence of such. We can proceed no farther
than to ask for the causes of particular effects; to ask for more
is to lapse into meaningless queries. The totality, therefore, is
totally without explanation; to ask for an explanation in terms
of finding a cause for it is meaningless.
Thus for Russell, the question, Is there a sufficient reason
for the totality of beings? is crucial to the cosmological argu-
ment. But such a query is meaningless. Thus the entire edifice
of the argument supposedly collapses around our feet.
CONCLUSION
It appears that none of Russell's arguments will support his
106 The Cosmological Argument
claim that we cannot proceed beyond the totality in order to
explain the existence of a contingent being. We have argued,
first, that to make the quest for the cause of the totality a
central part of the cosmological argument is not applicable to
our argument; no such contention can be found in the argu-
ment. Secondly, the appeal to an explanatory totality to which
causation does not apply can be countered by an argument
from the character of the parts of this totality to the character
of the whole. This argument shows that the totality must itself
be contingent and therefore cannot fill the role Russell assigns
to it, namely, of being a cause to which itself the notion of
causal explanation is inapplicable. Finally, we have contended,
contrary to Russell, that there is a meaningful frame of refer-
ence in which we can ask for the cause of the totality, namely,
a framework in which the notion of totality is extended beyond
its restricted usage with reference only to contingent beings.
Hopefully, then, the cosmological argument has met the chal-
lenge of Bertrand Russell.
NOTES
1. Frederick Copleston and Bertrand Russell, "A Debate on the Exis-
tence of God." In John Hick (Ed.), The Existence of God (New York,
Macmillan, 1964), p. 175.
2. Note 2, Ch. 1.
3. Copleston and Russell, op. tit., pp. 174-175.
4. Ibid., p. 175.
5. The argument form involved here would be modus ponens.
Premise 1: (x) (y) [(xRy • Px) D Py]
Premise 2: (3x) (3y) [(xRy • Px)]
Premise 3: .". (3y) (Py)
6. William L. Rowe, "The Fallacy of Composition." Mind, LXXI,
No. 281 (January, 1962), p. 89.
7. The causation involved is obviously not efficient causation. Prob-
ably the closest we could come to categorizing the cause would be to call
it a material cause.
8. Why this is so with respect to certain properties and not with others
is difficult to say. But the recognition that the part/whole inference works
with certain properties like colors and extension, but not for others like
small or round, is not contingent upon our presentation of a principle which
will enable us to decide automatically in every instance whether premise 1
is necessarily true.
Chapter 6
107
108 The Cosmological Argument
implies the existence of a necessary being . . . To say that there
is a necessary being is to say that it would be self-contradictory to
deny its existence.3
Moreover, upon concluding that it is logical necessity which
is involved in the conclusion of the argument, it is then
fashionable to quickly dispense with the question of the validity
of the cosmological argument on the grounds that it leads to a
logically impossible conclusion. For example, from the phras-
ing of the conclusion of the cosmological argument in terms of
logical necessity, Smart contends that the cosmological argu-
ment must therefore be invalid because this proposition is self-
contradictory. No existential proposition can be logically neces-
sary. " 'A logically necessary being' is a self-contradictory ex-
pression like 'round square.' "4 J. N. Findlay goes even farther,
arguing not simply that the cosmological argument is invalid
because the conclusion is self-contradictory, but presenting what
amounts to a disproof of God's existence on the same assump-
tion that since "necessary," when applied to the divine being,
is logical necessity, the existence of a necessary being is self-
contradictory and hence impossible.5
Thus, the claim is made that whether the cosmological
argument concludes that a necessary being exists or that a divine
being necessarily exists, in either case the necessity involved
is of one particular kind, namely, logical necessity. Consequent-
ly, no matter which way one formulates the conclusion of the
cosmological argument, it may be properly translated into "The
proposition 'God exists' is logically necessary," and this latter
proposition is evidently self-contradictory.
But are Smart and others correct when they claim, first
all, that the proposition "a necessary being exists" is equivalent
to the proposition "a being necessarily exists"?6 Is the word
"necessary" being used in precisely the same sense in both in-
stances? Secondly, are they correct when they claim that both
propositions are translatable into "the proposition 'God exists'
is logically necessary"? That is, can they justify their claim that
there is indeed only one kind of necessity, namely, logical
necessity? And finally, when the cosmological argument argues
to the existence of a necessary being, is it claiming that its
The Problem Of Necessity In The Conclusion 109
conclusion is a logically necessary proposition, as Smart and
others seem to think?
In this chapter we want to attempt to unravel the con-
fusion which surrounds the use of the word "necessary" as it
is used with reference to divine existence, and in particular, as
it manifests itself in the conclusion of the cosmological argu-
ment. Hopefully we can show that the majority of contem-
porary criticisms of the cosmological argument (i.e. those criti-
cisms which, like Smart's, attack the conclusion of the argument
as being self-contradictory) are not really so telling against the
argument as their proponents would have us think. Rather
these criticisms arise from the contemporary muddle which sur-
rounds the use of "necessary" in the cosmological argument. The
muddle, we will argue, has arisen because of the modern ten-
dency to group all uses of "necessary" under one category, viz.
logical necessity. Our contention will be that such a move is a
mistake and that a proper analysis of the different uses of "neces-
sary" involved in the conclusion of our argument will go a long
way toward making the conclusion of the argument, and the
theistic claims in general, both understandable and meaningful.
CONDITIONAL NECESSITY
Must a proposition, to be necessary, be logically necessary?
Must the opposite of every proposition which claims to be
necessary be seZ/-contradictory? Must all necessary propositions
be necessary simply in virtue of the meanings of the terms
contained within the proposition? I think not. There are proposi-
tions which derive their necessity, not from an analysis of the
meanings of the terms contained within the proposition, but
rather from being the conclusion of a valid argument. The
proposition follows necessarily from its premises.
That such concluding statements are not logically necessary
can be seen from the fact that they lack the two characteristics
which are essential for a proposition to be logically necessary.
On one hand, they are not self-contradictory. The denial of the
conclusion of an argument does not lead to self-contradiction.
Indeed, the opposite of such is perfectly conceivable; the con-
clusion, taken simply by itself, is nothing more than a contingent
statement. What it does contradict is a certain set of conditions
or premises. For example, in the standard argument, "All men
are mortal, Socrates is a man, therefore Socrates is mortal," the
denial of the concluding statement is not self-contradictory. If
we denied that Socrates is mortal, we would not be contradicting
The Problem Of Necessity In The Conclusion 115
ourselves, but rather contradicting a set of given premises or
conditions. Thus, on this count a concluding statement of an
argument is not logically necessary.
On the other hand, the conclusion of an argument makes
no claim to be necessary simply on the basis of an analysis of
the meanings of the terms involved in the concluding proposi-
tion. If such were the case, then no argument would be neces-
sary to support such a conclusion; the conclusion would be
self-evident. But such an argument is essential to establish the
type of proposition which we are considering. For example, it is
not self-evident that Socrates is mortal; we must appeal to
other relevant facts to support this. Thus, on neither count can
the conclusion of an argument be considered to be logically
necessary.
However, this is not to say that the conclusion is not
necessary. Indeed, it is necessary; its opposite does imply a
contradiction. But the contradiction which is implied is not self-
contradiction, but rather a contradiction of a certain set of
conditions or premises. The conclusion is necessary, but it is
necessary because it is the conclusion of a valid argument: the
conclusion follows necessarily from its premises. Thus, though
it is not logically necessary, the conclusion is still necessary;
it is necessitated by a certain set of conditions.
Consequently, we have an answer to our first question. A
proposition does not have to be logically necessary in order to
be necessary. It can be necessary in that to deny it involves
a contradiction with respect to a certain set of conditions or
premises. Thus, we have at least two kinds of necessity which
can characterize propositions: a proposition can be logically
necessary, or it can be conditionally necessary, i.e. necessary with
respect to certain conditions or premises.
CONCLUSION
We have, then, the answer to the questions which we posed
at the outset of this chapter. First of all, the proposition "a neces-
sary being exists" is not equivalent to "a being necessarily
The Problem Of Necessity In The Conclusion 121
exists." The necessity predicated of beings is different from
that predicated of propositions; these two different uses
must not be confused. Secondly, we have contended that there
is more than one kind of necessity. We have isolated at least
three kinds of necessity: that predicated of beings, logical ne-
cessity predicated of propositions, and conditional necessity pre-
dicated of propositions. Finally, it is not the case that the
necessity which is involved in the cosmological argument is
logical necessity. In particular, the cosmological argument does
not contend for a being whose existence is logically necessary.
Rather, it contends for the logically contingent, conditionally
necessary existence of a necessary being, a being which, if it
exists, cannot not exist. It concludes to a conditionally neces-
sary proposition about a necessary being, a being which is in-
dependent of all other beings, which can neither be caused to
exist nor cease to exist. And as such, the argument avoids the
contemporary criticism that its conclusion is self-contradictory.
Such a criticism is based on the mistaken assumption that the
necessity involved in the conclusion of the argument is of only
one kind, and that is logical necessity.
NOTES
1. J. J. C. Smart, "The Existence of God." In Antony Flew and
Alasdair Maeintyre (Eds.), New Essays in Philosophical Theology (New
York, Macmillan, 1955), p. 38.
2. C. B. Martin, Religious Belief (Ithaca, N. Y., Cornell University
Press, 1959), p. 151.
3. Paul Edwards, A Modern Introduction to Philosophy (New York,
Macmillan, 1965), p. 381.
4. Smart, op. cit., p. 39. See also Edwards, op. cit., "To talk about
anything 'existing necessarily' is . . . about as sensible as to talk about
round squares."
5. J. N. Findlay, "Can God's Existence Be Disproved?" In New
Essays in Philosophical Theology, (see note 1), p. 52.
6. They are obviously in good company. See Immanuel Kant, Cri-
tique of Pure Reason, A585, 587, 588, 606.
Also Frederick Copleston and Bertrand Russell, "A Debate on the
Existence of God." In John Hick (Ed.), The Existence of God (New York,
Macmillan, 1964), p. 169.
7. Smart, op. cit., p. 38.
122 The Cosmological Argument
8. Terence Penelhum, "Divine Necessity" Mind, LXIX, No. 274
(April, 1960), pp. 180-181.
9. Findlay, op. cit., p. 52.
10. Smart, op cit., p. 38. See Penelhum, op. cit., p. 185, "A proposi-
tion is necessary if its truth can be known without reference to anything
other than a clear understanding of what is said or implied in it."
11. See Findlay, op cit., p. 54.
Also A. J. Ayer, Language, Truth and Logic (New York, Dover Pub-
lications, 1946), pp. 22, 31, 73f.
12. Smart, op cit., p. 39.
13. Ibid., p. 38; Copleston and Russell, op. cit., p . 169.
14. Penelhum, op. cit., p. 185.
15. John Hick, Philosophy of Religion (Englewood Cliffs, N. J., Pren-
tice-Hall, 1963), p. 23. It should be made clear that I am not claiming that
this has been the historical position with respect to the meaning of "neces-
sary" as applied to beings. As Patterson Brown has shown in "St. Thomas's
Doctrine of Necessary Being," The Philosophical Review, LXXIII (January,
1964), this is certainly not Thomas's meaning. What I am claiming is that
this is a consistent and adequate analysis of "necessary" as applicable
to beings.
16. Alvin Plantinga, God and Other Minds (Ithaca, N. Y., Cornell
University Press, 1967), p. 22.
17. Ayer, op cit., p. 16.
18. Ibid., p. 79.
19. Ibid., pp. 79-80.
20. Plantinga, op. cit., pp. 22-23.
21. One should be careful not to confuse the real necessity predicated
of beings with that predicated of propositions (in Ch. 4 ) . A being is a
necessary being if it is such that if it exists, it cannot not exist. A
proposition is really necessary if its opposite is contradictory because it
denies or contradicts the very nature or essential structure of the world.
Chapter 7
123
124 The Cosmological Argument
being proven to exist, in order to "make the negative concept
of a non-contingent or necessary being into a positive one," the
concept of a most real being (ens realissimum) is introduced.
"Unless we can show that we possess such a concept—whether
it be the concept of a perfect being, a supreme being, or a most
real being—the whole cosmological argument must fail to prove
the existence of an absolutely necessary being in any positive
sense."2 Thus, since the cosmological argument must ultimately
appeal to the ontological argument, it is ultimately dependent
upon the ontological argument for its validity.
But this yields a curious conclusion. Paton writes, "If the
ontological argument is valid, the cosmological argument is
superfluous. If the ontological argument is invalid, the cosmo-
logical argument must be invalid too."3 Thus, seemingly, advo-
cates of the cosmological argument are impaled on the horns
of a dilemma: either the cosmological argument is superfluous
in that all that is necessary to prove the existence of the neces-
sary being is the ontological argument, or it is invalid in that
the ontological argument on which it is based is invalid. In
light of our argument in the first chapter, this is surely an un-
happy dilemma.
But is the cosmological argument dependent on the onto-
logical argument? Granted that it has been thought to be de-
pendent, why has it been thought to be such? What reason can
be given for such a claim? Paton's analysis is merely a summa-
tion of the fuller argument to which contemporary critics ap-
peal, i.e. Kant's argument as presented in the Critique of Pure
Reason. Let us turn to this traditional formulation as the locus
chssicus of the purported proof of dependence. What is said
with respect to Kant will, I think, apply equally to such con-
temporary critics as Paton.
Thus, the first step proceeds from the fact that a contingent
being exists to the necessity that there exist an absolutely neces-
sary something.
But this something which necessarily exists is indeterminate;
the first part of the argument only leads us to "the concept of
absolute necessity, but is unable to demonstrate this necessity
as belonging to any determinate thing."7 That is to say, the first
part of the argument only yields the fact that something neces-
sarily exists; it concludes with the existence of some necessary
being, but the argument does not inform us as to exactly what
or who this being is. "What properties this being may have,"8
how many such beings there are, who the being is—all this the
argument fails to tell us. Indeed, Kant contends that the argu-
ment leaves open the possibility that this being might even be
a limited being. He writes as follows:
[Even granting the argument] it by no means follows that the con-
cept of a limited being which does not have the highest reality is
for that reason incompatible with absolute necessity. For although
I do not find in its concept that unconditioned which is involved in
the concept of the totality of conditions, we are not justified in con-
cluding that its existence must for this reason be conditioned. . . .
On the contrary, we are entirely free to hold that any limited beings
whatsoever, notwithstanding their being limited, may also be un-
conditionally necessary. . . . Thus the argument has failed to give
us the least concept of the properties of a necessary being, and
indeed, is utterly ineffective.9
126 The Cosmological Argument
All that the first part concludes is that an indeterminate
something necessarily exists. Thus, a second part is required to
complete the argument, a part which provides us with a con-
cept by which we can know or recognize this necessary being.
In this stage, pure reason chooses from its warehouse of con-
cepts that one which is most suitable to and most compatible
with the concept of necessary existence. And the concept which it
finds to be most compatible is the concept of an ens realissimum.
"Now there is only one possible concept which determines a
thing completely a priori, namely, the concept of the ens real-
issimum. The concept of the ens realissimum is therefore the
only concept through which a necessary being can be thought."10
Thus reason, by its own workings, through its own con-
cepts, supplies us with the concept by which we can know that
which necessarily exists. It is reason alone, in the second part,
that determines which concept is most compatible with it. It
can appeal to nothing else to determine the possessor of ab-
solutely necessary existence.
Kant then raises his objection to the introduction of the
use of pure reason, for, to him, this means that we have sur-
reptitiously introduced the ontological argument in the second
part. Kant writes as follows:
But it is evident that we are here presupposing that the concept of
the highest reality is completely adequate to the concept of absolute
necessity of existence, that is, that the latter can be inferred from
the former. Now this is the proposition maintained by the ontologi-
cal proof; it is here being assumed in the cosmological proof; and
indeed, is made the basis of the proof. . . - 11
Kant's objection is that the cosmological argument is based on
the ontological argument, such that the validity of the former
is entirely dependent on that of the latter. The cosmological
proof intrinsically relies on the ontological argument for its
cogency.
However, Kant does not stop at this point, but even goes
further to say that the cosmological argument reduces to, is
nothing more than, the ontological proof.
Appeal is made to the agreement of two witnesses. . . . In reality
the only witness is that which speaks in the name of pure reason;
in the endeavor to pass as a second witness it merely changes its
Is Cosmological Argument Dependent Upon Ontological Agrument? 127
dress and voice. . . . Thus, the so-called cosmological proof really
owes any cogency which it may have to the ontological proof from
mere concepts. The appeal to experience is quite superfluous.12
This really boils down to saying that the cosmological argu-
ment is nothing more than the ontological argument in a new
dress. Though it appeals to experience, when we analyze it, it
is really arguing from a priori concepts all along. All that is
necessary to the cosmological argument is presented in the
second part, which argues from pure reason alone.
Moreover, since, according to Kant, the ontological proof
is invalid, so also is the cosmological proof which is based on
it. Thus if, as Kant claims, the cosmological argument is based
on or reducible to the ontological, and if the latter is invalid,
then the cosmological argument must be rejected as invalid, for
it yields no proof for the existence of any necessary being.
In dealing with this criticism there are two possible de-
fenses. On one hand we might attempt to defend the validity
of the ontological argument, and in this way rescue the cosmo-
logical. Or else, we might try to show that the cosmological
proof is in no way dependent on the ontological. It is the latter
which will be attempted here.
But how can we show that the cosmological proof is inde-
pendent of the ontological? To do such, it seems that we must
answer two questions. First, in what way does the cosmological
argument supposedly depend on the ontological argument? I
hope to show that Kant's claim of dependence cannot be sus-
tained, for the conclusion of the cosmological argument cannot
be converted into the principle of the ontological argument, as
he thought it could. The second question might be put, Why is
the cosmological proof thought to depend on the ontological?
Our answer to this will hopefully reveal both the basic mis-
conception on the part of Kant as to the nature of the cosmo-
logical argument and also his misconception as to the nature
of the necessity involved in it. Let us now consider these ques-
tions in this order.
CONCLUSION
We have attempted to show that the traditional reasons
given by Kant to show that the cosmological proof is dependent
upon the ontological proof cannot be sustained. Those who con-
tend that the cosmological proof needs the ontological in
order to identify the being to which we have argued not only
fail to see that the characterization of this being lies outside
the realm of the proof per se, but due to a confusion of the
necessity which is characteristic of beings with that which
is applicable to statements and propositions, they also fail to
see that the cosmological argument does indeed, in and of
itself, provide a concept by which we can identify this being
which exists. Furthermore, the demonstration given by Kant
to show that the conclusion of the cosmological argument is
the principle of the ontological argument does not yield
the above result. It is not arguing from concept to necessary
existence in reality, but concludes to the realm of existence from
a prior premise which had already established such existence.
But the latter is not the ontological argument, for whereas this
argument claims to be able to prove the existence of a necessary
being without prior proof, the conclusion in 8' depends upon
such prior proof of existence. Hence, those who contend in this
fashion that the cosmological argument rests upon the ontological
argument seem badly mistaken.
NOTES
1. H. J. Paton, The Modern Predicament (New York, Collier Books,
1955), pp. 202-203.
2. Ibid., p. 203.
3. Ibid., p. 203.
4. Immanuel Kant, Critique of Pure Reason, trans, by Norman Kemp
Smith (New York, St. Martin's Press, 1965), A584, 585, 606. It is ob-
viously the first part — the appeal to experience — which sets the cosmo-
logical argument apart from the ontological argument. Whereas the latter
appeals solely to pure reason, the former "really begins with experience,
and is not wholly a priori or ontological" (A605).
5. Ibid., A604.
6. Ibid., A584, 585, See Paton, op. cit., pp. 196-199.
7. Ibid., A607. See Paton, op. cit., p. 202.
8. Ibid., A606.
9. Ibid., A588.
136 The Cosmological Argument
10. Ibid., A605.
11. Ibid., A607. See Paton, op. cit., p. 203.
12. Ibid., A606-607.
13. Ibid., A608.
14. Ibid., A608.
15. Ibid., A607.
16. Ibid., A593, 594.
17. Ibid., A606. This likewise seems to be the objection raised by
Paton (see note 1), p. 202.
18. Aquinas himself recognized this distinction. The cosmological ar-
guments presented in Question Two of the Summa Theologica intend to
establish that, for example, a first cause exists, but it is the task of the further
reasoning of Question Three and following, which are outside the domain
of the cosmological argument, to establish what properties this being has,
what characterizes it.
19. Ibid., A606. See also A585, 587, 588, 607. Also see Paton, op.
cit., p. 199.
Chapter 8
137
138 The Cosmological Argument
The same point was made several centuries ago by David Hume:
"Why may not the material universe be the necessarily existing
being, according to this pretended explication of necessity?"5
Those who argue in this fashion are making two points.
First, they are claiming that as the cosmological argument
stands it cannot validly conclude that this necessary being is
God. In order to make such a move legitimately, the argument
must be further supplemented. Matson, for example, goes on
to suggest that the cosmological argument should be supple-
mented by the argument from design. With this general con-
tention I agree; in light of the premises of the argument, to
The Necessary Being And God 139
move from the necessary being to a personal deity requires
additional argumentation. We shall say more on this later.
Secondly, and more importantly, they are contending that
one need not conclude that the necessary being in S6 is divine
at all. Indeed, it is quite conceivable (or even evident) that
this necessary being can be identified with the world or material
universe. We need not conclude to a personal divinity, intro-
duced from a religious sphere external to the argument; the
world itself could be the necessary being, as that which "non-
temporally actualized itself."
But is the world or material universe a proper candidate
for the position of necessary being? I think not. The totaHty or
world or universe is nothing over and above the sum total of
its constituents. But all its constituents are contingent beings;
what exists in the universe could conceivably not exist. Now if
the components, as contingent, could conceivably not exist, then
the totality or world which they compose could likewise con-
ceivably not exist, for if all the constituents ceased to exist at
the same time (which is possible), the totaHty of which they
constitute the parts would likewise cease to exist. And if the
whole which now exists can conceivably not be, it is contingent.
Thus, contrary to what Matson claims, the world cannot be the
necessary being for which we argue, for if it were, it would
possess contradictory properties: it would be necessary because
it is the necessary being to which the cosmological proof argues,
and it would be contingent because it is the totality of con-
tingent beings.
Matson responds to this by noting that it is meaningless
to apply the notion of "cause" to the totality of contingent be-
ings. That which is contingent is caused; but causation can
only meaningfully take place within the context of a totaHty.
"The universe is the framework within which causal explana-
tions operate. . . . It is quite beyond their scope to link the uni-
verse to anything [beyond itself]. To ask for the cause of the
universe is to ask a question similar to 'When is time?' or 'Where
is space?'" 6 Consequently, to inquire about the cause of this
totaHty is to move outside the framework within which the
question of causation can be meaningfully asked. As such, the
140 The Cosmological Argument
contention that the totality is contingent and caused is meaning-
less.
Since we have attempted to answer this objection in detail
in Chapter 5, a summation of our contention there will suffice
at this point. Though it be granted that causation can take
place only within a totaHty, there is no reason to restrict the
notion of "totality" simply to the totality of contingent beings.
There are many kinds of totalities, some of which are relevant
to the problem before us. There can be, for example, a totality
of all existent beings or a totality of all things which are causes.
That is, the notion of "totaHty" can be satisfied in the inclusion
of all contingent and necessary beings in the whole, and it
is within this resulting totality of existents that we can mean-
ingfully inquire concerning the contingency or causation of
the totaHty of contingent beings. Thus, far from being meanr
ingless, the question of the contingency or causation of the
totaHty of contingent beings can be asked within a meaningful
framework. Just as we can ask for the cause of the totaHty of
tulips within the totaHty of the world, we can ask for the cause
of the totaHty of contingent beings within the totality of ex-
istents or totaHty of causes.
Thus, it seems that we cannot identify the world or ma-
terial universe with the necessary being, for the world itself is
contingent and requires an explanation or cause for its existence.
INDEX
A D
Accident, 9, 61-63 Distinguishable, 56-60
Accidently related, 21-22, 88 Duns Scotus, 21-22, 95-97
Alston, William, 23-24, 47, 137
Analytic proposition, 73-83, 86 E
A priori Edwards, Paul, 107
analysis, 8, 126-128, 132 Ens realissimum, 8, 124, 126, 128-
concepts, 85, 87 130, 133-134
propositions, 74-79 Epistemological conditions, 59-60
Aquinas, S t Thomas, 3-5, 20-21, 95- Essentialism, 88
97, 136-137, 140-141 Eternal, 110, 118, 134, 142-143
Aristotle, 3, 11, 21, 40, 66 Existentialism, 70
Ayer, A. J., 73-81, 92
F
B Faith (and reason), 144-146
Fallacy of composition, 99-103
Braithwaite, Richard B., 34-38
Findlay, J. N., 108, 113, 117
Brown, Patterson, 16
Flew, Antony, vii
c G
Camus, Albert, 52, 70-71 Gilkey, Langdon, 52
Categories (Kantian), 85-86 God, 52, 87, 107-108, 116-117, 123,
Causal 137-146
efficacy, 26-27, 40-46
impression, 46-47 H
laws, see Covering law Hanson, Norwood Russell, 79
principle, 8-9, 51-71, 73-74, 79, Hume, David, 3, 25-46, 5 1 , 53, 56-60,
85-87, 91 138
Cause, 5, 8-67, 73, 94-106, 110, 120,
139-140, 142-143 I
Constant conjunction, 25, 27-28, 32-
38, 4 1 , 47 Impression, 25, 27, 42, 45-47, 57-60
Contingent, 5-7, 20, 63, 99, 102-107, Inductive generalization, 34-35, 38, 50
117 Infinite series, 10-11, 16-18, 20-21,
beings, 5-13, 16-20, 23-24, 51-56, 95-98
61-63, 67-68, 73, 94-99, 102-
110, 116-117, 139-140 K
Conventionalism, 76-77 Kant, Immanuel, 3, 71, 78, 85-87,
Copleston, Frederick, 65, 94-97, 104, 124-135
137 Kierkegaard, S0ren, 144-146
Covering law, 33-39, 47-48 Kinetic energy, 40
149
150 The Cosmological Argument
S
N
Separable, 56-60
Necessary Simultaneity, 11-15, 28-32, 35
being, 18-20, 51, 95, 98, 105, 107- Smart, J. J. C , 73, 107-109, 112,
112, 116-135, 137-144 114, 117
conditions, 15, 17, 23, 25, 28-32, Sufficient
39, 4 1 , 47-49, 51, 53-55, conditions, 15, 17, 23-25, 32-33,
68, 74, 87 35-36, 39, 4 1 , 47-48, 51-55,
connection, 26-27 68, 74-75, 87
propositions, 73-91, 107-122, 129- reason, 10-11, 13, 16-20, 52-55,
130, 133, 135 68, 94-98, 105, 134, 142
truth, 56, 65, 100-103 principle of, 52-56, 67-71
Necessity Synthetic
conditional, 114-117, 121, 130 apriori, 77-79, 85-86
logical, 6 1 , 73-74, 87-88, 9 1 , 107- propositions, 73, 75, 77-79
109, 112-121, 129-130
real, 6 1 , 81, 87-91, 120-122 T
Nicene Creed, 142
Tautology, 75-76, 79-80, 113, 119
Taylor, Richard, 30-31, 41, 69
o Temporal priority, 11-12, 25, 27-32,
Ontological 35, 41
argument, vii, 8, 112, 119, 123-135 Toulmin, Stephen, 34
conditions, 59-60 Transitive series, 13-16, 31-32
Owens, Joseph, 62
u
P - '- ' Universe, 138-140
Pap, Arthur, 83
Paton, H. J., 123-124, 137 V
Penelhum, Terrence, 112, 118
Veatch, Henry, 82
Perception, 42-46, 94, 98
Plantinga, Alvin, 119-120
Principle of
w
causation, see Causal principle Wisdom, John, 81