The Cosmological Argument for Dummies

Given that I’m the author of several For Dummies books and a title on natural theology, I thought I might be qualified to offer some simplified (but hopefully not simplistic!) thoughts on cosmological arguments.

My aim for this post is modest: to give people an idea of what cosmological arguments are— their general approach and strategy—not to defend any particular formulation. Still, I hope this serves a useful purpose since many seem to reject cosmological reasoning based on an extremely cartoonish understanding of what it is (I can already hear someone shrieking “god of the gaps!”). If nothing else, I hope this brief introduction encourages people to give more rigorous formulations a fair hearing.

So, what are cosmological arguments? Speaking generally, a cosmological argument seeks to explain why anything exists (rather than nothing) by identifying certain features of reality—such as change, contingency, or compositeness—that cannot fully explain themselves.¹ It argues that these features require an external cause or explanation, ultimately leading to the conclusion that there must be a fundamental, necessary being—something unlike the contingent or changing or composite things we observe—which serves as the ultimate source of all existence. This being is typically understood as God.

Note: Sometimes cosmological arguments are called “first cause” arguments, but this label can be misleading, which is why I resist it. It can be misleading because many cosmological arguments—in fact, I think the strongest ones—aren’t concerned with finding a causal entity that is temporally first, like something that existed at the beginning of time. Rather, they focus on identifying an explanatory entity that is primary and fundamental, operating within a nested causal chain that produces things here and now (often called a synchronic causal hierarchy). In other words, these arguments are less concerned with what caused things in the past and more with what ultimately produces things that, by their nature, need not exist right now.

Put differently, cosmological arguments claim that there must be some cause categorically distinct from the caused category under consideration—e.g., an unchanging thing to explain changing things (Aristotle’s way), a noncomposite thing to explain composed things (Plotinus’s way), a necessary thing to explain contingent things (Leibniz’s way), and so on. This holds true—so the cosmological reasoner will argue—no matter how many things fall into that category (whether finite or infinite) or how those things are arranged (linearly, in a circle, etc.). And while it’s not always the goal of the cosmological argument to fully examine the nature of this uncaused (or unchanging, non-composite, or necessary) entity, many argue that this is what people mean by God.²

To put it more colloquially, cosmological reasoners say this, essentially: “There’s something about certain things that requires us to invoke an outside explanation if we are to make sense—good, adequate sense—of why these things are here, existing, or have the characteristics they do, or both. This explanation cannot be provided by those things themselves, no matter how many there are, and no matter how they’re arranged. To get an adequate explanation, we must transcend these types of things entirely. For example, to explain why there are any contingent things at all, we must eventually arrive at something necessary—something that exists in the most robust way possible, that cannot not exist, and thus serves as the foundational existential anchor for everything else. Otherwise, there would be no contingent things in the first place! Moreover, here are reasons to think that this something is deserving of the description ‘God…’”

Again, a crude statement, but hopefully illustrative for introductory purposes. At this point, perhaps the best way to further elucidate cosmological reasoning is to spell out some of the obstacles it must overcome to establish the existence of God. While there are many obstacles and objections to cosmological reasoning (I examine a whole “museum” of them in my book), the major ones are as follows:

1.    The Justification Problem

2.    The Regress Problem

3.    The Gap Problem³

The Justification Problem

The justification problem is this: why should anyone believe in the causal or explanatory principle required to run a cosmological argument? (In other words, what justifies it?) For example, why should we believe that all contingent things are caused, or more broadly, as other formulations hold, that everything is intelligible—meaning everything has an adequate explanation, either through an external cause or the principles of its own nature?

The principle that everything is intelligible—that everything has an adequate explanation for its existence—is a broad statement of the Principle of Sufficient Reason (PSR). This principle denies the existence of brute facts. A brute fact is something that seemingly requires further explanation—something with limited intelligibility, especially regarding why it exists—yet has no answer. It’s the idea of, “It just is, and that’s all!”

It’s generally safe to say that classical theists are committed to the idea that reality is fully intelligible. This doesn’t mean reality is fully intelligible to us, but rather that it could be fully understood, insofar as there is an adequate answer to every coherent question that could be asked—even if only God has those answers!

To frame it differently, one might think of these explanatory principles as the engines driving cosmological arguments. For them to work effectively, they must be powerful engines indeed, capable of plowing through all natural entities and pushing toward the transcendent—no easy task, as what counts as “natural” isn’t always clearly defined.⁴ Nevertheless, to the extent that a causal or explanatory principle applies to certain types of things—not just contingent things, but composite things, changing things, or qualitatively finite/arbitrarily limited things—the more obvious it becomes that whatever this principle points toward is quite unlike anything most of us would consider natural. In other words, it points to something beyond space and time, without arbitrary limits in power, shape, position, knowledge, etc.

There are various ways to motivate these explanatory principles (shameless plug: I develop several in my book). Some arguments appeal to self-evidence or common sense, others to inference to the best explanation, and still others use dialectical strategies to show the catastrophic consequences of denying such principles (for example, leading to empirical skepticism or undermining scientific knowledge). Some thinkers even believe that denying these principles leads to contradiction.

Again, I leave these matters open for exploration; I’m just reporting how this business is typically done.

The Regress Problem

The regress problem addresses the suggestion that an infinite regress or loop of certain things, like contingent things, could itself be a sufficient explanation. If one posits an infinite regress as a brute fact, this relates more to the explanatory issue. For the regress issue to stand on its own, it must propose that such a regress somehow satisfies the explanatory principle.

Cosmological reasoners typically respond to the regress issue in two ways. The first is to argue that it’s irrelevant; the second is to argue that it’s impossible. Each strategy has its nuances, but the general response (omitting many details) is something like this: Even if there were an infinite regress or collection of contingent things, the principle of explanation demands we move beyond the category of contingency to necessity in order to find an adequate explanation for why there are any contingent things in the first place. This approach claims that an infinite regress is not incompatible with the need for a fundamental cause.

Think of it this way: if the principle of explanation demands a reason for why anything contingent exists, then pointing to any particular contingent thing—or even an infinite regress of them—still assumes the very thing we are trying to explain: contingency itself. To assume the thing to be explained in your explanation is hardly sufficient.

Leibniz offers a famous example involving an infinite chain of geometry books, where each book is a copy of the previous one. He argues that appealing to an infinite chain here is not a fully adequate explanation, because it never addresses why there are geometry books at all—let alone why the subject is geometry and not, say, biology.

The second strategy—arguing the impossibility of an infinite regress—divides into two general approaches as well:

1.    Causal Finitism: This asserts that an infinite sequence of causes is impossible (i.e., no effect can be preceded by an infinite number of causes).

2.    Essential vs. Accidental Causal Series: This distinguishes between causal series that must have a primary member and those that might not. This approach is more specific than the first but also more modest insofar as it allows infinite causal series of a certain sort.

The latter approach requires understanding Aquinas’s distinction between essentially and accidentally ordered causal series. In short, Aquinas claims that causal series exhibiting “deep dependence” relations—specifically when the very existence of something is immediately dependent on a prior condition (or set of conditions), which is itself dependent on another prior condition, and so on—cannot regress infinitely. Why? Because the entity in question does not possess existence by virtue of what it is; it receives existence for as long as it exists from another source. In such a causal series, we must trace back to a fundamental cause that does have existence by virtue of what it is—what Aquinas says must, if we are assuming his metaphysic, be something whose essence is its existence (i.e., God). Without such a fundamental cause, we lack an adequate explanation for how "existence" entered the causal series in the first place and produced the ultimate effect (in this case, the existence of the contingent thing in question). Otherwise, existence would have simply come from nothing, which Aquinas considers absurd since he was definitely committed to PSR or something near enough. Thus, an infinite regress of things whose existence is essentially derivative does not explain how any such things actually exist. Fairly intuitive, I think.

As noted, Aquinas’s argument that some causal series must terminate leaves open the possibility that other causal series are non-terminating. For Aquinas, causal series that are accidentally ordered—like a father begetting a son, and that son begetting his own son—do not need to trace back to a primary member. This is because the causal property in question (the power of begetting) is possessed by the thing in virtue of its nature. Humans, for example, have the power to beget because they are rational animals. That power is not immediately "borrowed" from some prior condition in its use. In fact, this distinction between essentially and accidentally ordered causal series is key: if, in causing Z, Y must borrow the power from X, then the causal series X → Y → Z is essentially ordered. If Y does not need to borrow the power from X to cause Z, then the causal series X → Y → Z is accidentally ordered.

I believe Aquinas is right that causal series exhibiting deep dependence relations are either necessarily terminating or non-explanatory, and I argue this case more rigorously in my article Is Grounding Essentially Ordered Causation? (December issue of The Review of Metaphysics).

Causal finitism is a stronger thesis than Aquinas’s since it entails the impossibility of even an accidentally ordered causal series, which Aquinas did not strictly rule out. Nevertheless, this thesis has been strongly argued in recent years, particularly by Alexander Pruss and Robert Koons. Pruss’s strategy is to present a number of paradoxes (sometimes complex but all very clever), all of which can be resolved by adopting this thesis. I won’t get into the details here—for those interested, consult his book. I’ll just say that if one adopts this stronger thesis, it unlocks a greater variety of cosmological arguments.

The Gap Problem

The gap problem is essentially this: how can we say that whatever we’ve reached through the cosmological argument—say, a necessary being—is actually God as traditionally understood by theists (i.e., all-powerful, all-knowing, and all-good)?

There are several strategies to address this. The first is the most modest and simplest. Some cosmological arguments don’t aim to go far; they may be content with establishing a necessary being—whatever that may be—and leave it there. Proponents of these more modest approaches might suggest that it’s odd to think of a necessary being as any sort of physical reality—it sure doesn’t seem like anything physical is necessary— but they don’t necessarily seek to rule it out.

Others go further, but how far they go, and how they identify the fundamental entity as God, varies significantly depending on the type of argument used and the explanatory principle deployed.

For example, some argue that moving partway toward the divine attributes—or even just identifying one divine attribute or something similar—is enough. Suppose a cosmological argument leads us to a being that exists outside physical reality and is responsible for the existence of physical reality. This gives us two characteristics: something non-physical and extremely powerful. While this may not fully define God or omnipotence, it seems compatible with, and points in the general direction of, God. So, unless there’s a reason to think it isn’t God, why not conclude that it is? Especially since it might be simpler—according to the virtue of theory selection—to 1) rely on one productive principle rather than many, and 2) choose one that isn’t arbitrarily restricted along certain dimensions, as God isn’t (traditionally, God has unrestricted power, knowledge, goodness, etc.).

Additionally, proponents of this more modest approach may supplement the cosmological argument with other arguments in natural theology, such as the fine-tuning argument, which might suggest a “super-calculating super-intellect” is at work. The result is a being that is extremely powerful and intelligent. Again, while this might not match the fully robust conception of classical theism, it’s certainly not incompatible with it and is suggestive of it, especially if one has other, independent reasons for believing in God. (This approach is taken by Michael Rota in Taking Pascal’s Wager; he essentially blends the cosmological argument with the fine-tuning argument to make a cumulative case for God).

Of course, there are more robust approaches to bridging the gap as well. Joshua Rasmussen, for example, proposes a principle of explanation that seeks to explain all arbitrary, non-maximal limits. This leads to a fundamental reality with no arbitrary limits whatsoever, including in power, knowledge, or goodness (which Rasmussen argues are legitimately attributed to a fundamental reality based on what it must explain). This approach gets you close, if not all the way, to traditional theism—assuming one accepts his principle and follows it through. Others, like Ryan Byerly, argue you can move from the idea of a necessary being to a perfect being (Byerly’s work can be found here). I believe there are several promising approaches along these lines, even if they don’t fully match the traditional conception of God.

Ultimately, my preferred approach is to run a cosmological argument within a specific metaphysical system—namely, Thomas Aquinas’s, being the, ahem, unreconstructed manualist two-tier Thomist that I am (that’s not true, by the way, but I do like the way it sounds). This approach leads to the conclusion of a being whose essence just is its existence (sounds strange, I know, but you’ll have to read up on it). Given Aquinas’s theory of existence, this being must be: 1) uniquely unique (there can only be one, out of metaphysical necessity), 2) purely good, 3) omnipotent, 4) omniscient, 5) eternal (outside of time), and so on. In Aquinas’s system, the divine attributes are metaphysically entailed by pure actuality, so there really is no "gap" issue.

It’s actually a bit amusing because the so-called "gap problem" is often treated as a contemporary worry, as if it’s something newly discovered. People don’t seem to realize that Aquinas spent considerable time addressing this, showing why a being whose essence just is its existence—or is purely actual—fits the definitive description of God. He argues for attributes like simplicity, perfection, immutability, unity, infinity, knowledge, will, justice, mercy, beatitude, etc. (Summa Theologica. Part I, Questions 3–26.). Check it out if you want.

While Aquinas's approach may rely on a "controversial" (oh no!) metaphysical system, it has the distinct advantage of logically ruling out the possibility that anything other than God could be the ultimate, foundational reality reached through cosmological reasoning. Personally, I think that’s worth the trade-off because 1) I find his metaphysical system to be impressively robust, and 2) it tends to provide far more conceptual clarity about the nature of God and His relationship to the world than other approaches.

Summary

In a nutshell: cosmological reasoning identifies certain aspects of the world—or specific things in the world—that seem to shout, “I am insufficient; something causes me!” There must be something that accounts for these features, which is itself devoid of them—is not contingent, changing, composite, or whatever else. Philosophical analysis of such a reality points toward something very much like God.

I pray this introduction has, at minimum, clarified what cosmological arguments are aiming to do—I haven’t tried to fully articulate or defend any specific argument here, but I do think several of them are quite compelling. Here, then, are some recommended readings on the subject.

Of course, I’m going endorse my book first—The Best Argument for God.

Other popular level, but still highly instructive books on cosmological reasoning, include: How Reason Can Lead to God, Taking Pacal’s Wager, Five Proofs of the Existence of God, The Philosophical Approach to God, and Introduction to Philosophy of Religion.

If you want more advanced material, then: The Blackwell Companion to Natural Theology (Alex Pruss’s article), Classical Theism: New Essays on the Metaphysics of God (Robert Koon’s article), Aquinas’s Way to God and Collected Articles on the Existence of God.

1. I once heard a philosopher explain cosmological arguments as tools for investigating the foundation of reality. I like that!

2. This paragraph taken from my book The Best Argument for God (though with slight modification).

3. I borrow this scheme from Robert Koons, who attempts to overcome each of these issues—which I believe he does—in his article God’s Existence, featured in Neo-Aristotelian Perspectives In Metaphysics.

4. For example, a causal principle that states “only things after the first physical thing—whatever that is—require a cause”, would not be powerful enough to run a cosmological argument of theistic significance.