Contingency Argument

William Lane CraigThe theologian William Lane Craig presents a version of Wilhelm Leibniz’s contingency argument[i] as follows:

  1. Anything that exists has an explanation of its existence, either in the necessity of its own nature or in an external cause.
  2. If the universe has an explanation of its existence, that explanation is God.
  3. The universe exists.
  4. Therefore, the universe has an explanation of its existence (from 1 and 3).
  5. Therefore, the explanation of the existence of the universe is God (from 2 and 4).


All evidence for premise 1 is drawn from our success in finding ex­planations within the natural realm—material explanations trans­latable into the language of physics. Let me emphasize that these explanations, these physical causes, are invariably within the natural realm. Yet premise 4 presumes that, if there’s anything beyond the natural realm, it must (via premise 1) have an ex­planation as well.

You may recognize this as an appeal to the principle of sufficient reason, de­bunked earlier. Extrapolating outside the relevant domain is an error well understood by statisticians studying phenomena with­in the nat­ural realm. Premise 4 commits this blunder in the worst imag­inable way by assuming that we can extrapolate from premise 1 to draw con­clusions completely beyond the natural realm.

Given that all evidence supporting premise 1 consists of material causes, we might be tempted to conclude that, no matter how far back we look in the chain of causation, we’ll always find another material cause. That’s sheer conjecture. Granted, it’s more sensible in light of our experience than any alternative conjecture, but it’s still conjecture. To commit oneself to this conjecture would be a mistake. Craig com­mits a far more egregious mistake by say­ing that premise 1 supports the specula­tion that the cosmos has an immaterial cause.

In support of premise 2, Craig rightly points out that if a cause is a material cause then it is, itself, part of the material realm. There­fore it cannot count as the cause of the material realm.

He goes on to say that for something to be the cause of the mate­rial realm, that cause must be immaterial. In other words, if we trace back through all the causes within the material realm, and if we encounter the very first material cause, which we can call M, then if we find the cause of M, that cause must be immaterial.

The supporting argument just described takes a form that phi­losophers call a disjunctive syllogism. Craig is arguing that if there’s a first cause, (1) M or Not-M is the first cause, (2) M is not the first cause; therefore, Not-M is the first cause. The second premise of this supporting argument is logically equivalent to saying “Not-M is the first cause.” That statement, you’ll notice, is identical to the conclu­sion of the supporting argument. Placing the conclusion of an argu­ment in one of its prem­ises is an informal fallacy known as “begging the question.” This kind of argument proves nothing—or more pre­cisely, it can be used to prove anything. For that reason, philoso­phers dismiss this kind of argument as ineffective.

Insisting that the cause of the material realm must be immaterial discounts the possibility that the material realm has no cause, that it’s material causes all the way back, the only conclusion congruent with our experience. It also presumes the plausibility of immaterial causation, even though no imma­terial cause has ever been identified or even adequately defined.

The “material-immaterial” dichotomy is analogous to creating two classifica­tions of humans: warm-blooded and cold-blooded. As far as we know, there are no cold-blooded humans. Likewise, as far as we know, there are no imma­terial entities. The existence of cold-blooded humans is much more congruent with our experience than is the existence of immaterial entities.

Craig says that if the material realm has a cause, it must be imma­terial. As impressive as that sounds, it’s a trite state­ment. All it means is this: if both ma­terial and immaterial causes exist, and if we filter out all ma­terial causes, we’re left with only immaterial causes. That tells us nothing of theologi­cal significance. It’s logical legerde­main.

Returning to Craig’s main argument, premise 2 says, “If the uni­verse has an explanation of its existence, that explanation is God.” Note that Craig has substituted the term “God” for “imma­terial cause.” When challenged on the legitimacy of this substitution, Craig shrugs that these two terms are equiva­lent.

Then why make the substitution? Let’s get real. Craig is pulling a fast one. We all know that God is taken by most people in Craig’s audience to be a con­scious being, whereas “immaterial cause,” to the extent that it has meaning, doesn’t imply any such thing. An im­material cause might be transient or imperma­nent. Craig smug­gles in a portrait of Divinity by using the heavily freighted term God.[ii]

As noted earlier, all evidence for premise 1 consists of mate­rial causes. Craig says premise 1 refers to efficient causes, a concept introduced by Aristotle. We might say, for instance, that the efficient cause of a painting is the painter. The efficient cause of the painter’s sunburn was a defect in her sunscreen.

It may sound as though “efficient” cause is just another name for “material” cause. That’s not always the case. One might say, for instance, that a child’s guardian angel was the efficient cause of the child’s stepping onto the sidewalk just in time to avoid a speed­ing car. The term effi­cient cause is broad enough to encompass both material and immaterial causes.

Premise 1’s being about efficient causes raises problems for Craig’s argument. Whatever credibility premise 1 has is owed strictly to our experience of material causes. Craig him­self, in defense of premise 1, provides examples only of material causes, never of im­material causes. We can’t infer immaterial causes from having observed only material causes. It makes no sense to cry out, “Ooh, I found a material cause! And another! And another! Since I found this abundance of material causes, there logi­cally must be an immaterial cause!”

Craig, after relying on material causes to establish premise 1, switches to immaterial causes in premise 2, without alerting his audi­ence that he’s no longer speaking of material causes. Most audience members probably never notice Craig’s guileful shift from material to imma­terial causes. Critics accuse Craig of equivocat­ing.

Equivocating is a major no-no in philosophical circles. The rule against equivocation pro­hibits speak­ers from trick­ing listeners by surreptitiously switching between alternate meanings of a word that has multiple meanings.

For example, suppose I said to you, “Yesterday I saw a huge boa and took a fancy to it. Today I bought a boa.”

I might be talking about a snake. I might be talking about a fluffy pink stole made of feathers. Which­ever I’m talking about, I should try to be clear. Right?

But what if I snickered and told you that I meant snake in the first sentence and stole in the second? You’d probably think I was being purposely deceitful.

Craig de­nies[iii] equivocating between material and immaterial caus­es, saying that he meant efficient causes all along. If we accept that defense, and I do, then Craig isn’t technically guilty of equivocating.

Still, using a word in a context where the audience likely won’t rec­ognize this switching back and forth between mean­ings is effectu­ally, even if not technically, equiv­ocating. Craig’s approach, if prac­ticed by a door-to-door salesman, would be classi­fied by the legal profession as a bait and switch scam.

Even if we let Craig off on a technicality, he’s not yet in the clear. Returning to the boa example, suppose you complained that I misled you about whether I was talking about a snake or stole. I could mim­ic Craig’s defense and say that I was talking about “effi­cient boas,” which en­compasses both snakes and stoles.

Using the term “efficient boa,” I could argue as follows: We have seen efficient boas (by which I mean snakes) within the park (a snake in the meadow, a snake by the stream, and a snake in a tree); therefore, an efficient boa (by which I mean a stole) exists outside the park.

Notice that the weakness of this argument would be less apparent if I strike all references to efficient boas, snakes, and stoles and use only the word boas: We have seen boas within the park; therefore, boas exist outside the park.

This rewording of the argument doesn’t make it better, just more cunning. Craig engages in pre­cisely this sort of wordplay.

Craig’s argument not only exploits deceptive wordplay, but it also incorporates fallacious logic. The conclusion of his argument (state­ment 5) is that imma­terial causation (God) exists. Once we under­stand that premise 1 refers to efficient causes, it’s obvious that prem­ise 1 presupposes immaterial causation. The argument’s conclusion is therefore embedded in one of its premises. As discussed earlier, that’s an informal fallacy known as beg­ging the question. Sometimes it’s called petitio principii or circular reasoning.

Craig defends himself from the charge of circular reasoning by saying that all deductive arguments are circular, “In a deductive argument, the conclusion is implicit in the premises.”[iv] Craig’s con­tention that all deductive arguments are circular is false. Otherwise, all deductive arguments would be fallacious. His state­ment that a de­ductive conclusion is implicit in the premises is irrele­vant. In a valid de­ductive argument, the conclu­sion is de­rived by com­bining the logic of the various premises, whereas Craig’s con­clusion (immate­rial causation exists) is direct­ly en­com­passed by the very term effi­cient cause. Premise 1 flat-out stipu­lates his conclusion.

Craig is mired in a catch-22 predicament. His only defense from the charge of cir­cularity is to insist that a plain reading of premise 1 doesn’t stipulate im­material causation or efficient causes. But if that is so, then Craig is guilty of equivocation.

Craig has offered us a fallacious rendi­tion of the cosmological ar­gu­ment. Unfortunately, no apologist has offered a better rendition.[v]


[i] Craig often describes his premises as “more probably true than false,” and then says that, this being the case, we should embrace his conclusion. It doesn’t follow. Suppose an argument has three premises, each of which we judge to be true with 51 percent confidence—more probably true than false. The probability of the conclusion (the conjunction of all three premises being true) equals .51 X .51 X .51 = 13.2651 or roughly 13 percent. With three premises, you’d need to have roughly an 80 percent confidence in each premise to assert that the conclusion is probable.

[ii] Craig insists that he has other arguments that prove that the immaterial cause mentioned in premise 2 possesses godlike qualities and a personal nature. Even if that were so, that conclusion (being derived from external arguments) does not belong in a premise of his contingency argument. His substitution of the word God for immaterial renders the form of his argument invalid.

[iii] William Lane Craig, “Objections So Bad I Couldn’t Have Made Them Up (Worst Objections to Kalām Cosmological Argument)”, (Posted 2/2/2012), Note: Craig is speaking here of the Kalām argument rather than contingency argument, but the objections and defenses largely overlap.

[iv] Ibid

[v] Here’s an imaginary dialog about the contingency argument. The participants include a believer (Bob) and an atheist (Alice). Bob: “Did that oak tree have to exist?” Alice: “No. The acorn might have been eaten by a squirrel. Or there could have been a drought that killed off the sapling.” Bob: “And Earth itself is contingent, right?” Alice: “I imagine so.” Bob: “But we can’t trace backwards forever, always saying that everything is contingent. There must have been something that started this whole causal sequence. That something must not have been a contingent thing, but a necessary thing. We can call that necessary thing God.” Alice: “Wait. You’re saying a necessary God had to create the universe?” Bob: “Correct, except that God technically didn’t HAVE to create the universe. He has free will. He chose to create the universe.” Alice: “Okay, so God, a necessary being, chose to create the universe, which is contingent?” Bob: “Precisely. So now you believe in God, right?” Alice: “Not so fast! You said that God has free will and that his decision to create the universe was therefore contingent.” Bob: “Yes.” Alice: “Was it contingent on something God saw somewhere or dreamed about? What led him to his decision to create the universe? There must have been something that triggered his decision, or else his decision was not a contingent thing.” Bob: “Well, there was no environment to trigger him, since he hadn’t created anything yet. It had to be something within his mind.” Alice: “Did he create his own mind?” Bob: “No. God has always existed.” Alice: “Then there was something about the nature of God’s mind that caused his decision, right?” Bob: “Right. It was God’s nature, just the way his mind works, that led to his decision to create the universe.” Alice: “Let me get this straight. The universe was contingent on God’s decision to create the universe. God’s decision to create the universe was contingent on God’s mental functioning, which was contingent on God’s nature, which could be no other way than the way it is. It seems to me that if there are no degrees of freedom in God’s nature, then everything down the causal stream is strictly determined. It couldn’t have been any other way.” Bob: “That can’t be right, though.” Alice: “Why not? The only way out of this conclusion is for you to abandon your assertion that God’s nature is necessary. You would have to admit that his nature COULD have been otherwise. But if you do that, if you say that God’s nature is contingent on something else, then God is a contingent entity. And your whole contingency argument collapses.” Bob: “Jesus loves you, Alice, but he’s probably getting pretty fed up with you right about now.”


Leave a Reply

Your email address will not be published. Required fields are marked *