Introduction to the Kalam Cosmological Argument

[NB: I do have some reservations about this argument. This is not supposed to represent an endorsement of the argument, but merely serves as an introduction to it].

Introduction to the Kalam Cosmological Argument

This is a particular form of the cosmological argument developed in a modern form primarily by William Lane Craig. Though cosmological arguments in general have been one of the traditional groups of arguments for theism, the kalam cosmological argument has not enjoyed as much attention until Craig’s more recent exposition. It has now become an extremely popular argument both in apologetics as well as Philosophy of Religion. As Quentin Smith puts it, “[m]ore articles have been published about Craig’s defense of the kalam argument than have been published about any other philosopher’s contemporary formulation of an argument for God’s existence”. Though it has roots in ancient (Aristotle) and medieval (Aquinas, al-Ghazali) philosophy, modern interpretation and proposition of the argument has seen Craig’s work as formative, and we will primarily consider his line of argument here.

The argument begins with the syllogism:

1. Everything that begins to exist has a cause.
2. The universe began to exist.
3. Therefore, the universe has a cause.

A criticism typically advanced is that this doesn’t show that God exists – only that the universe has a cause. This is an unfortunate consequence of the fact that most debates involving the argument tend to spend most time discussing this part of the overall argument, and this sub-argument is often seen as constituting the kalam argument in its entirety. The reality is that most proponents of kalam recognise that this part of the argument has relatively limited conclusions, and give further arguments for why the cause might have further attributes of God.

Another objection to cosmological arguments is that they do not explain why, if the universe needs a cause, God does not also need a cause. Putting the argument this way will make it clear why this is so, but yet the objection is occasionally still made. God, on this argument, would not need a cause because only things that begin to exist have a cause – this does not include God. “So how do you know the universe began to exist?” We will answer this question imminently.

Everything that begins to exist has a cause

Craig holds this to be evidently true, and more plausibly true than its negation. Indeed, its negation would be, in the words of philosopher of science Bernulf Kanitscheider, in “head-on collision with the most successful ontological commitment” in the history of science, namely, the metaphysical principle that, out of nothing, nothing comes. This principle of ex nihilo nihil fit is the primary reason Craig gives for accepting premise 1: “To suggest that things could just pop into being uncaused out of nothing is to quite doing serious metaphysics and to resort to magic. Nobody sincerely believes that things, say, a horse or an Eskimo village, can just pop into being without a cause. But if we make the universe an exception to [this principle], we have got to think that the whole universe just appeared at some point in the past for no reason whatsoever.” With an abundance of evidence vindicating this metaphysical axiom, then, and no known evidence to the contrary, it seems reasonable to accept this premise as being probably true.

Moreover, critics do not give good reasons why universes would constitute an exception to this principle. It is not enough to say that the Causal Principle applies only to things inside the universe, since it is a metaphysical rather than a contingent, physical principle. Nothingness cannot be discriminatory, according to Craig, since it has no properties. Why should nothingness have the capacity to birth universes, yet not have the capacity to give rise to any of its constituent parts? Indeed, if we accept the idea that the universe can exist uncaused, we have also put an end to the main competing (or complementary) search for universal origins: astrophysical cosmology. Objections that things at a first moment of time need no explaining are similarly unconvincing, since they rarely give any reason why there is a relevant difference in the causal question between first moments of time and embedded moments of time.

Finally, there is overwhelming experiential confirmation of this principle – “[s]cientific naturalists thus have the strongest of motivations to accept it”. Even JL Mackie conceded that it is “constantly confirmed in our experience”, though he thought that creation by a god was even less plausible than the Causal Principle’s negation. And, similarly, Mackie’s refutation of the cosmological argument made an appeal to Hume, who himself wrote that he “never asserted so absurd a Proposition as that anything might arise without a cause”.

The Universe began to exist

Both philosophical and scientific arguments have been offered for this premise, the former related to the existence of actual infinites and the latter pertaining to cosmological evidence for a beginning point of the universe, and of time.

In philosophy, the argument might run as follows:

1. An actual infinite cannot exist.
2. An infinite temporal regress of events is an actual infinite.
3. Therefore, an infinite temporal regress of events cannot exist.

The first premise here follows Aristotle in affirming that no actually infinite magnitude can exist. At this point, many might offer putative examples of infinites – one could divide up any finite length of time into an infinite number of sections, and thus have an infinite number of existing things. But here the distinction between actual and potential infinites – a distinction found in Aristotle but taken up by the great Cantor later on – becomes important. Indeed, it is telling that it is a finite length of time which is here divided into an infinite number of sections – Aristotle would agree that something might be infinitely divisible, but this is far from the actual infinite which would be constituted by a series of past events. Most people will only be familiar with potential infinites, denoted by ∞. These are the kind found when, for example, curves on a graph “tend to infinity”, but the reader will note that the curves never actually reach some kind of “infinity”. Actual infinites, in contrast, are denoted by א (the Hebrew letter aleph) and are widely held to be limited to transfinite arithmetic and similar domains – and not present in reality.

Craig usually presents some examples for why the idea of an actual infinite is absurd – the foremost being the hotel of the famous mathematician David Hilbert where all א of the rooms are occupied, and yet there is still room for more guests to stay in new rooms! Such a hotel, according to Craig, would lead to any number of absurdities: “suppose that the persons in rooms #4, 5, 6, … checked out. At a single stroke the hotel would be virtually emptied, the guest register reduced to three names, and the infinite converted to finitude. And yet it would remain true that as many guests checked out this time as when the guests in rooms #1, 3, 5, … checked out! Can anyone believe that such a hotel could exist in reality?”

Craig offers a second philosophical argument, namely that P1) a collection formed by successive addition cannot be an actual infinite, P2) the temporal series of events is a collection formed by successive addition, and therefore that C) the temporal series of events cannot be an actual infinite. It is not the task of this introduction to go into depth on these arguments; much more has been written elsewhere.

All of this is confirmed, it is argued, by emerging scientific evidence that the universe (including time) did indeed have a beginning. This has, in the past, been extremely controversial: until partway through the 20th century, it was taken as axiomatic that the universe had existed eternally in the past. Take, for example, von Weizsacker’s recollection of an event in which Walther Nernst challenged von Weizsacker’s presentation on the age of the universe:

“He said, the view that there might be an age of the universe was not science. At first I did not understand him. He explained that the infinite duration of time was a basic element of all scientific thought, and to deny this would mean to betray the very foundations of science. I was quite surprised by this idea and I ventured the objection that it was scientific to form hypotheses according to the hints given by experience, and that the idea of an age of the universe was such a hypothesis. He retorted that we could not form a scientific hypothesis which contradicted the very foundations of science. He was just angry, and thus the discussion, which was continued in his private library, could not lead to any result.”

As with the Copernican revolution, then, this traditional cosmology required a Kuhnian shift when increasing amounts of evidence emerged that supported the universe having a beginning. This led renowned physicist Stephen Hawking to state: “All the evidence seems to indicate, that the universe has not existed forever, but that it had a beginning, about 15 billion years ago. This is probably the most remarkable discovery of modern cosmology. Yet it is now taken for granted … [T]he universe has not existed forever. Rather, the universe, and time itself, had a beginning in the Big Bang, about 15 billion years ago.”

Craig cites the work of Arvind Borde, Alan Guth and Alexander Vilenkin, who demonstrated that “any universe which is, on average, in a state of cosmic expansion throughout its history, cannot be infinite in the past, but must have a past space-time boundary”. And elsewhere, Craig and Sinclair write, “[o]ur survey shows that contemporary cosmology is quite supportive of the second premise of the kalam cosmological argument. Further, this conclusion is not reached through ferreting out elaborate and unique failure conditions for scores of individual models. Rather, the repeated application of simple principles seems effective in ruling out a beginningless model … It seems that the field of cosmology, therefore, yields good evidence that the universe began to exist.” As with the philosophical arguments, there are volumes that could be written, but that is not the purpose of this introduction.

With good support for both the first two premises, then, that everything which begins to exist has a cause and that the universe has a cause, it follows necessarily that the universe has a cause.

What must the cause be like?

It is often assumed that the argument finishes at this point, with the rather unsatisfactory (yet still controversial) conclusion that the universe has a cause. “But that doesn’t prove that God exists!” complains the critic. Indeed, this first part of the argument was never intended to demonstrate that God exists – but there has also been a great deal of literature dealing with the question of what properties a cause of the universe must have.

Uncaused: This is not strictly implied by the argument, for one could posit a plurality of explanatory steps – the universe might be caused by X, which in turn is caused by Y, which in turn is caused by Z, and so on. But this seems to violate the principle of parsimony expressed by Ockham’s Razor – we ought not to posit causes beyond necessity. Since our primary cause of the universe would not begin to exist, it would not need a cause on the grounds of the arguments given above, and so it is simplest to suggest that it is, itself, uncaused. This is especially so if the arguments from infinity work, and are applied to an infinite number of causal entities.

Beginningless: If the first premise of the kalam argument is true, that everything that begins to exist has a cause, it follows that nothing which is uncaused can begin to exist. If, therefore, we accept the parsimony of our cause of the universe being uncaused, it would follow from this, along with premise 1 of the original argument, that this cause does not begin to exist.

Changeless, immaterial and atemporal: Changelessness is different from immutability in that the latter affirms that the cause cannot change, whereas we are affirming only that, sans the universe, the cause does not change. The cause’s changelessness, immateriality and atemporality are all closely linked, with each tending to imply the others. For example, atemporality implies changelessness (since change requires time), and changelessness in turn implies immateriality, since matter is constantly changing at the micro-level. But yet both the cause’s immateriality and its atemporality is implied by its causing space and time (the universe) and therefore transcending it. Similarly, the cause’s spacelessness is implied by these latter two, in that no spatial entity can be both immaterial and atemporal.

Powerful: It follows also that this cause must be incredibly powerful, “since it brought the entirety of physical reality, including all matter and energy and space-time itself, into being without any material cause.” This is not necessarily omnipotent, but it is still extremely powerful.

Finally, Craig offers three main lines of argument for why the cause is plausibly, but not necessarily, personal. First, he draws on Richard Swinburne’s distinction between personal and scientific explanation, where the former explains something in terms of agents and volitions, and the latter explains something in terms of laws acting on initial conditions. Since there is no time before the universe, it cannot be explained in terms of laws operating on initial conditions, and therefore it can only be accounted for by personal explanation.

The second argument relies on the characteristics the cause has already been demonstrated to have: there are, Craig argues, only two categories of things which are potentially immaterial, beginningless, uncaused, timeless, and spaceless. These are, on the one hand, abstract objects, and on the other, unembodied minds. This is not to say that humans have unembodied minds, or even that human minds are not dependent on the brain, but rather it is to admit the possibility of such an entity. But since abstract objects cannot cause anything, it follows that the only plausible cause is something similar to an unembodied mind.

The third reason for agent-causation is that a first temporal effect from a changeless cause can only arise from free, personal explanation. If the necessary and sufficient conditions existed changelessly and eternally, it is hard to believe that the effect would not also exist changelessly. But clearly the origin of the universe is not eternal and changeless, so how can this be? Craig answers, “The best way out of this dilemma is agent causation, whereby the agent freely brings about some event some event in the absence of prior determining conditions.” His conclusion? “On the basis of the kalam cosmological argument, it is therefore plausible that an uncaused, personal Creator of the universe exists, who sans the universe is beginningless, changeless, immaterial, timeless, spaceless, and enormously powerful.”

Is the argument useful?

This is one of the primary points of doubt when considering the kalam cosmological argument. Even if the argument succeeds in demonstrating all this, the objector says, it ultimately has no value, since it doesn’t show that God exists – it has nothing to say about God’s moral character, and especially about particular Gods. This type of objection is exemplified by Stephen Law’s “Evil God” hypothesis, where there exists an uncaused, personal Creator of the universe, who sans the universe is beginningless, changeless, immaterial, timeless, spaceless and enormously powerful. The only difference is that this God is exceedingly malevolent. On the kalam argument, Law contends, this Evil God is just as probable as a good God, and therefore the argument does not demonstrate that a good God exists, and does not even form part of a cumulative case for such a God.

There are different ways to respond to this, the first being to respond in agreement with the objection, but without conceding that this weakens the argument. Perhaps the argument isn’t intended to prove a good God, only to serve as a defeater for atheism. As Craig put it in his debate with Law, “it is a strange form of atheism, one not worth the name, that admits that there is a beginningless, uncaused, spaceless, timeless, immaterial, enormously powerful, personal Creator of the universe.”

But Law’s claim that the kalam cosmological argument cannot form part of a cumulative case for God, on the grounds that it also supports the Evil God hypothesis is palpably false. Suppose we were investigating a murder and found a gun at the crime scene, from which a DNA sample was obtained and traced back to Thomas. We would all reasonably count this as quite substantive evidence that Thomas was the killer. Even if it turned out that Thomas had an identical twin, John, would we really count the evidence worthless, and maintain that it did not support the case against Thomas at all? Of course, it wouldn’t be conclusive evidence, but it is most unreasonable to suppose, just because the evidence pointed also towards John, that it does not constitute substantive evidence towards the culpability of Thomas. After all, it seems to rule out a great number of other suspects, and unless the DNA evidence distinguishes between Thomas and John, we may reasonably suppose that it increases the probability of their culpability proportionally:

And so it is difficult to see why the existence of some other hypothesis, for example, Evil God, should reduce the kalam argument’s inductive force. Swinburne puts it well:

Note that it is no objection to a P-inductive or C-inductive argument from e to h that some contrary hypothesis h* is also compatible with e, as some writers on the philosophy of religion seem to think. They seem to think that if, for example, the order in the universe is compatible with ‘God does not exist’, then there is no good argument from it to ‘God exists’. But one has only to think about the matter to realize that this is not so. In any non-deductive argument from e to h, not-h will be compatible with e; and yet some non-deductive arguments are good arguments.

But what of Law’s argument? For Law is not simply saying that Evil God is compatible with the existence of the universe but that, if the kalam argument is sound, then the argument supports Evil God as much as it does a good God. But Swinburne goes on:

Note also a further interesting feature of good C-inductive arguments. In such an argument from e to h, P(h|e&k) > P(h|k). It may be the case that also for some contrary hypothesis h* there is a good C-inductive argument from e– that is, also P(h*|e&k) > P(h*|k). The fact that certain evidence confirms a hypothesis does not mean that it does not also confirm a rival hypothesis. Once again, this should be immediately clear if one thinks about it.

And so we might represent the kalam argument similarly to our example with Thomas and John – unless kalam offers us reason to differentiate between a good God and Evil God, then it seems reasonable that they should fill up the probability space left by the exclusion of other hypothesis in a manner proportional to their prior probabilities. And, since Law’s entire objection is to the effect that kalam does not let us differentiate between the two, that is what we will do:

Of course, there are some other hypotheses compatible with the kalam argument, but the point is clear: the success of the kalam argument is perfectly compatible with both God and Evil God being more probable in retrospect. When seen this way, it is very strange that people should deny there is any cumulative force in the argument, so long as it works to its conclusion.

And finally, note that this can be applied also to the first steps of the argument, to the conclusion that the universe has a cause. Since this excludes the possibility of an uncaused universe, it seems reasonable that it should raise the probability of theism as an explanatory hypothesis accordingly, even if it simultaneously raises the probability of other explanatory hypotheses.

As hinted by Swinburne, this can be more clearly seen by a Bayesian analysis. Since this is only an introduction, I have tried to stay away from introducing Bayesian probability analysis, but it is worth noting that it would be a most fruitful way to demonstrate the probabilistic force of kalam, given that one is familiar with Bayesian theory.


Personally, I am not convinced of all parts of the kalam argument. I am convinced that it has some cumulative force as per the last section, and I think it is more than reasonable to conclude that the universe has a cause. The move to a God-like cause, though not as illogical as many would assume, is not yet entirely convincing to me, but I hope to have provided a basic introduction to and defence of the key aspects of the argument.

Further reading

For more, see Craig, W.L. & Sinclair J.D. (2009) The kalam cosmological argument. In Craig, W.L. & Moreland, J.P. (eds.), The Blackwell Companion to Natural Theology, 101-201. Oxford: Blackwell.

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 )

Google photo

You are commenting using your Google 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