The Kalam Cosmological Argument and The Problem of Induction

By R.N. Carmona

Though the Kalam Cosmological Argument (KCA) is a deductive argument, it imports inductive reasoning. Since there’s a distinction between argument and reasoning, it should apply to, at least, some deductive arguments.1 To see where this occurs, it is necessary to restate P1 of the KCA: Whatever begins to exist has a cause.2

Before demonstrating how P1 imports inductive reasoning, a distinction must be made between the common application of the Problem of Induction and what it is it actually entails. The common application of the problem focuses on the notion that given the uniformity of nature, the future will be like the past. For Hume, the uniformity of nature is used to justify induction. It is, however, important to note that Hume never explicitly mentioned induction. The Problem of Induction is derived from Hume’s discussion on causation.3 This notion is at the center of the common application of the Problem of Induction. It does, however, have broader application.

The problem should focus on beliefs concerning matters of fact that extend beyond experience and observation. These beliefs could be about past, unobserved present, and future events, objects, or phenomena.4 This is key to seeing how the first premise of the KCA imports inductive reasoning.

Disregarding discussions on the nature of time, i.e. whether time corresponds with A-theory or B-theory, the past and the future, intuitively speaking, have something in common. In science, predictions are often made about the future and the past; the latter is increasingly being referred to as a retrodiction.5 The reason predictions are necessary about the past and the future is because they are out of the reach of our experience and observational apparatus. Since we cannot directly observe or experience the past and the future, inductive inferences are necessary to draw conclusions about both.

When one argues that, whatever begins to exist has a cause, one is employing inductive reasoning. During one’s lifetime, one has noticed that every effect is preceded by a cause. However, in order to argue that this has always been the case, per Hume, one has to assume the uniformity of nature. It follows that one is making an inductive inference about the past.

From an epistemic standpoint, the Problem of Induction is also the problem of transfer of belief in the evidence. Though it is true that people believe things that lack evidence, conclusions are often supported by evidence. So the problem occurs in the bond between a conclusion and the evidence said to support it–specifically in the transfer from belief in the evidence to belief in a given conclusion. This transfer of belief is what Hume sought to address. He didn’t want to call into question the belief itself, but rather, the grounds of said belief.6

Much of this shouldn’t be news for the Christian or theist. When addressing proponents of scientism or when attempting to downplay the superiority of scientific over religious reasoning, they are both fond of invoking the Problem of Induction.7 Unfortunately, they fail to realize that it’s imported into an argument they’re also quite fond of. If it’s an actual problem, then the implications of the problem apply to the KCA. In other words, if the Problem of Induction cites a lack of justification for beliefs about the past that are out of the range of experience and observation, the Christian and theist aren’t justified in saying that whatever begins to exist has a cause. Since the premise lacks justification, there’s no way to establish its soundness. Therefore, given P2’s connection to P1, we cannot establish the soundness of P2. It follows that we cannot establish the soundness of the argument as a whole.

In the past, I’ve been pointedly critical of the KCA. I focused, for instance, on the concept of causation.8 I also located a common fallacy within the argument.9 One would be hard pressed to establish the soundness of the argument even if P1 didn’t feature inductive reasoning, which the Christian and theist admit has problematic implications. With that said, this isn’t so much a rebuttal of the argument, but rather, another spotlight on the double think of theists, Christians in particular.

It seems as though Christians don’t care to correct the inconsistency within their arguments. They can’t invoke the Problem of Induction whenever it suits them only to discard it whenever it doesn’t. If one is intellectually honest, consistency is required.  If they admit that there’s a Problem of Induction and that it is still without sufficient reply, as apologists have argued, then the implications of the problem are also imported into P1 of the KCA. I find that this conclusion is inescapable.

Works Cited

1 Will, Frederick L.. Is There a Problem of Induction? The Journal of Philosophy, Vol. 39, No. 19, pp. 505-513. 1942. Print. 

2 Craig, William L. “In Defense of the Kalam Cosmological Argument”Reasonable Faith.

3 Burns, Samuel R.. The Problem of Deduction: Hume’s Problem Expanded. University of Michigan, Ann Arbor. 2009. Print.

4 Fritz Jr., Charles A.. What is Induction? The Journal of Philosophy, Vol. 57, No. 4, pp. 126-138. 1960. Print. 

5 Barrett, Martin, Sober, Elliott. Is Entropy Relevant to the Asymmetry Between Prediction and Retrodiction? (1992); Steinitz, Yuval. Prediction versus Retrodiction in MIll (1994); Begun, David R. Human Evolution: Retrodictions and Predictions (2005); Christopher J. Ellison, John R. Mahoney, James P. Crutchfield. Prediction, Retrodiction, and The Amount of Information Stored in the Present (2009).

6 Oliver, Donald W.. A Re-examination of the Problem of Induction. The Journal of Philosophy, Vol. 49, No. 25, pp. 769-780. 1952. Print. 

7 See “Does scientific knowledge presuppose God? A reply to Carroll, Coyne, Dawkins and Loftus” by Vincent Torley

8 See “Dispotions, Causality, and the Kalam Cosmological Argument”

9 See “A Simple Rebuttal: Kalam Cosmological Argument”


  1. Steven Jake

    I’d have to disagree here. The principle of causality is not an induction based statement but is, rather, a necessary condition of change at all, and is arrived at a priori. Here’s why I claim this. If something comes into being, then there are only two options left over if one is to deny the principle of causality: 1) the substance caused itself to come into being, or 2) the substance came into being from nothing. 1) would seem to be false because in order for a thing to be self-caused it would have to exist before it existed in order to cause itself to exist, which is logically incoherent. 2) is false because nothing cannot be a cause of anything because nothing cannot do anything, since it is, by definition, the absence of anything.

    The point here is that none of this depends on induction. Rather, it is deduced a priori that the principle of causality is, not only true, but necessarily true.


    • R.N. Carmona

      The difference is between Hume’s view of causality and Kant’s. On Hume’s view, causality is arrived at by induction; more specifically, it’s a habitual and customary connection we make. Also, there’s no manner in which we can prove causal connections. On Kant’s view, which is the view you allude to, causation is an a priori judgment. It is an object of pure reason. I can get much more specific, but there’s no need to.

      All that needs to be said is that it isn’t conclusive that Kant’s response to Hume succeeds. Furthermore, there’s no consensus as to whose account is better or whether they actually disagree. Kant sounded quite certain when responding to Hume; however, he would later be perplexed by causation and expressed, at the very least, caution.

      In any case, your examples fail because the universe isn’t a substance, whatever is meant by that.* The universe also isn’t a mere object. The KCA applies the so called principle of causality to the universe. Interestingly, there are cosmological models for a self-caused or alternatively, a causally closed universe; in fact, most modern cosmological models are of that nature. There are also philosophical arguments for a self-caused and/or self-organizing universe (see Quentin Smith, for example).

      Ultimately, Kant and Hume agree that causal connections cannot be proved. This is peculiar because necessary and universal judgments are self-evident. If causality was as self-evident, Kant would not have been perplexed. He would not have, in other words, sided with caution in his later writings. He would have continued to express the same level of certainty. Also worth exploring is whether accepting Kant’s definition of causality commits one to accepting his philosophy: objects of pure reason, phenomena and noumena, and so on. I think that it might suggest this, but this is tangential.

      At any rate, Kant’s definition is not as conclusive as perhaps you hoped. Furthermore, this was a smaller criticism of the argument–one that attempts to show a contradiction that exists in the minds of, at least, some espousers of the KCA. More forcefully, even if causality fits Kant’s definition better than it does Hume’s, there’s still the fact that the universe is unlike the objects we speak about in relation to causality. Causality, as normally construed, simply does not apply to the universe. We can also point out flaws in the underlying Aristotelian metaphysics, but again, that would bog us down in a tangential discussion that isn’t required to debase the KCA. There are just too many issues with the argument, not least of which is its incompatibility with modern cosmology.

      *The term substance has been used differently by a number of philosophers. There is no consensus on what that word means or whether it even has an ontological place in the universe, i.e., whether there’s such a thing we can identify as substance.


    • The Thinker

      Steven, your response for one thing relies on the A-theory of time since you assume things “come into being.” You therefore cannot arrive at it a priori because you’re making some assumptions about the nature of time that have no a priori basis. Secondly, your (2) is not correct. The thing wouldn’t have come into being from nothing, it would just come into being without a cause. There were things that existed prior to it, just nothing that caused it. And your response to (2) is not a logical fact, it is itself based on inductive reasoning.

      Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s