The Kalam Cosmological Argument is an argument for the existence of God that Craig developed during his doctoral work in the 1970s. While it can be formulated in various ways, here I will consider the following version:
P1. The universe began to exist.
P2. If the universe began to exist, it had a cause.
P3. If the universe had a cause, that cause must be a personal agent.
C. A personal agent caused the universe to begin to exist.
There is a simpler form of the Kalam which ends at premise two, and hence concludes merely that the universe had a cause. However since Craig uses the Kalam to demonstrate the existence of God, I believe it is important to make this aspect of the argument explicit by including premise three.
The Kalam has attracted a great deal of attention in recent years, both in popular debates and also in the philosophical literature. However, in my view the argument is fatally flawed. In this article I shall briefly outline five major problems with Craig’s defence of the Kalam, showing how he fails to establish the conclusion that the universe had a personal cause. Readers interested a more detailed discussion of these issues should consult my book, Unreasonable Faith.
The problem of the extent of the present
Craig has stated that when he talks of the universe ‘beginning to exist’, he means this in a very sense. Specifically, he means that the universe came into being from a state of nonbeing. As Craig has said many times, this depends on a particular understanding of the nature of time called ‘presentism’. According to this view, only the present exists; the past and the future do not exist. The past consists of a series of moments which came into being and then went out of being, while the future consists of a series of moments which will come into being, but haven’t yet done so. This view of time contrasts with a competing view known as the tenseless theory of time. This view of time holds that the past, present, and future all exist alongside each other, just at different temporal parts of the history of the universe. In this view, there is no objective ‘present’. Instead, the past is just the part of the universe that exists in the ‘before than’ direction, while the future is the part of the universe that exists in the ‘later than’ direction. Under the tenseless theory of time, a universe with a finite age would not have come into being from nonbeing, but would simply be extended a finite time in the ‘earlier than’ direction. This means that under a tenseless theory of time, the universe could not have ‘begun to exist’ in the way that Craig believes is necessary for the Kalam to succeed. Because of this, Craig has stated definitely that “The kalam cosmological argument presupposes from start to finish a theory, not of tenseless time, but of tensed time”.
According to Craig, presentism, the view that only the present exists, is vital for the Kalam to succeed. But this leads naturally to the question: what exactly counts as present? The problem is that Craig has no sensible answer to this question. One possibility is that the present could consist of an infinitesimally short instant, with an infinite collection of such instants forming a finite interval (just like the real numbers). But this won’t work, because it would imply that actual infinities could exist, which conflicts with Craig’s various philosophical arguments against the possibility of an infinitely old universe. A second possibility is that the present could consist of a discrete unit of time, with successive discrete presents following one after the other with no time in-between (just like the natural numbers). But this won’t work either, as it would mean that God was somehow restricted to act only at these discrete moments, but couldn’t act in between them. If neither of these possibilities is suitable, what is left? Craig’s answer is that the present consists of the present ‘non-metrical interval’, the length of which varies depending on the context. He describes this as follows: “the extent of the present depends upon the extent of the entity described as present… the duration stipulated to be present will be an arbitrary, finite duration centred on a conceptually specified instant”. Craig even says that this interval can be further divided into present, past, and future parts. Although creative, this solution is completely absurd; indeed it is essentially unintelligible. How can it be the case that we bring into being a period of time just by referring to it? It seems that according to Craig’s view, dinosaurs were brought back to life when humans discovered them and began referring to them! Also, if as Craig claims only the present exists, then how can there be parts of the present interval that are past and parts that are future, neither of which exist? How can something both exist and not exist? Craig must resolve this problem of the extent of the present if the Kalam has any chance of succeeding.
The problem of the relevance of thought experiments about infinity
Craig provides many philosophical arguments for why the universe cannot be infinitely old. These arguments generally consist of thought experiments which aim to provide illustrations of absurd or impossible scenarios that allegedly follow from the assumption that actual infinities can exist. Some of these thought experiments, such as Hilbert’s Hotel, aim to show that no truly infinite thing can exist in the real world. Hilbert’s Hotel is an imaginary hotel with infinitely many rooms. If such a hotel existed, then it would be possible for the hotel to start out with all its rooms occupied, then for the owner to shuffle the infinitely many guests around between rooms in such a manner as to free up an infinite number of rooms, thus allowing another infinite number of guests to check in. Craig argues that thought experiments such as this illustrate the absurdity of actual infinities, hence demonstrating that they cannot exist in reality.
Another type of thought experiment focuses specifically on infinities produced by processes of successive adding of one item to the next, since this is the type of infinite process which (so the argument goes) would be involved in giving rise to an infinitely old universe. While there are many such thought-experiments, one fairly simple example involves a man who has been counting down from infinity, and at the present moment finally finishes and reaches zero. According to Craig, such a scenario is impossible because the man should have finished yesterday, since yesterday he already had an infinite amount of time to count. However, he already had an infinite amount of time even before yesterday, and so he should have actually finished the day before yesterday – on so on backwards in time. Thus, according to Craig, we cannot make any sense of an infinite count arriving at the present at any particular time. From this Craig concludes that we should reject the possibility of arriving at the present after an infinitely long count.
While one can argue the details of each particular thought experiment, there is a more fundamental problem with such reasoning. Any specific example of an absurd or impossible situation can only ever show that the particular case in question is absurd or impossible. For example, Hilbert’s Hotel shows at most that infinitely large hotels with moveable guests are absurd. Likewise, the counting down from infinity example shows that it is not possible to begin with an infinitely large set and subtract one element at a time to eventually reach zero. The crucial question, however, is why we should infer from such cases that the universe cannot be infinitely old? The universe is not an infinite hotel, nor is there some cosmic counter marking down the time from infinity past to the present. Other thought experiments which readers may be familiar with, such as the Grim Reaper Paradox and the Tristam Shandy Paradox, similarly involve bizarre situations (such as infinitely many grim reapers appearing, or a man writing an infinitely long diary), neither of which are entailed by an infinitely old universe. As such, the most these thought experiments can ever show is that certain types of infinities are absurd or impossible, not that the universe must have begun to exist. Craig has argued that the relevant aspect of all such cases is the notion of infinity rather than any other details, but this can always be disputed. For instance, Hilbert’s Hotel assumes that infinitely many guests can be moved around, while past events cannot be moved around. The man counting down from infinitely had some starting point that he began counting from (since this is how counting works), and yet an infinitely old universe would have no such beginning point. There is thus always a further question about which part of the thought experiment renders it absurd or impossible, and Craig fails to show that this must be notion of infinity itself. This is a fundamental weakness of this type of argument: why should we infer that because certain situations involving actual infinities are impossible, that therefore the universe cannot be infinitely old?
The problem of the difference between physical and metaphysical time
To supplement his philosophical arguments, Craig also appeals to various results from contemporary cosmology to argue that the best empirical evidence supports the beginning of the universe. The result he most often appeals to is called the Borde-Vilenkin-Guth Theorem, which shows that under certain plausible assumptions, all past-directed paths through spacetime must terminate after a finite distance, and hence cannot be infinitely long. From this and other related results, Craig infers that there is strong empirical confirmation from cosmology that the universe began a finite time in the past. The problem with all such arguments is that the Borde-Vilenkin-Guth Theorem, along with all of modern cosmology, is founded upon the general theory of relativity. In general relativity, time is treated as a fourth dimension alongside the three dimensions of space, which can be bent by the existence of massive objects and is affected by relativistic effects such as time dilation. Time is treated in a tenseless manner, with all times existing alongside one another in a four-dimensional ‘spacetime’.
It is very difficult to see how this notion of time can be reconciled to Craig’s presentist philosophy of time, according to which the present is an objective feature of reality independent of the matter within the universe, and the past and future do not exist. Indeed, in his publications on the philosophy of time, Craig himself acknowledges this, arguing for example that “if we draw a distinction between metaphysical time and physical time as Newton did, it is quite evident that a beginning of the latter does not imply a beginning of the former”. He is also clear that the four-dimensional spacetime used in general relativity “serves as a convenient calculational and diagrammatical aid… but says absolutely nothing about ontology… the four-dimensional continuum should therefore be regarded as a useful tool, and not as a physical ‘reality’”. This means that Craig faces with a stark choice. If he holds firm to his presentist philosophy of time, then he must give up all the arguments (such as the Borde-Vilenkin-Guth Theorem) for the finite of the past that are based on physical cosmology. On the other hand, if he wishes to keep using these arguments, then he must give up presentism, the very philosophy of time he believes is essential for the Kalam to succeed.
The problem of establishing a cause of the universe
Craig believes that the universe must have had a cause of its coming into being. While this may seem reasonable on the face of it, on further examination it is actually very hard to see how Craig could possibly know such a thing. The most common argument Craig presents in favour of contention is simply that it is a basic metaphysical principle that “something cannot come out of nothing”. Yet this is begging the question. Someone who doubts that the universe had a cause will not simply accept that this is a basic metaphysical principle! Furthermore, Craig’s talk of ‘something coming from nothing’ is inaccurate, as this wording implies that somehow there was nothing, and then the universe ‘popped into being’, as Craig often says. Yet this clearly doesn’t make any sense. There wasn’t anything, no time or space, before the universe, so it wasn’t as if there was a black emptiness and then suddenly there was a loud ‘pop’, and the universe appeared. The idea is rather that if you could somehow go back to the first moment of time, the universe would already be there, and didn’t require anything to make it exist. It just exists all by itself, without an external cause. It appears that Craig has nothing to say about this more reasonable conception of a universe that began without a cause.
A second argument that Craig gives in support of the idea that the universe had a cause, is that we have overwhelming evidence both from science and everyday experience that when something begins to exist it has a cause. By analogy, he argues, it is therefore reasonable to believe that the universe as a whole must have had a cause. Such an inference, however, is simply not warranted. Everything we observe in science and everyday life is some object or process within the universe. The universe as a whole is not analogous to some particular thing within the universe, and so we have no reason to think that just because things within the universe have causes, that therefore the universe as a whole must also have a cause. This is similar to inferring that because every man has a mother, that therefore mankind as a whole also has a mother.
Craig’s third and final argument that the universe had a cause asks the question: if universes could come into existence without a cause, then why not anything? Craig asks “why do bicycles and Beethoven and root beer not pop into being from nothing?” Yet the answer to such questions is obvious. Bicycles and Beethoven and root beer are things that exist within the universe, and have known necessary causes for their coming into being. By contrast, the universe is not something that exists inside the universe with known necessary causes, but instead is the totally of space and time itself, for which it is unknown whether there is any necessary cause. It is therefore entirely unsurprising that things do not pop into being within the universe. This tells us nothing about whether the universe itself required a cause, since obviously the universe would not have come into being inside the universe. Craig’s arguments notwithstanding, the question therefore remains: why couldn’t the universe have simply begun without a cause?
The problem of non-personal, timeless causes
Building upon the first two premises of the Kalam, Craig also argues that if the universe had a cause, that cause must be a personal agent. Craig argues that only an immaterial mind with libertarian free will could have the ability to bring about the beginning of a temporal series of events (i.e. create the universe) from an existing timeless state. To justify this, Craig argues first that the cause of the universe obviously can’t be an event in time, because then this would just be part of the universe, and would require its own cause. He also argues that the cause of the universe could not have been an impersonal timeless state, because any effect caused by such a state would have to be co-eternal with the cause. According to Craig, only an agent with libertarian free will could exist timelessly, and then by exercising their libertarian agency, bring about an effect that exists in time. Such an explanation leads naturally to the question – what is so special about libertarian agency? Why can’t some timeless, non-personal entity be the case of the universe? Why does it have to be a personal agent? Craig’s answer is simply that libertarian agents have a special type of causal power called ‘agent causation’, and only agent causation can bring about a temporal effect from a timeless cause. Why is that the case? Craig doesn’t appear to have any clear answer for this, other than his assertion that this is just the way it is. For instance, he claims “because the agent is free, he can initiate new effects by freely bringing about conditions which were not previously present. For example, a man sitting changelessly from eternity could freely will to stand up; thus, a temporal effect arises from an eternally existing agent.” Effectively this amounts to asserting that agent causation, and only agent causation, can bring about a temporal effect from a non-temporal initial state, because agent causation is just special that way. Needless to say, I do not find this to be a persuasive argument.
Indeed, I can see no reason why any sort of timeless cause could not function as the cause of the universe. Why could the cause not be some kind of timeless quantum field, such as is postulated by various models in quantum cosmology? Or possibly it could be an immaterial entity driven by instinct, such as the gunas from Hindu philosophy. Or maybe it could be a metaphysical force such as the Dao from Chinese philosophy. Or perhaps it could be abstract information with causal powers, as scholars like Paul Davies have recently argued. Or just as likely, it could be something else entirely that we haven’t even conceived of yet! Craig relies entirely on his personal intuition that a personal agent is the only sort of thing that could bring the universe into being, without even considering these other possibilities. Indeed, it seems on the face of it highly unlikely that we humans could have any reliable method of knowing what exists outside of the universe. Yet Craig not only claims to know this, but claims to have come to this conclusion on the basis of reflection and intuition, without even needing to do any scientific research or empirical investigation. As such, I believe Craig’s argument as to why the cause of the universe must be personal is entirely unpersuasive. He seems to have no persuasive answer to the simple question: why can’t a non-personal, timeless entity be the case of the universe?
Craig’s Kalam Cosmological Argument faces major problems. These can be summarised as a series of questions. First, what is the temporal extent of the present? Craig needs to provide a cogent answer to this question before he can develop a philosophy of time necessary for the Kalam to work. Second, how can we determine that because certain situations involving actual infinities are impossible, that therefore the universe cannot be infinitely old? This difficulty in making the leap between hypothetical thought experiments and the finitude of the past is a major limitation in philosophical arguments for a beginning of the universe. Third, is the notion of ‘time’ in general relativity equivalent to ‘real’ metaphysical time? It seems that Craig must answer ‘no’ to this question if he is to maintain his presentist philosophy of time, and yet such an answer means that he cannot appeal to scientific results (which depend on general relativity) to establish the beginning of the universe. Fourth, why can’t the universe just begin without a cause? Craig’s quip that ‘from nothing, nothing comes’ is not a sufficient answer, since the real question not whether the universe could have ‘popped into being’, but whether the universe could simply have begun a finite time in the past without any external cause. Fifth, why can’t a non-personal, timeless entity be the cause of the universe? Craig asserts that only an immaterial mind with libertarian freedom can cause a temporal state to begin from an initial timeless state, but has no non-circular justification for this, and simply ignores the many other possible causes that could bring about this effect. Unless Craig is able to provide cogent and consistent answers to these five questions, his Kalam Cosmological Argument will continue to be plagued by major unresolved problems, and fail to persuade skeptics such as me that God is the best explanation for the origin of the universe.