Causal Necessity and the Ontological Argument

Author(s): James M. Humber

Source: Religious Studies , Sep., 1974, Vol. 10, No. 3 (Sep., 1974), pp. 291-300
Published by: Cambridge University Press

Stable URL: https://www.jstor.org/stable/20005176

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access
to Religious Studies

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms
Rel. Stud. IO, pp. 291-300

JAMES M. HUMBER
Professor of Philosophy, Georgia State University

CAUSAL NECESSITY AND THE

ONTOLOGICAL ARGUMENT

The ontological argument appears in a multiplicity of forms. Over the past

ten or twelve years, however, the philosophical community seems to have
been concerned principally with those versions of the proof which claim that
God is a necessary being. In contemporary literature, Professors Malcolm
and Hartshorne have been the chief advocates of this view, both men holding
that God must be conceived as a necessary being and that, as a result, his
existence is able to be demonstrated a priori.' This claim has not gone
unchallenged; indeed, numerous writers have argued that neither Malcolm
nor Hartshorne has exercised due care in his use of 'necessary'.2 That is,
critics charge that the arguments of both men have only the appearance of
validity, for in their reasonings the defenders of the a priori proof have
tacitly assumed that God is a logically necessary being. Whether or not a
being can be logically necessary, however, is a quaestio disputata. In fact,
until recently the question was not in dispute at all-virtually all 'competent
judges' agreed that only propositions could be spoken of as logically necessary,
and thus that God must be defined as a physically or factually necessary
being. But is the statement, 'a physically necessary being exists', logically
true? Critics of the ontological argument think not; and in support of this
view they offer analyses of 'physical necessity' which, they fee], not only
give meaning to the phrase, but also show that a physically necessary
being's existence can be proven only by some kind of aposteriori investiga
tion.3
Not content to let matters lie, Professor Alan G. Nasser has recently
published an article in which he seeks to turn the arguments of these critics
inward upon themselves.4 Accepting their analyses of 'physical necessity',
Professor Nasser offers an argument which purports to show that 'whatever
1 Norman Malcolm, 'Anselm's Ontological Arguments', The Philosophical Review, Vol. LXIX
(I960); Charles Hartshorne, The Logic of Perfection (Springfield, Illinois, 1962), pp. 50-I.
2John Hick, 'God as a Necessary Being', journal of Philosophy, 57 (I960); 'A Critique of the
"Second Argument"' in The Many Faced Argument, ed. John H. Hick and Arthur McGill (New
York, I967), pp. 34I-56; Terence Penelhum, 'Divine Necessity', in The Cosmological Arguments,
ed., Donald R. Burrill (New York, i967), pp. I43-6I; A. C. A. Rainer, 'God and Necessity',
Mind, 58 (I949), pp. 75-7; G. E. M. Anscombe and P. T. Geach, Three Philosophers (Oxford, 1961)
pp. 114-I5.
Ibid., Hick, Penelhum, Rainer.
'Alan G. Nasser, 'Factual and Logical Necessity and the Ontological Argument', International
Philosophical Qzuarterly, Vol. XI (September, I97I).

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms
 292 JAMES M. HUMBER

is factually necessary must be logic

Nasser's study is worthy of considerat
argument is unsound, and should not be
however, in uncovering the reasons f
that, contrary to current opinion, the
ment, namely, that which takes 'exis
of 'God'.
In developing my thesis, my procedure will be as follows: in section one
Nasser's argument will be considered and his conclusion shown to be false;
in section two Nasser's error will be explained in terms of his having accepted
an incorrect definition of 'physically necessary being'; finally, in section three
'physically necessary being' will be properly analysed, and an argument
offered in support of the contention that there is only one ontological proof.

As a first step towards showing that God's existence can be demon

a priori, Nasser contends that:

'... every first level concept (a concept considered as the conjunction of

descriptive constituents whose possession by an object is both necess
sufficient for that object being an instance of the concept in question) ha
the following instantiational modalities: either (i) it is necessarily insta
(instantiated in all possible worlds), or (2) it is necessarily uninstantiate
stantiated in all possible worlds), or (3) it is neither necessarily instantiat
necessarily uninstantiated (instantiated in some possible worlds, uninstant
others).'2

A concept's modality of instantiation is not a matter for empirical discovery,

but is fully ascertainable a priori, i.e. it is given with the descriptive content
of the concept under consideration. For example, elephant and unicorn
obviously belong to (3), square circle to (2) and God, as factually necessary
being, falls under heading (i)-or so Nasser believes. In addition, whichever
modality does attach, must attach, i.e. a concept could not have either of
the modalities it does not have. Hence, if God is not necessarily instantiated,
it could not be necessarily instantiated. This being so, it is also true (by
transposition) that if God could be necessarily instantiated, it is necessarily
instantiated.
In order to demonstrate that God could be (and therefore is) necessarily
instantiated, Nasser accepts the 'physical necessity'-'logical necessity'
distinction insisted upon by critics of the ontological argument. But what
meaning attaches to 'physically necessary being'? In an attempt to give
meaning to the phrase, Nasser cites the following passages:
1 Alan G. Nasser, 'Factual and Logical Necessity and the Ontological Argument', International
Philosophical Quarterly, Vol. XI, p. 386.
2 Ibid., p. 397.

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms
 CAUSAL NECESSITY AND THE ONTOLOGICAL ARGUMENT 293

'What may properly be meant then, by the statement that God is, or has, necessa
as distinguished from contingent being is that God is, without beginning or e
and without original cause, or ground of any kind whatsoever."
'The necessary existence of an object, x, is defined as the existence of x with
beginning or end and without dependence upon anything other than itself.'2
'As applied to things or events, "contingent" will mean "dependent" or "cause
one thing or event being contingent upon another, "necessary" will mean "
dependent upon any other ... ." '3

Having accepted these statements, Nasser arrives at three conclusions. Fi

he agrees with Hick and others that God is, by definition, a physically
factually necessary being.4 Second, 'God is a necessary being' is taken t
equivalent to 'God is eternally independent'.5 And finally, an eternall
independent being is held to be one 'whose existence does not depend u
the existence or non-existence of anything'.6 Given these conclusions, then
Nasser argues that God's existence must be logically necessary, for a be
who depends upon nothing is one who

'. . . is conceived as capable of existing in all possible worlds (i.e., the concept
is capable of being instantiated in all possible worlds). And to say this is to say th
He could exist necessarily (God could be necessarily instantiated). Is not a be
who could exist necessarily, in the sense of in all possible worlds, a being who coul
be logically necessary?. . . and ... a being who could be logically necessary (a concep
that could be necessarily instantiated) is logically necessary (is necessar
instantiated).'7

At first glance it may seem that Nasser has succeeded in formulatin

valid a priori demonstration of God's existence. Given closer scrutiny
however, certain doubts arise. First, one finds it difficult to ascertain precis
what Nasser takes to be the meaning of 'physically necessary being'. To
sure, he accepts the Hick-Penelhum analysis of that concept, claiming t
it is quite sensible to speak of God as an eternally independent being. B
what does this mean? Certainly there can be no problem as regards Nass
use of 'eternal'-an eternal being is one without beginning or end. But w
meaning does he see as attaching to 'independent'? On the face of it, th
would appear to be two possible interpretations: (i) Nasser could be usi
that term to mean 'totally non-dependent or uncaused'; or again, (ii)
could be taking it to mean 'not dependent upon any other, self-dependent,
self-caused'.
Although certainty is unattainable in these matters, a close reading
1 Nasser, p. 389; Hick, 'God as a Necessary Being', p. 733.
2 Nasser, p. 389; Hick, 'A Critique of the "Second Argument" ', p. 348.
3 Nasser, p. 390; Penelhum, p. I59.
4 Nasser, p. 390: 'Hick would, I am sure, be prepared to admit that "God is a factually niecess
being" is a necessary truth. For any being who lacks either eternity or ontic independence (or b
is not properly called "God".'
5 Ibid. 6 Ibid., p. 401. 7 Ibid.

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms
294 JAMES M. HUMBER

the above-cited passages (notes I, 2,

and Penelhum probably opt for defin
some evidence that Nasser accepts (i)
being 'whose existence does not depen
of anything'). Now if this is what Nass
out of hand. The reasoning is simple
uncaused, it is true that his existenc
non-existence of anything. If this is gr
must be taken to cover God himself
he did not exist. Obviously, rather th
non-dependent God cannot exist in any
self-contradictory.
Clearly, if Nasser is to have any ho
that he is using 'independent' in sens
his meaning, can one justifiably co
entity could exist in all possible worlds
we must be clear as to what it means
or eternally self-dependent. A close ana
First, 'self-caused' cannot mean 'self
no first cause. This being so, Nasser's
who would insist that it is impossible
this criticism would be legitimate if G
being, for in this case 'self-caused' w
created being has to exist prior to hi
Given that God is conceived of as etern
be seen as having avoided the force o
Second, if an eternally self-depend
could end his existence. Clearly, noth
definition, totally independent of all
in his own causal activity, for action o
is to say, to hold that God could termi
to assert that the statement, 'a being
tence can have an end to his existenc
self-contradictory, its truth value can
Finally, it is evident that 'physical nec
equated, for God necessarily exists on
present and necessarily productive of
must be a contingently instantiate
instantiated one as Nasser believes
caused being is still a dependent being,
be seen as being contingent upon the
of God, this cause would appear to be h
But it is not at all clear why this pow

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms
 CAUSAL NECESSITY AND THE ONTOLOGICAL ARGUMENT 295

one possible world in which it need not be exercised (and there is noth
the definition of God as eternally self-caused which would precl
possibility), then God could not exist in that world. And if there
possible world in which God could not exist, his existence cannot be lo
necessary.'

I I

If our analysis thus far is correct, Nasser cannot have demonstrated God's
existence a priori. Now though valuable, knowledge of this sort does not
carry us very far, for it is extremely specific and almost wholly negative in
nature. At the same time, our study is not complete, for if I am not mistaken,
continuation of our analysis will lead to conclusions of much greater import.
First, in explaining why Nasser erred, we shall be driven to reject the Hick
Penelhum definition of 'physically necessary being'. Further, once the
reasons for this rejection are made known, we shall have the foundation we
need in order to provide a proper analysis of that concept. Given this
knowledge, then, we shall be in a position to argue that if Kant and Frege
are correct in their criticisms of the ontological argument, no version of
that proof can hope for success.
In attacking those forms of the ontological argument which define God
as a necessary being, both Hick and Peneihum insist upon distinguishing
between physical and logical necessity. Now as we have seen, the concept
of physical necessity makes sense only if it is equated with causal necessity.
Hence, when Hick and Penelhum claim that a physically necessary being's
existence is a contingent matter of fact, they must mean that it is an effect
contingent upon the actualisation of some cause. So much is clear. But
does this not show that their 'physically necessary' being is an entity which
is not at all necessary? What, after all, are the critics denying when they
hold that the question of God's existence can be settled only by some empirical
inquiry? Obviously they are denying that God must be conceived as existing
in all possible worlds. But this is true only because they are also denying
that there is anything in the definition of God which forces us to conclude
that he must exist in our world. That is, what the critics are really saying
is that if God were found to be eternally self-caused in our world, then he
would have to exist here. But whether or not an eternally self-caused being
exists in our world is an open question. Consequently, God cannot be
known to be logically necessary (existent in all possible worlds) because he
1 It should be clear that Nasser cannot claim that God's power of self-existence could be exercised
in all possible worlds and that, as a result, God still could exist (must exist) in all such realms. An
argument of this sort is ruled out by the fact that if this reasoning were accepted, God's power of
self-causation would have to be itself either: (A) eternally uncaused, or (B) eternally self-caused.
If (A) is accetped, the power can no more exist than an eternally non-dependent being can. Given
(B), however, the power is contingent and could not exist in that possible world in which its cause
was absent.

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms
 296 JAMES M. HUMBER

cannot be known to be physically necess

world.)
The same point can be made in a somewhat different way. What do we
mean when we say that it is causally or physically necessary that some
thing occur? Ordinarily we say that an effect (E) is causally necessary when
we believe that it must occur by virute of the activity of some cause. But
what sense of 'must' is here being employed? Do we mean that E must
occur in all possible worlds? Hardly. We all admit, for example, that there
are possible worlds in which fire need not produce heat, or ice need not be
cold. Well then, do we mean that E must occur in some possible world?
Again this does not seem to be the case, for an analysis of this sort captures
no sense of necessity. That is, even if a specific event would be causally
necessitated in some possible world, there is no sense in which it could be
said to have to be actualised, for its being is contingent upon the existence
of that world in which its cause is present and efficacious. Clearly, then, an
E is said to be physically necessary only if it is thought that it must occur in
our (actual) world. But the Hick-Penelhum definition of 'physically necessary
being' fails to capture this sense of necessity, for as they themselves realise,
an eternally self-caused being does not have to exist in our world. True,
God would be physically necessary in that world in which his power of
eternal self-causation was actualised. But this just shows that an eternally
self-caused being is one which must exist in some possible world; and as
we have seen, a being of this sort is not at all necessary, but contingent.
If the above considerations do not distort, certain conclusions seem
unavoidable. First, Hick and Penelhum succeed in showing that God is
not logically necessary (not existent in all possible worlds), only because
they hold that he is not necessary in any sense. But if this is so, their position
must be seen as inconsistent, for although they first claim that God is, by
definition, physically necessary, they later define him as contingent. In
addition, two questions remain unresolved: First, what is a physically
necessary being, i.e. what kind of being is it that must exist in our world?
And second, if the nature of such a being were able to be specified, could
Nasser's argument be reformulated, so that God's existence could be
demonstrated a priori?
As confused as things are, in Nasser's hands the issues become even more
clouded. Really, he seems to have been caught between his ordinary under
standing of what a physically necessary being is, and what Hick and
Penelhum say it is. To make sense of the concept, then, Nasser would appear
to have reasoned as follows: Even if it is true that an eternally independent
being does not have to exist in our world, he could. Further, he could exist
in all possible worlds, for as eternal and independent absolutely nothing can
stop his existence. But a being which could exist in all possible worlds,
does exist in all possible worlds. And a being which does exist in all possible

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms
 CAUSAL NECESSITY AND THE ONTOLOGICAL ARGUMENT 297

worlds is a being which is logically necessary. In addition, a logical

necessary being is one which obviously must exist in our wor
ABRACADABRA, the ordinary meaning of 'physically necessary being' h
been salvaged.
Now if the above adequately represents Nasser's line of thought,
error is both obvious and understandable. When he holds that God cou
exist in our world (or in any particular world for that matter) he must
accepting a definition of 'independent' as 'self-dependent or self-caused', for
as we have seen, an uncaused or totally non-dependent being can exist
no world. When he speaks of God as being able to exist in all possible world
however, he must be seen as having shifted his use of 'independent' f
'self-dependent' to 'non-dependent', for it is only in this way that one
make sense of his belief that the presence or absence of nothing can hin
God's existence. In essence, then, it appears that Nasser is trying to m
sense of 'physically necessary being' by conflating the two uses we ha
delineated, thus holding to the view that God is at once both self-dependent
and totally non-dependent. But if this is so, there can be no mystery as
why he felt he had validly demonstrated God's existence a priori-given
self-contradictory premises, anything can be derived. At the same time, hi
error is understandable, for in his reasonings he is simply trying to ma
some sense out of the contradictory position taken by Hick and Penelhu
After all, it was these critics who first told us that God was both physically
necessary (by definition), and yet not really necessary in any sense (b
implication from their definition of 'physically necessary being').

III

If our analysis thus far is correct, we know th

being is one which must exist in our world. We a
of God as eternally self-caused does not capture t
necessary', for an eternal and self-caused being does
world. The remaining question, then, is whethe
spoken of as having to exist in this world. I believe
entity can be specified, but before this task can
conceptual 'tools' necessary for the continuation
introduced.
As we have seen, sense can be made of God's p
'physical necessity' and 'causal necessity' are equa
then, God can be known to be physically necessary,
that causality is, as our common sense tells us,
When one turns to the philosophy of science for a j
however, one soon discovers that causal theorists
similar to that of Hick and Penelhum. Hume lon

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms
 298 JAMES M. HUMBER

in simplest form he argued as follow

nection, the conjunction of cause (
tradictory. The conjunction is not self-
logically possible that nature may
world could take the place of ours)
self-inconsistent, and causality cannot b
What Hume's argument makes clear
E must occur by referring to some p
is doomed to failure. For example, we
has the power to crush a container that
has the power to explode when ign
holding that it is the interaction of
which causally necessitates E (e.g. giv
and a p (container) with the capacit
crushing must occur when a vacuum
Hume is correct, our belief in the nec
totally without justification, for it i
could change its course (the atmosph
the container could lose its capacity t
then, we cannot say that the contain
inside.'
Until recently, attempts to avoid Hume's conclusion have met with
little or no success. In the past few years, however, Professors Madden and
Harre have provided new hope for those dissatisfied with Hume's scepticism.2
I have elsewhere identified myself with their thesis, and hold that it vitiates
Hume's scepticism by providing us with the proper analysis of physical
necessity.3 Now although I do not wish to imply that either Harre or Madden
would analyse 'physically necessary being' precisely as I do, there are
certain features in Madden's formulation of that theory which may be of
use to us in our investigation.
First, Madden recognises the force of Hume's C. E argument, and for
this reason holds that the mere presence of powerful p's guarantees nothing.
On the contrary, rather than explaining phenomena, it is the actualisation
of powers and capacities which themselves require explanation. This is done
by referring to the natures of the particulars involved. In our crushing
1 Admittedly, this analysis grossly simplified Hume, and fails to do justice to the variety of ways
in which 'cause' is used in ordinary language. Simplification does not necessarily lead to distortion,
however, and since the above statement is adequate for our purposes, nothing would be gained
by the introduction of greater detail.
2 R. Harre, The Principles of Scientific Thinking (London: 1970); E. H. Madden, 'A Third View
of Causality', The Review of Metaphysics, Vol. 23 (I969); 'Hume and the Fiery- Furnace', Philosophy
of Science (197I); 'Causality and the Notion of Necessity', Boston Studies in the Philosophy of Science
(Dordrecht, Holland: Reidel, I969).
3 James Humber and E. H. Madden, 'Nonlogical Necessity and C. J. Ducasse', Ratio, Vol. XIII
(December I97I); James Humber and E. H. Madden, 'Natural Necessity', The New Scholasticism,
Vol. XLVII (Spring, I973).

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms
 CAUSAL NECESSITY AND THE ONTOLOGICAL ARGUMENT 299

paradigm, for instance, the weight and pressure of the air explains why i
must have the power to crush; the shape, and material out of which the
container is made explain its capacity to be crushed, etc. Given this kind o
analysis, then, one can make sense of the statement 'container x must
collapse when it has no air inside'. If a change in the course of nature wer
to occur and the atmosphere's power to crush or the container's capacity
to collapse disappear, the loss of power and/or capacity could be explaine
only in terms of some change in the natures of the p's involved. But if on
or more of the p's changed in its nature, we would not have the conjunction
of C (air, container, plus vacuum) and r E, but 'C' (non-air, or non
container, plus vacuum) and ,E. Needless to say, that ' "C". -' E' i
always logically possible, in no way demonstrates that 'C. E' is self
consistent.'
Hume's error was that he thought ' "C". r E is never self-contradictory'
entails 'C. E is never self-contradictory'. Or to put it another way, he
continually confused two senses of necessity. An expression is logically
necessary if it holds true in all possible worlds. In this sense, of course, the
statement 'container x collapses when it has no air inside' cannot be neces
sarily true, for there are possible worlds in which the p's involved could have
natures quite unlike those they now possess. But the fact that this must be
admitted does nothing to alter the conclusion that, given the present natures
of the p's of this world, it is physically necessary that container x collapse
when the air is pumped from its interior.
Accepting Madden's analysis, it is an easy matter to explain why Hick
and Penelhum failed to capture any sense of necessity in their definition of
'physically necessary being'. Simply put, their definition as 'eternally self
caused' presupposes that it is God's own power of self-causation which
physically necessitates his existence. But unless it can be shown why this
power must be actualised in this world, God's existence cannot be known to
be physically necessary. Now how could one go about showing that the
power must be actualised ? Clearly there is only one way, namely, by showing
that there is a real (i.e. existent) nature which has the character of being
eternal and independent of all others. But if this is so, 'physically necessary
being' means 'existent being, eternally independent of all others'. And this
insight is of extreme importance, for once it is understood the monogenetic
character of the ontological argument can be demonstrated.
First, there are only two ways in which one could go about showing
that God is a physically necessary being: he could try to do it in an a priori
manner, or via aposteriori means. For those, like Nasser, who seek to establish
God's existence a priori, the latter alternative is ruled out. If one chooses the
former course of action, however, his procedure must be illegitimate, for if

1 I have been extremely liberal in my statement of selected aspects of Madden's theory. I trust,
however, that his intent is fairly represented.

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms
300 JAMES M. HUMBER

the insights of Kant and Frege are so

physically necessary being, i.e. as existe
others. Indeed, one cannot (as Hick a
physically necessary and then see it as a
exists. On the contrary, the real que
necessary?' Then too, there is another
as physically necessary. That is, any at
ignores the distinction between logical
because if one holds that 'God is a phys
definition, God's physically necessary exis
being turned into logically necessary ex
hold that God is physically necessary and t
logically contradictory, for if God is defin
physically, but logically necessary. Reme
is one which must exist only in our world
necessary (existent), he must, by definitio
what else could one mean by 'logically n
exists in all possible worlds'? This was N
gation simply gives added support to his
In sum, then, what do we have? Wel
necessary being' is correct, there can b
and this is the one which takes existenc
concept God. Clearly there is no significan
God is, by definition, existent, and that he
eternal, and independent of all others. T
logically necessary being, one can only me
as existing in all possible worlds. But a b
defined as existent. In each and every cas
true, only if 'existence' can be used as a pr
'defenders of the faith' had best return
in an attempt to show that they are eit
it is certain that God will be on their si
anyone seriously doubt that they will be u

This content downloaded from

45.188.74.121 on Fri, 20 Jan 2023 19:34:51 UTC
All use subject to https://about.jstor.org/terms