Luke Morris

4/25/2004

The Benefits of Natural Theology

            Since the Middle Ages, philosophers have made numerous attempts to prove the existence of God.  They did not develop such arguments so much because God’s existence was in doubt, nor did they mean for logical proofs to serve as the ultimate means of conversion to Christianity.  They rather meant to use natural theology as a tool for reaffirming faith – hence Augustine’s “faith seeking understanding.”  In contemporary thought, philosophers use these arguments for a similar purpose:  in short, to demonstrate that belief in God is rational.  Granted, this is not natural theology’s only use, but it is an important one.  And the three main proofs of God’s existence – the ontological, teleological, and cosmological arguments – offer valid demonstrations of this conclusion.  While these arguments may not convert many people to a genuine belief in God, they do help to show that such a belief need not be anti-rational.

            The ontological argument, first developed by St. Anselm, is a fairly simple a priori proof from definition.  To begin with, Anselm defines God as that being than which a greater cannot be conceived (98).  And since it is greater to exist in reality as well as in the understanding than in the understanding alone, that than which a greater cannot be conceived must be conceived as existing.  For if we imagine this being as not existing, than we can imagine an even greater being that does exist – which shows that the first being we imagined was not really that than which nothing greater can conceived, that it was not really God.  Therefore, God, by our conception of Him, must exist; we cannot even conceive of Him as not existing (99).  The first question this proof raises, of course, is why it is greater to exist in reality than to exist in the understanding alone.  But if we define greatness as any positive quality or attribute, then we would have to admit that something that exists in reality can affect reality in a way that a purely mental concept cannot, which would seem to give it another positive quality over and above our mere conception of it.  So this question poses no real difficulty for Anselm.

            Immanuel Kant, however, offers an objection that does pose a problem for the advocate of the ontological argument.  Existence, Kant argues, is not a property or an attribute at all.  Not every predicate serves as a straightforward modification of the object.  While my conception of “a dog” might alter were I to think of “a brown dog,” my thinking about “an existent brown dog” does not alter my conception of the dog I was already thinking of – it merely alters my conception of the world, in that I now imagine myself as living within a world that contains a brown dog.  But an attribute or property of a thing must change our conception of that thing.  Hence, our conception of God as an omnipotent, omniscient, omnibenevolent, existent being is no different, and no greater, than our conception of this being as non-existent, since the only thing that changes in thinking of God as existing is our conception of the world, not our idea of God. 

            The second major objection to the ontological argument deals with whether the term “God” is used as a referring or distributed singular term.  A referring singular term entails a particular thing, while a distributed singular term makes a general statement, as one might apply to a species.  In other words, “this chalkboard” is a referring term; “the chalkboard is a teaching tool” uses “chalkboard” as a distributed term.  Obviously, then, we cannot give a stipulative definition of a referring singular term, as that term only applies to one particular.  The problem this shows in the ontological argument is that the fact that a description of a thing might be right, such as “God exists,” does not mean it is true by definition.  If we use “God” as a referring singular term throughout the argument, the first premise of the argument is incoherent – we cannot define “God.”  If, on the other hand, we treat “God” as a distributed singular term throughout, then by reason of the first premise, nothing could count as God by the third premise (to exist in reality is greater than to exist in the understanding alone) unless He exists; or, if God exists, He exists, which is trivial.  But if we use the term “God” as distributed in the first premise and as referring in the conclusion, we commit the fallacy of equivocation.  The ontological argument, therefore, does not seem to prove much of anything.

            Alvin Plantinga restates the argument, though, in a way that allows it to slip past these objections.  Plantinga’s version rests on the assumption of metaphysical possibility, made concrete through hypothetical “possible worlds.”  The first premise states that “there is a possible world in which maximal greatness is instantiated” (119).  A being has maximal greatness only if it has maximal excellence in every possible world.  Maximal excellence includes omniscience, omnipotence, and moral perfection – the qualities we ascribe to that than which a greater cannot be conceived.  According to Plantinga, “now we no longer need the supposition that necessary existence is a perfection; for obviously a being can’t be omnipotent . . . in a given world unless it exists in that world” (117).  Thus he sidesteps Kant’s objection.  But is it just an assumption to say that there is such a possible world?  Not really.  Because there is no contradiction in imagining a universe in which maximal greatness is instantiated, such a state of affairs is not logically impossible; if it is not logically impossible, then it is logically possible; if it is logically possible, the argument goes, then it is logically necessary.  Therefore, God exists.

            This argument is certainly a strong one, but even Plantinga admits that it does not prove the existence of God.  Thomas Aquinas also rejects the ontological argument, not because he sees it as invalid or unsound, but because he does not believe that it is a proof for us.  As a purely a priori, analytic argument, the ontological proof does not give us any way to connect this concept of a necessary being to our experience of everyday reality.  Or, we might say, it feels like a trick – we cannot see how it relates to our lives.  I am inclined to agree with Aquinas on this one.  As Plantinga states, “An argument may be sound . . . without in any useful sense proving God’s existence” (120).  Still, the argument does show, in one way, that belief in God is not unreasonable.

            To give a proof for God’s existence with a stronger empirical connection, then, some philosophers have used the teleological argument, the argument from design, which asserts that the universe is too orderly, with too fine an adjustment of means to ends, to have arisen by mere chance.  The ordered structure of the world must have some purpose, or telos, which necessarily presupposes an intelligent designer.  The basic formulation of the argument is William Paley’s analogy of the watch, in which the design of the universe is likened to the design of a machine.  Such a miraculous design as we find in nature cannot have arisen from the mere workings of the laws of nature, for “a law presupposes an agent, for it is only the mode according to which an agent proceeds: it implies a power, for it is the order according to which the power acts” (160).  It thus makes no sense to ascribe the telos of design to the law of nature, since that law cannot be an efficient cause of phenomena.  David Hume restates this analogy in his Dialogues through the character of Cleanthes, as he says, since the world is one great machine composed of an indefinite number of smaller machines, “the Author of Nature is somewhat similar to the mind of man, though possessed of much larger faculties” (169).  The advocate of the Design argument attempts to prove the existence of God a posteriori.

            Hume argues, then, that the design analogy is weak.  We have no real justification for anthropomorphizing the causes of the universe.  But man cannot know anything of the universe in purely a priori terms, since “experience alone can point out to him the true cause of any phenomenon” (171).  But if all we know of the order of the universe are matters of fact, we cannot prove that the arrangement and adjustment of final causes necessitates a designer as efficient cause, since all we know of cause and effect are our past experiences of their constant conjunction.  That is, there is no contradiction in thinking that all matter contains its origin within itself, and falls into order by itself, any more than a contradiction arises in prescribing an intelligent designer to that order.   All of this, of course, the proponent of the teleological argument may agree to, since it generally amounts to a refutation of a priori evidence of God, while the design argument runs a posteriori.  But Hume then takes this one step further, presenting a number of arguments to show that the disproportion between a man’s creating a machine and a god’s creating all of nature is so great that any comparison is mere speculation – we are not justified in drawing a conclusion of the whole from our experience with its parts (173).   

            In arguing his case, Hume puts forth a number of suppositions that he claims are at least as likely as the analogy of intelligent design.  R.G. Swinburne, though, takes issue with Hume’s major objections, attempting to show that they all amount to invalid straw-man arguments (189).  Swinburne claims that the teleological proof, properly articulated, argues from regularities of temporal succession, the operation of natural laws, to a designer of that succession, rather than from regularities of co-presence or spatial order.  Arguments from spatial order open themselves up to empirical objections to which arguments from the order of succession do not – occasions of spatial disorder, for one, or the possibility of a purely scientific explanation of the phenomena (194).  Using the latter form of the argument, Hume’s objections do not hold.  Using Ockham’s Razor, Swinburne argues that Hume in several cases unnecessarily complicates things, by claiming that we are not justified in only partially anthropomorphizing God (why not go all the way and give Him a body?) or in positing only one designer (why not a Pantheon?), and that the simpler explanation of a free, non-bodily, rational agent as the cause of the universe makes more sense (199).  And on Hume’s assertion that the world’s generation is more vegetative than mechanical, Swinburne shows that, though this may be correct as regards the moment-to-moment spatial order that we see, what we are trying to explain is the cause of the laws of nature according to which a vegetative organism develops, and this regularity of succession can only be satisfactorily explained by an intelligent designer (200).  Still, Swinburne does grant Hume’s conclusion that, at most, we are justified in believing that the design of the universe bears some remote resemblance to the creations of human intelligence.  While both man-made order and the order of the world are types of order, we have difficulty in saying how similar they are.  Indeed, the universe is ordered on so magnificent a scale that we cannot really explain what it would be like for a rational being to hold responsibility for all of it (201).

            The most popular objection to the teleological argument (it is not one that Hume advocates) states that, despite the extreme improbability that the world would, by chance, become orderly enough to support human life, this would eventually happen given an infinite sequence of chance events (203).  [This is similar, but not the same, as Hume’s argument from the principle of plenitude – that all possible universes come to pass, and we can only exist in an orderly one, so of course ours is orderly (208).]  Ian Hacking shows, though, that this objection commits what he calls the “inverse gambler’s fallacy” (202).  One who commits this fallacy assumes that, because he sees one unlikely result, that it must be the result of many attempts that would have made that result more likely.  [Hume’s argument does not necessarily commit this fallacy – it does in Wheeler’s adaptation of the long sequential string of universes, but not in Carter’s theory of the coexistence of all possible universes (209).] The likelihood of any one chance event occurring does not increase with subsequent independent chance events.  Just because a gambler rolls an unlikely double-six does not mean that he has been rolling for a long time – indeed, he could be on his first roll (205).  The same holds for the formation of the universe.  An infinite sequence of chance events makes it no more likely that our universe would have the order it has than would a single chance event (207). 

What the dispute over the teleological argument reduces to, in Hacking’s view, is a case of those who say that the world is indeed strange, but strange things happen by chance, versus those who claim that the world is strange, therefore it must have an intelligent designer (211).  Hume would say that we are not justified in positing such a designer, since we do not know what it would be like for a being to create the laws of nature that bind us.  The design advocate might respond that our limited knowledge does not prevent us from drawing the analogy from the knowledge that we do have.  While it does shakily stand on the strength of the analogy, though, the argument does have the benefit that it is empirically based – we can relate it to our experience of the world, which we cannot do with the ontological proof. 

The cosmological argument for the existence of God seems to draw a middle ground between the a priori and a posteriori, and hence it seems to stand on the firmest foundation – it argues from the causality of existence, a matter of fact, and from the analytic nature of dependent and independent beings.  A dependent being finds the explanation for its existence in the causal efficacy of another, whereas an independent being’s cause lies within its own nature.  Samuel Clarke sums up the argument thus:  “An infinite succession . . . of merely dependent beings, without any original independent cause; is a series of beings, that has neither necessity nor cause, nor any reason at all of its existence, neither within itself nor from without:  that is, ‘tis an express contradiction and impossibility” (138).  In other words, a series or set of beings, each dependent on another, requires a primary cause for its own existence.  Even if the series were to extend infinitely in either direction, it still requires an external reason why the series has the members it has, rather than other members or none at all.  As William Rowe states it, every being is either dependent or independent; there either is and independent being or all beings are dependent; and since not all beings are dependent, there must be an independent being, and hence a necessary being (142). 

Hume criticizes the cosmological argument on the grounds that one cannot demonstrate a matter of fact.  As an a priori demonstration, the cosmological proof is not justified in speaking of causes or of necessarily existent beings at all.  As Cleanthes states, “There is no being . . . whose non-existence implies a contradiction.  Consequently there is no being whose existence is demonstrable” (140).  Hence, there is no being whose existence is necessary.  Indeed, it is absurd to even search for a cause for an infinite causal chain, since nothing could be prior in time to that chain.  Why do we even seek a cause for the whole series?  Each particular finds its cause in its predecessor; the “uniting of these parts into a whole . . . is performed merely by an arbitrary act of the mind, and has no influence on the nature of things” (141).  When we have explained the cause of each member of the series, we have explained the cause of the whole.  In fact, attributing the ultimate cause of the existence of all the dependent beings in the universe to a single necessary being would limit our admiration of the order of things, and hence our admiration of God, by applying necessity to them.

The real strength of the cosmological argument, though, rests on its first premise: the strong form of the Principle of Sufficient Reason, which states that each being must have an explanation for its existence, a reason why it is existent rather than non-existent, either within itself or from some outside causal source (143).  [The weak form of the Principle claims that there must be an explanation for every being that begins to exist; this form does not support the argument as applied to an infinite chain.]  It seems that we cannot defeat the argument unless we reject this first Principle, as Hume wants to do.  Since we can reject this with no contradiction, there is no necessity that denies that the world could have just sprung into existence, or that it could have always existed in itself, as a brute fact.  Rowe shows, however, that if we do accept the principle – and common sense might say we should – then two strong criticisms of the argument cannot work.  The first objection, from Bertrand Russell, claims that “it makes no sense to ascribe the property of having a cause or explanation to the infinite collection of dependent beings” (152); and if we treat the whole set as an abstract entity, it might be illogical to attribute a cause to it.  But, Rowe counters, it does make sense, according to the Principle of Sufficient Reason, to search for a reason why the set has the members it does rather than none at all – and if the explanation cannot be found within the infinite collection of dependent beings itself, this reason must lie outside of the collection, an independent being.  The second objection, then, comes from Hume and Paul Edwards, who hold that “If the existence of every member of a set is explained the existence of that set is thereby explained” (154).  The problem with this criticism is similar to the previous one.  If all we know is that each dependent being’s existence is explained by another dependent being, we still do not know why these dependent beings exist rather than none.  Thus, “from the fact that the existence of each member of a collection is explained it does not follow that the existence of the collection is thereby explained” (155).  If I go to the zoo to accompany someone else, and that person went to accompany another, and on and on to an infinite collection of people, each going there for the sake of another, why did any of us go to the zoo at all?

Again, the cosmological argument assumes that there must be a reason why something exists instead of nothing – that is, existence requires a causal explanation.  If we accept this strong form of the Principle of Sufficient Reason, but deny the existence of an independent, necessary being as first cause, we run into an infinite causal regression of dependent beings.  It seems that, to properly reject the argument, we would need to reject the Principle on which it stands.  And, as shown before, there is no contradiction in doing this.  We could just treat the world as a brute fact, with no explanation for its existence, and be rationally justified in doing so.  In other words, we could reduce ourselves to Humean skepticism.  But this might be an unsatisfying view of the world – a world without necessary causes. 

We now see the strength of natural theology.  Though none of the arguments are invincible, they are all valid supports for belief in God.  The ontological argument is a nifty piece of a priori reasoning, even if it is divorced from the phenomenal world.  And the teleological proof, though rested on a potentially weak analogy, gives us a sensible, human notion of the order of the universe.  The cosmological argument then delivers the punch of causality – if we are to believe in sufficient causal conditions for existence over non-existence, we must believe in a primary, independent, necessary cause.  While none of these proofs would convert me to a genuine belief, they do provide rational grounding for the theist’s faith.