By A. Joubert
The cosmological argument is one of the oldest and most established arguments for the existence of God, coming in a variety of forms and spanning more than two thousand years (from at least Plato’s Laws). Although there are a many different forms (Aristotelian, Platonic, Thomistic, Leibnizian, Kalam, etc.), they are united by appealing to the contingency of the universe in some way or another. In other words, they appeal to the fact that the universe is not self-sufficient. This contrasts with metaphysical necessity, where if a being is metaphysically necessary, its existence is not dependent on anything other than itself. Abstract objects and universals, for example, are metaphysically necessary. The universe is contingent, because it is not existentially independent. I want to take a look at some common objections to the cosmological argument.
In order to introduce the argument, we’ll take a look at Leibniz’s cosmological argument, which relies on the Principle of Sufficient Reason. (While this argument is only one of the many forms, it is suitable as an introduction, because it is appropriate for the objections we will address). The Principle of Sufficient Reason entails that “nothing occurs without a sufficient reason why it is so.” Consequently, there must be a sufficient reason for the universe as a whole. The sufficient reason must be necessary or self-existent (its sufficient reason can be found in itself), because no matter how far back in the series of sufficient reasons we go, the series as a whole is still contingent and therefore requires an explanation that is not itself. The entity that is the sufficient reason of the universe must itself be necessary, because if it is contingent, then it would mean that it is merely another link in the chain of sufficient reasons that make up the universe. The only way to terminate this regress of sufficient reasons is to postulate a necessary being that holds its sufficient reason in itself.
One common objection is that the universe itself is necessary or self-existent. Why should the universe itself require a cause? Why can the infinite regress of contingent causes not terminate before it has even started, with the universe? For one, Big Bang cosmology has given us good reason to think that the universe is contingent. It is difficult to make the claim that something that begins to exist, which modern science tells us happened to the universe, is metaphysically necessary. Remember, metaphysical necessity means that it is existentially independent (does not depend on anything other than itself for its existence). If the universe began to exist, then it is dependent on some type of event or condition. Secondly, the universe is physical. From observation, we know that physical things tend to be contingent. Physical things are not metaphysically self-sufficient. Everything in the universe, that constitutes it, is impermanent and depends upon other physical things for its existence. It is therefore unlikely that the universe itself, which is physically constituted, should detract from this rule. But you may retort: just because everything within the universe has a cause (and is contingent) doesn’t mean the universe as a whole has a cause. Doesn’t that reasoning commit the fallacy of composition? It is important to note, firstly, that the contrary viewpoint also commits the fallacy of composition. To say that it is sufficient to explain the parts of a thing and that one does not therefore have to explain the whole thing is also to commit the fallacy of composition. Or, in other words, if we say that explaining the parts of the universe sufficiently explains the universe itself, is also the fallacy of composition, because one is saying that as a result of the parts being explained, the whole is also explained. Secondly, it is not always logically invalid to reason from composition. For example, one can say that because every part of a book is constituted by paper, therefore the book itself is constituted by paper. In addition, as Bruce Reichenbach observes, the contingency of the parts of the universe makes an explanation of the whole of the universe still necessary: “If they are explained in terms of something else, the entire collection still remains unaccounted for.” This means that reasoning from composition in terms of the contingency of the universe is not logically invalid. The fact that everything in the universe and everything of which it is constituted is physical, means that the universe itself is also physical. The universe, to state it simplistically, is a composite object of physical things, and it will be just as contingent as any other composite object, no matter how different in size or complexity they may be.
This ties in with a very popular objection, not only to the cosmological argument, but also to the teleological argument: namely, who designed the designer? Bertrand Russell, old Mr. Dawkins (and countless others) have taken this line of argument. But it rests on a misunderstanding of the arguments. Firstly, both the cosmological argument and the teleological argument do not imply the existence of a contingent designer. The creator would only need a creator in turn if it were a contingent being (so that it is not metaphysically self-sufficient). All forms of the cosmological and teleological arguments imply a being that transcends matter and time. If this being created matter and time (and, indeed, all the properties of the physical universe), it must transcend the properties of the physical universe. If space, time, and matter came into being through it, it cannot be dependent on any of those properties. And, therefore, it transcends the limitations of the physical universe (such as contingency). Indeed, the entire point of the Thomistic Cosmological Argument, for example, is that one cannot have an infinite regress of contingent causes, because it is logically incoherent. (One difference between the Thomistic and Leibnizian cosmological arguments is that the former focuses on the concept of “motion” while the latter focuses on the concept of explanation).
If the chain of causation carries on forever, then the universe would not exist, because the cause of the universe would never “arrive.” Because the chain stretches back into infinity, the link that causes the universe would never be “reached.” This, in turn, means that the universe would not exist. The only way to avoid this incoherence is to posit a necessary entity (i.e. one that would not require a cause of its existence). Secondly, as William Craig has noted, this objection distorts the real nature of explanation. If an explanation is proposed, one does not immediately demand an explanation of the explanation. If you always needed to provide an explanation of every explanation you come up with, you would never explain anything at all. You would, in every instance, need to explain your explanation, and then come up with an explanation of that explanation of that explanation… ad infinitum. You would never succeed in explaining anything, because you will always be busy trying to justify your explanations and your explanations of your explanations, etc. Of course, it would be chaos if this criterion, which Russell and others want to apply to Natural Theology, were actually applied to scientific theories.
We’ve seen that the cause of the universe needs to transcend time, space and matter, because it caused the existence of these properties. But why would the cause need to be personal? Scientific explanation has a naturalistic metaphysics, which means that it can only explain in terms of objects in the universe (i.e. material objects). Scientific explanation here is inappropriate since, as we’ve seen, the cause cannot be naturalistic. This means that we can only explain the universe in terms of personality. The only two types of explanation available to us are personal explanation that we use when agents, such as humans, are involved, and the impersonal, mechanistic explanation that is used when explaining the inanimate world (i.e. scientific explanation). You may retort that the cause of the universe could be some sort of non-naturalistic mechanism that is unknown to us. Yet, an impersonal mechanism would need to be “set in motion,” to use Aquinas’ cosmological language. An impersonal mechanism cannot stop and start in ways determined only by itself, because it has no cognizance of itself and its activity. It cannot organize itself. Or, more aptly, it isn’t conscious. Such a mechanism, therefore, cannot be self-existent, or necessary in the way required. Since it cannot organize itself and “set itself in motion,” it would need a cause as well, and we are back at square one. Only a volitional agent can be metaphysically necessary in the way required: something that does have cognizance of itself and its activity, something that is self-conscious. At this point, it become appropriate to talk about “someone.”
 Bruce Reichenbach, “Cosmological Argument”, Stanford Encyclopedia of Philosophy, updated October 26, 2012, accessed June 21, 2015, http://plato.stanford.edu/entries/cosmological-argument/.
 J. L. Mackie, “Critique of the Cosmological Argument,” in Philosophy of Religion: Selected Readings, ed. Michael Peterson, William Hasker, Bruce Reichenbach and David Basinger, 4th ed. (New York: Oxford University Press, 2010), 205.
 The fallacy of composition is an informal logical fallacy where something is assumed to be true of the whole because it is true of a part. For example, that part of a sofa is soft does not imply that the whole sofa is soft. Some parts of the sofa are hard (such as the legs). However, as I indicate, it is not always fallacious to reason from composition, because sometimes properties of the whole are shared by each of its parts.
 Reichenbach, Bruce R., “The Cosmological Argument,” in Peterson et al, 193.
 William Lane Craig, Reasonable Faith (Wheaton: Crossway, 2008), 152-53.