The term “cosmological argument” (hereafter CA) refers to a whole cluster of arguments or patterns of thinking, all of which draw the common conclusion that God is real based on observations that things we see around us cannot exist unless something or someone else makes them exist. Diamonds, dandelions, dromedaries, and davebecks: none of them can exist apart from a whole range of surrounding factors and causes. This argument says that we must think of God as the originating cause, or the initiating source of things and events, because there cannot be an infinite series of things that make the things around us exist.

Arguments like this seem to be present in every culture and religion. Ancient Greek philosophers Plato and Aristotle developed it into clear form. Christian, Jewish, and Islamic traditions all know it. It can be found in African, Buddhist, and Hindu expressions as well.

Within Western traditions, the best known version is the brief statement of Thomas Aquinas found in chapter 15 of the Summa contra gentiles. Here he says, “We see things in the world that can exist and can also not exist. Now everything that can exist has a cause. But one cannot go on ad infinitum in causes...Therefore one must posit something the existing of which is necessary.”1

As I see it, there are three basic steps in this argument.

Step 1: What We Observe and Experience in Our Universe Is Contingent

Note that this is an observation about the things we actually see and know in the real world around us. It is not intended to be about everything in the universe, only what we have actually experienced. The key word in this sentence is “contingent.” In this context, this means that something owes its existence to something else; it does not exist by itself. It needs a cause. And everything we know anything about is contingent.

So the universe consists of series and networks of causes, which are in turn connected into whole systems of causes. That is, a is caused by b, but only as b is caused by c, and so on. Everything we know exists and functions only as it is caused by other factors in its causal chain. We know of nothing that spontaneously initiates its own causal activity. (Note that nothing here turns on our having to know about everything. Even if something did turn out to spontaneously initiate, it would have no effect on CA.)

Step 2: A Network Of Causally Dependent Contingent Things Cannot Be Infinite

The idea here is that regardless of how complex and interconnected it might be, the series or system of causally related contingent things cannot be infinite. Thomas uses the illustration of a hand moving a stick, which in turn is moving a ball. Perhaps the most frequently used picture in recent discussions is the train.

Imagine you are seeing for the first time a train moving past you. Baffled, you wonder what is causing the boxcar to move. You come to realize that it is being pulled by another boxcar in front of it, and so on, down the tracks beyond your view.

This picture allows us to visualize the various atheist scenarios, so commonly heard in our society, that attempt to describe how it is that things exist in our universe. “The cosmos is a great circle of life,” we are told. But if we were to connect the boxcars all the way around the universe in a circle until the last one hooks up to the first, that will still not explain the motion of even the first boxcar, let alone any of the others. If contingent things cause each other to exist in a circle, there is still nothing to initiate the causal process. The atheist offers perhaps a more promising scenario: “The cosmos is an intricately evolved ecosystem in which everything is related causally to everything else.” So boxcars clutter the universe in an unimaginably complex system of railroading such that in some way every boxcar is coupled to every other one and therefore every boxcar is pulling that first one. Still, we have no accounting for the motion of that first boxcar. Likewise, the notion of an evolving ecosystem does not account for the existence of any actual thing in our universe.

It is always tempting, of course, to say that it is enough to know that each boxcar is being pulled by the boxcar in front of it. In one sense it is clearly true that boxcar a is pulled by boxcar b. But b can pull a only because c is pulling b at the same time. The pulling action of b is transferred from c. And so it is also true that a is being pulled by c. The same is true, of course, about d, and about e, and so on. In practice, it is enough to know that AIDS is caused by HIV, but it is sometimes important and always possible to keep asking the questions if we really want to understand our reality.

One last option suggests itself. Suppose that there are just infinite boxcars or, as the atheist might say, “The intricacy of the universe is lost in infinite complexity.” But infinite boxcars, no matter how complex their arrangement, still leave unexplained why our first boxcar is moving and hence why any of them are. Letting the questions of cause go to infinity fails to explain anything at all.

Step 3: A Network of Causally Dependent Contingent Things Must Be Finite

This last idea simply draws the obvious conclusion from step 2. If the network cannot be infinite, then it must be finite. No other option remains, unless one wanted to argue that nothing actually exists. But that is not a rational option.

Conclusion: There Must Be a First Cause in the Network of Contingent Causes

If the causal sequence is finite then by definition there must be a first cause. This concept of “first cause” obviously contains two ideas. If it is the first cause then it neither requires nor has a cause of itself. First is first! So it is fundamentally different from every other cause in the series. It depends on, is limited by, or exists because of absolutely nothing else. It truly initiates causality.

On the other hand, to say of the conclusion that it is the first cause is to define its relation to everything else in the network: it is the cause of all of them. It initiates all of the causal activity without negating the fact that each subsequent cause is itself a cause of the following one in the series.

To switch back to our train analogy, the only explanation for the moving boxcars is that somewhere there is a locomotive with the capacity to pull the entire train. The notion of “first cause” is richer than it might at first appear. It is always the case that there are two correct answers to the question of what is causing the existence of something. The immediate cause or causes as well as the first cause are equally correct answers. Both the locomotive and the boxcar in front of boxcar a, namely b, are the cause of the motion of b.

Some Objections

Atheists typically bring four kinds of objections to this argument. The first and certainly the most frequent criticism of CA is the “It’s not God” objection. What the conclusion delivers is only a vague and undefined first cause: an initiating source of the universe. This may just be the Big Bang, elementary particles, energy state, or even an original vacuum. In any case, the conclusion does not deliver an infinite creator God who is personal and relational.

In response we could vote that what CA does give us, after all, is a first and uncaused cause. This is already enough to defeat atheistic naturalism, which holds that the universe is a complex causal system existing on its own, purely by chance, without any external source, and consisting of only natural and finite components. The argument demonstrates that there is at least one non-contingent component.

On the other hand, if the objection is that the conclusion is too “small,” that God is so much more, then we should just agree. It certainly tells us only a little about God. Those who use this objection often seem to suppose that unless we know everything about God then we do not know anything. This is obviously false. We know much about many things without knowing everything about any of them.

A second objection is the “Infinite series are possible” claim. CA depends on a denial of an infinite series of causes. Yet the sequence of cardinal numbers, as we all learned in elementary school, is infinite. Since we can assign a cardinal number to each member of a causal sequence, we then have an infinite sequence of causes.

This objection overlooks the specifics of the network of causes in CA. It has four critical characteristics:

  1. It is a system, an interconnected network of causes and effects
  2. Each cause is contingent; each one itself needs a cause
  3. The dependency in the Aristotelian/Thomistic CA is concurrent not chronological; it refers to concurrent dependency relations (think of the moving train analogy!) within a system of causes
  4. The specific relation to which the general CA refers is the causing of existence itself

The key point in CA is that there cannot be an infinite series of causes with all four of the above characteristics. This is not to say that there could not be infinite series of other kinds, including some very similar ones, such as chronological sequences of causes in time (such as a series of parent/child relationships). However, there cannot be an infinite series of contingent things.

A third typical objection is the “We do not know about the whole universe” contention. We have no way of knowing that everything is contingent. The simplest way to answer this is to admit that it is true. Obviously we do not know about everything in the universe. However, the conclusion holds regardless. What the argument shows is that if there is something that is contingent, then there must be something that is not.

Initially this might seem to leave the possibility of multiple gods. Granted, CA by itself does not eliminate that option. However, as Aquinas learned from Aristotle (and as Parmenides understood much earlier), there can only be a single uncaused or infinite being. A second infinite being would have to be different from the first in some way, but one infinite being cannot be either more or less than another. We all learned early on that infinity minus or plus anything is still infinity. So there can only be one infinite being.

A fourth objection is the “What caused God?” counter. If the universe is a network of causes and effects, then you cannot arbitrarily stop at some point and just call it “God.” This, however, simply misses the whole point of the argument. The CA shows that a series of contingents must be finite: it must eventually lead to a non-contingent. It would be nonsense to ask what causes this first uncaused cause. So this objection simply fails to understand the argument.

In conclusion, it is worth pointing out two things. First, this is of course a very brief and overly simplified discussion of an argument with a long and diverse history, about which many books have already been written. This has hardly done it justice. Second, this argument opens up a wealth of additional conclusions that follow from the concept of non-contingency. Together, they round out a much fuller concept of God not apparent in the simple CA itself. Both of these points encourage a fuller study of CA. But what this argument does show is that we cannot make sense of the universe, the reality in which we live, apart from there being a real God.

Notes

  1. Thomas Aquinas, Summa contra gentiles 15.124