Monday, April 03, 2006

Arguing for God


There is, in polemical and philosophical theology, a tradition of arguing for the existence of God. This is primarily an exercise in offensive apologetics. For you have unbelievers who argue against the existence of God. Hence, you have Christians who mount counter-arguments.

Because the Bible is written by and for the Jews, it doesn’t generally argue for the existence of God.

But when, on rare occasion, it does turn to a pagan audience, we see it move into the realm of apologetics. Acts 17 is the best-known example, but a neglected instance is Isa 40-48. As one commentator puts it:

“The unit opens with Yahweh summoning the nations to appear in court for a trial. The claims of the foreign gods will be tested according to legal rules…the force of the argot in both parts of the trial appears to be that the claim to true divinity rests on the ability not only to control the course of future evens, but also to have predicted the events before the occurred. Consequently, the ability to match the prediction with is fulfillment can then be tested rationally in the trial,” B. Childs, Isaiah (Westminster 2001), 317,21.

A commonly lodged complaint concerns the intricacy of the theistic proofs. Only an intellectual can follow the often-recondite reasoning.

There’s some truth to this, but it’s hardly distinctive to the existence of God. There’s something inherently artificial in trying to prove the existence of many things we take for granted.

When you look out a window and see a tree, how do you prove that there really is a tree. This is no easy matter. It is well-nigh impossible to prove the existence of your humble tree, even though just about everyone believes in the existence of the tree they see out their kitchen window.

So while the exercise of proving God’s existence may often seem to be pretty artificial and abstruse, that is true for proving the existence of most anything, even the most mundane and common sense objects of human experience.

What I’ll do in this little essay is to outline some of the arguments for the existence of God. In the age of modal logic and Bayesian probability theory, one could devote a book to each theistic proof, or variation thereof.

So my aim is not to present full-fledged arguments. Rather, my more modest aim is to present certain argumentative strategies which can be fleshed out to any desired degree of sophistication.



This argument reasons from the existence of the world to the existence of a divine creator. There are many versions.

One popular argument makes use of modern cosmology. But a danger with scientific versions of the cosmological argument is that they are tied to the current state of physics, and scientific theories about the origin of the universe are vulnerable to revision.

Another difficult is attempting to square modern cosmology with a traditional reading of Gen 1.

Metaphysical versions of the cosmological argument are more stable in that, if they are cogent, their cogency is independent of the vicissitudes of modern science.

i) The Kalam cosmological argument

This goes back to the Byzantine theologian, John Philoponus. He argued against the eternity of the world on the grounds that a world without a beginning would have an infinite past, which is impossible.

His argument turns on Aristotelian and pre-Socratic objections to the actual infinite. Zeno argued that it would be impossible to traverse even a finite span (in time or space) by finite increments—much less an infinite span. Aristotle tried to resolve the paradoxes of Zeno by affirming a potential infinite, but rejecting an actual infinite.

The argument has been updated in light of modern set theory. Set theory admits the existence of an actual infinite. Indeed, a hierarchy of infinite sets.

However, the actual infinite of set theory is a timeless or abstract infinity. Indeed, that’s what makes it actual. It is a given totality.

So the essential logic of the Kalam argument remains.

There can be an actual infinite, but only an abstract, not a concrete infinite. You cannot have a temporal infinitude, for a temporal infinite is incomplete, whereas an actual infinite is unincreasable.

This argument has been popularized by William Lane Craig and J. P. Moreland, although I don’t think they cast it in its strongest possible form.

Here are some examples:

ii) The PSR-cosmological argument

This version of the cosmological argument is a special case of the Principle of sufficient reason.

Leibniz wrote:

“Our reasonings are founded on two great principles, that of Contradiction…And that of Sufficient Reason, in virtue of which we consider that no fact can be real or actual, and no proposition true, without there being a sufficient reason for its being so and not otherwise, although most often these reasons cannot at all be known by us.”

Put more compactly, the principle of sufficient reason says that every truth of fact (contingent true proposition) has an explanation.

This is not an axiom you can directly demonstrate, but it’s difficult to challenge without undermining rationality.

The PSR-cosmological argument has both a general form and a specific form:

i) Why is there something rather than nothing?

ii) Why does this particular world exist instead of some other possible world?

An underlying of the general form is that contingent existence is not self-explanatory, unlike nonexistence. It also assumes that the universe is a contingent state of affairs.

One reason for the assumption of contingency is that it’s easy to conceive of any number of trivial variants on the real world. Little things that could have been otherwise. If I was five-foot eleven instead of five-foot ten. If I had blue-gray eyes instead of blue eyes, and so on.

The question, then, is what accounts for the instantiation of this world rather than some other possible world?

Answer: an omniscient agent had to choose to instantiate this possibility out of all other possible worlds.

This argument has the advantage, not only of implying the existence of a divine creator, but an agent who vastly exceeds the effect inasmuch as we must account, not only for the particular effect, but the relation between this effect and all the alternatives.

A possible blocking maneuver against the specific form of the argument is appeal to the multiverse. In one version of cosmology, every possibility is, in fact, realized in some parallel universe.

But there are a couple of problems with this move:

i) It’s an ad hoc maneuver.

ii) Even if, for the sake of argument, we were to grant that hypothesis, it would run afoul of the general form. Appealing to a multiverse only pushes the problem back a step, for you must still account for the existence of the ensemble as a whole.

For examples:


i) Argument from analogy

In its intuitive form, the teleological argument is often taken to be an argument from analogy.

Nature exhibits a level of order and complexity which gives it the appearance of design. Just as we attribute design to a human artifact, we should attribute design to a natural object or relation if it exemplifies the same sort of features.

If you concede that nature has the appearance of design, then the burden of proof is on the unbeliever to explain away this appearance without recourse to actual design.

This inference has been attacked on several grounds:

a) It begs the question. In the case of a human artifact, like a watch, we know it to be of human manufacture, so we infer design on the basis of experience.

However, that seems to be a pretty weak objection.

For one thing, we might stumble across an unfamiliar object that we would still ascribe to intelligent design. Suppose we discovered an alien spacecraft that crash-landed on earth.

At first the technology would be so advanced as to be unrecognizable. But as we began to figure it out, we would rightly attribute the artifact to intelligent design.

For that matter, is the design inference based on experience? How do I know that a watchmaker intends to make a watch? I don’t have direct access to his motives. I simply judge by the end-product. Since the artifact functions as a timepiece, I assume that it was designed to function as a timepiece.

b) Another objection is that every argument from analogy is also argument from disanalogy.

This is true, but analogical reasoning is a fundamental feature of human reason.

c) Yet another objection is that the argument from analogy is anthropomorphic. We are projecting ourselves onto the object, which begs the question.

But this objection either proves too much or too little. It’s only natural for a human observer to take himself as the point of reference, for he has no other point of reference to assume. If this is illicit, then all of natural science is illicit.

In addition, it isn’t clear that the teleological argument is, in fact, an argument from analogy. Suppose I stumble upon a rock with what appears to be a pitted surface. The indentations appear to be random. They appear to be produced by natural erosion.

I photograph the rock with my digital camera, and input the image into my computer, then run a comparative analysis. To my surprise, the indentations correspond to the star chart of a constellation.

I therefore infer that, appearances notwithstanding, the markings were left there by a human stargazer.

Or suppose they correspond to a constellation which was not known to ancient, naked-eye astronomy, but only discovered by high-powered telescopes?

I’d therefore infer that ETs had visited our planet some time in the past.

ii) More recently, Bayesean probability theory and the “law” of specified complexity (an extension of information theory) have been devised to give our pretheoretical intuition of design a more quantifiable and testable form.

These are not teleological arguments, but rather, the technical apparatus for generating teleological arguments, such as:

iii) Cosmic fine-tuning

Recent appeal is made to the highly unlikely convergence of various conditions or constants without which life would be impossible.

In its particular form, this argument is vulnerable to scientific revision. Still, it seems that any scientific revision would merely refine rather than redefine the basics of the argument.

This argument is best deployed by someone with the requisite scientific expertise.

An objection to the fine-tuning argument is the anthropic principle. The contention is that there’s nothing surprising about the fact that we find ourselves in a universe fine-tuned for life; after all, if the universe were not fine-tuned for life, there would be no observers in the first place!

But this objection confuses the order of being with the order of knowing. The fact that certain physical conditions must be in place for there to be observers of those very conditions does nothing to probabilify the conditions in question.

iv) Irreducible complexity

It is said that certain biological structures cannot be erected from simple to complex by a linear, step-wise process, for they entail a relation or interrelation in which two or more relata must be in place for the relation to obtain. So there is a least-upper threshold of complexity. A mousetrap is a convenient illustration.

This argument is only as good as the biological phenomenon it educes. Is a given biological structure in fact irreducibly complex?

One objection to the teleological argument is that naturalistic evolution has displaced the need of teleological reasoning. For order and complexity can emerge by a non-directive process.

But there are several problems with this objection:

i) Taken on its own grounds, evolution, even if true, would not necessarily displace intelligent design. Rather, teleology is an ends-means relation, and evolution might be the appointed means of achieving a certain end, a la theistic evolution.

So one could construct an evolutionary version of the teleological argument.

ii) The objection is more germane to biology than cosmology.

iii) The objection is only as good as the theory of evolution. Many opponents of evolution remain critical of this theory precisely because, in their view, evolution is unable to explain certain teleological features in nature.

Richard Swinburne, William Dembski, and Michael Behe are contemporary proponents of the teleological argument. For examples:


As I understand it, Anselm's argument takes the form of a logical dilemma. An atheist cannot deny the existence of God unless he has a concept of God. This concept is the idea of a greatest conceivable being. What is more, even though the concept is mental, it is a concept of an extramental being. Although the thought of God subsists in the mind, the object of thought is the idea of God as an extramental existent. That much is necessary even for purposes of negation.

So what is the gold-standard that redeems the object of thought? What is it that makes either the affirmation or negation of a proposition to be true or false? And what if the proposition is about a being that is greater than the thinker who thinks him?

Gaunilo replied by introducing the counterexample of the blessed isles. But this fails on couple of counts: (a) it raises a similar dilemma, for what is there that grounds both the bare possibility and its practical denial? Of what is this possible? Of what is this false? To what does it refer? (b) Unlike the idea of the blessed isles, the idea of God is the idea of a being that is more than an idea.

As a general matter, how you rate the ontological argument has less to do with the specifics of the argument than with your general theory of knowledge. If you're in the nominalist/empiricist camp, you're predisposed to reject it; if you're in the realist/rationalist camp, you're predisposed to accept it.

Alvin Plantinga has formulated a modal version of the ontological argument, assuming axiom S5 of modal logic:

(i) By definition a maximally great being is one that exists necessarily and necessarily is omniscient, omnipotent and perfectly good. (Premise)
(ii) Possibly a maximally great being exists. (Premise)
(iii) Therefore, possibly it is necessarily true that an omniscient, omnipotent and perfectly good being exists (By i and ii)
(iv) Therefore, it is necessarily true that an omniscient, omnipotent and perfectly good being exists. (By iii and S5)
(v) Therefore, an omniscient, omnipotent and perfectly good being exists. (By iv and since necessarily true propositions are true.)

The axiom S5 says that if a proposition is possibly necessarily true, then it is necessarily true.


The basic idea is that truth is a relation between a true belief or mental representation and a corresponding object.

Most secular thinkers adopt some version of the correspondence theory. If so, then truth is a relation between a truth-bearer and its object (a fact).

Philosophers differ over what functions as the truth-bearer, whether a belief, proposition, sentence, or something else.

Speaking for myself, to identify a truth-bearer with a sentence only pushes the question back step, for a sentence is simply a way of encoding a proposition—a storage and retrieval mechanism.

Likewise, the same proposition is expressible in different words.

The most plausible account is to say that a truth-bearer is a belief, which, in turn, presupposes a believer.

So you have a relation between a believer—via the truth-bearer (a belief) and the factual object to which that belief does or does not correspond.

The belief is a true belief if it corresponds to the factual object or state of affairs, and a false belief if it fails to correspond to the factual object.

But being a relation, truth cannot exist unless both relata—object and truth-bearer—coexist.

If there were no sapient beings, there would be no truth.

Assuming there is no God, nothing is true before the advent of intelligent life, or true after the extinction of intelligent life. It isn’t true that the earth was the third planet from the sun until a human being came on the scene which entertained that proposition.

Further refinements are possible, between object-based and fact-based correspondence theories, but that doesn’t affect the primary thesis.

Or, if you prefer the coherence theory of truth, then truth is once again a relation property—in this case, a relation between ideas. This also requires a mind.

There are other, less dominant theories of truth.

For an example of self-refuting secularism in relation to truth, cf.


This is a transcendental argument. It assumes the existence of abstract objects, viz., numbers, logical laws, propositions, possible worlds.

These abstract objects or universals function as truth-conditions. Without them, rationality is impossible.

Without them you cannot have infinite mathematical sets or propositions or counterfactuals or logical laws.

The modal argument then seeks to ground the existence of abstract objects as properties of a property-bearer. They are mental entities which inhere in the infinite and timeless mind of God.

Here are some examples:


This is another transcendental argument. It assumes the existence of human reason. It then argues that a precondition of reason is a reliable belief-forming mechanism in conjunction with an environment conducive to the formation of true beliefs. It then argues that only intelligent design can meet these conditions.

Alvin Plantinga has put this argument on the map:


The basic argument is that unless human beings are the handiwork of a divine creator, creatures which exemplify abstract ethical attributes, creatures designed to function in a certain way, then there is no basis for moral absolutes.

The strength of the moral argument is that it enjoys more popular appeal than arguments of a more cerebral nature.

The weakness of the moral argument is that an unbeliever may agree with the conclusion and adopt an amoral worldview.

For example:


Just as most unbelievers are not intellectuals, most believers are not intellectuals. The primary reason that a Christian believes is due to his personal experience of God’s grace and providence in his life.

A limitation of the argument from experience is that you have to be an insider to appreciate the full force of the argument. So it’s rather irrelevant to the outsider.


A common objection to theistic proofs is that they all fall short of proving the God of classical Christian theism.

There’s some truth to this. The theistic proofs have been devised over the years by a variety of different thinkers. Hence, they’re somewhat piecemeal.

But, in principle, there is a way of filling the gaps. You simply draw up a list of divine attributes or activities, then devise a rational argument for each.

Your doctrine of God supplies the object of proof. Hence, your doctrine of God also supplies the blueprint. You structure your arguments around the nature of the object.


  1. Very impressive summaries.

  2. If you're in the nominalist/empiricist camp, you're predisposed to reject it; if you're in the realist/rationalist camp, you're predisposed to accept it.

    Could you briefly describe these two camps?

  3. A nominalist/empiricist believes that the object of knowledge is limited to sensible particulars. Hence, inductive observation, which can only yield a probabilistic conclusion, constrains the scope of our knowledge.

    By contrast, a realist/rationalist believes that the object of knowledge consists of abstract universals or truths of reason (necessary truths). Hence, he believes in the possiblity of apodictic proof on the basis of a priori arguments.

    Both positions amount to half-truths.