Deus Ex Machina

In some of the recent articles on this blog, we tackled the structure of logic, the nature of truth, and the meaning of existence. With these beginner topics sorted out, we can now talk about God.

Rest assured, I’m not going to guilt you into going to church. Instead, we’ll talk about an ancient philosophical frontier: the efforts to use reason to prove the existence of God. These proofs are commonly lampooned by atheists, but little effort is made to represent the arguments, ahem, faithfully.

We’ll focus on the arguments for and not against because to get rid of God, you seemingly only need to embrace scientific materialism. The hard version of it has no room for otherworldly beings; the soft variant may be indifferent to supernatural entities, but denies them them opportunity to interact with matter. Well — it’s not actually so simple, but we’ll get back to that in a while.

Arguments from design

The most familiar and least interesting category of proofs of God aren’t proofs at all; they’re pragmatic statements about the balance of probabilities. These teleological arguments tend to follow the same outline:

1. There exists x.
2. God is the best explanation for the existence of x.
3. Therefore, God (likely) exists.

In this formula, x is some chosen property of reality that lacks a convincing natural explanation; this might be a nebulous concept, such as beauty or morality, or it might be a firmer observation, such as that the parameters of the universe appear to be fine-tuned for life. To flesh out that last idea, there are maybe six key constants in the standard model of particle physics that, if they differed even by a tiny amount, would probably have precluded the formation of the solar system and the emergence of carbon-based life. As the argument goes, we could have ended up here by extreme luck, but it seems more plausible that they values have been picked with a specific outcome in mind.

The weakest critique of teleological arguments is to sneer at the logical leap in step #2, from “there exists a phenomenon that can’t be explained” to “a bearded man lives in the clouds”. This confuses philosophy with theology; the “God” of the argument is simply an unknowable entity to which the universe owes its shape. The conclusion is broadly compatible with any religion that has a creation myth, and — as we’ll show later — with the beliefs of some avowed atheists.

A more substantial doubt about the argument comes from the observation that much of what used to be inexplicable can be now adequately explained by evolutionary biology, physics, and chemistry. There is little hope for a complete theory of why we’re here, but the realm of gods appears to be shrinking over time, so there’s something cheap about the “god of the gaps”.

In all fairness, the argument about fine-tuned cosmic constants is not so easily dismissed; indeed, it’s the only one that’s considered interesting by many atheists. It can be unconvincingly waved away by invoking the anthropic principle: the assertion that, had the constants been any different, we wouldn’t be here to ask questions about them. However, for this to work, we also need to baselessly assert that some form of cognition is likely in all possible realities. If not, we somehow won the cosmic lottery.

The ontological argument

The ontological argument has fascinated philosophers for more than a thousand years. This is not because it’s particularly believable, but because in contrast to arguments from design, it can be expressed in a way that’s conductive to deeper philosophical inquiry. Further, as Bertrand Russell put it, “it is easier to feel convinced that [the proof] must be fallacious than it is to find out precisely where the fallacy lies”.

The proof can be stated in countless ways; we’ll go with fairly natural phrasing:

1. God exists “in the understanding” — as a concept.
2. God is the most perfect being — that is, the concept possesses all the positive (desirable) attributes it could possess.
3. Actual existence of God is a positive (desirable) attribute.
4. Therefore, God actually exists.

The same proof structure can be used to conjure a maximally-negative devil. Instinctively, the argument feels wrong because it takes agreeable premises and then wills a concept into actual existence. That said, as noted earlier, the flaw in the proof is hard to articulate. There’s nothing overtly wishy-washy or ambiguous about the syntax; in fact, the argument can be spelled out using a flavor of formal logic.

Perhaps the most lucid objection comes from Immanuel Kant, who targeted step #3. Kant posited that actual existence — the instantiation of a conceptual object into the world — can’t be treated as a proper property (predicate) of the concept itself. This has the smells of trying to avoid a paradox of self-reference; it also meshes well with the Meinongian theory of objects from the earlier article on ontology. Alexius Meinong posited that to avoid a variety of nonsensical results (in our example: existent non-existent unicorns), an object’s “so-being” must be treated as disjoint from its physical existence.

If you don’t like artificially constrained theories of objects, another line of attack is to question step #2, which implicitly commits us to a specific definition of God; perhaps it needs to be prefixed with “By my definition, …” or “If God actually exists, …”, this qualifier then carrying over to the conclusion in step #4. Such an intervention turns the proof into a vacuous statement: “therefore, by my definition, God actually exists” or even more damningly, “therefore, if God actually exists, then God actually exists”.

The cosmological argument

Compared to what we discussed before, the cosmological argument for the existence of God is far more compelling, although it doesn’t give the resulting deity a discernible shape — not even in the terms of “perfection” or “desirability”.

To build the argument, we start with a fairly sensible philosophical principle: nothing happens without a sufficient reason. The reason doesn’t need to be good, logical, or even scrutable, but we take it as axiom that some reason exists for all the key features of reality. With this assumption in mind, the simplest version of the cosmological argument is the notion of the first cause: the universe could not exist unless something caused it to be.

At first blush, this argument is mildly incoherent: we demand a reason for everything except the first cause. That said, this is resolved quite elegantly: we posit the existence of necessary and contingent features of reality. A contingent event is something that, when considered in isolation, is neither logically necessary nor impossible. You could have been born or not born. Every contingent event needs a sufficient reason: for you to be born, your parent needed to meet (and then some). A contingent event can’t justify itself, directly or indirectly, as that would make it a logical necessity; in other words, an internal justification would promote it to the category of necessary events.

In this view, if we accept that the universe is contingent — if it did not logically need to exist — then by definition, it can’t justify its own existence. From this, we can view it as a product of a temporal causative chain; at the beginning of that chain, we ought to find a necessary, self-justifying event — our cosmological God.

A subtle issue with this argument snapped into focus with the advances in abstract math: the field gave us coherent models of infinite processes, including sequences that have no start and no end. We don’t bump into actual infinities in the material world, but we could entertain the idea that the universe is a part of an endless cycle; if so, trying to follow the temporal causation chain won’t get us anywhere. If there’s no beginning, there’s no first cause, and by extension, there might be no cosmological God.

The setback can be addressed by ditching the temporal chain in favor of a set: the universe is a collection of contingent features that can’t all circularly justify themselves, so there must have been a necessary entity that provided the initial reason for some slice of it.

Short of rejecting the principle of sufficient reason, another escape route is to posit that the universe contains some necessary elements within; in other words, at least part of it needed to come into existence, and then justified the rest. That said, I think most people would find this idea weird. If it feels better, you still have the option of saying “yes, the argument points to something, but that something ain’t your God”.

The materialist god-entity

At the start of the article, I remarked that the surest way to excise God is to embrace scientific materialism — that is, the belief that the world is governed solely by the laws of physics and that there is nothing beyond the physical realm.

But is it really so simple?… Again, in hard scientific materialism, the behavior of universe is described by the laws of physics; many argue that these laws can be expressed in algorithmic terms. If so, a sufficiently capable computer should be able to simulate entire worlds, along with their sentient inhabitants.

From this premise, Nick Bostrom argues that, unless some unforeseen complication prevents civilizations from reaching the requisite level of technological proficiency, future humans will likely spin up numerous simulations for a variety of reasons, including legitimate scientific inquiry. In fact, who’s to say they haven’t done so already? There’s only one “base layer” of reality and a high likelihood of multiple simulations, so wouldn’t it be logical to assume we’re probably living in one?…

This is the simulation hypothesis. It’s wildly popular on the internet: it’s commonly cited on Hacker News and in other techno-rationalist communities. It also appears to be favorite philosophical position of Elon Musk and hundreds of thousands of his most ardent followers. Last but not least, multiple mainstream pop science outlets, including Scientific American, ran full-length features about the hypothesis and presented it as worthy of serious debate.

But then, if you think about it, what do scientific materialists care? In their view, the real universe is a computer executing executing the laws of physics too. If the reality is purged of free will and other metaphysical ghosts that could tamper with the process, it makes no difference if the computation is executed on the prime computer or if it’s carried out by a sub-computer nested within. Mathematically, the scenarios are indistinguishable.

The truth, I think, is that people fascinated by the simulation hypothesis are not ready to banish their metaphysical ghosts. They imagine the creator of the simulation as a free-willed entity not bound by the rules of our universe. It’s an unknowable, omnipotent, and omniscient God, no different from any big-ticket religious belief. In fact, the simulation-God is uniquely capricious and frightening: he might pull the plug on his creation the moment the research grant runs out.

Some further reading on this blog:

I write original, long-form articles on a variety of geeky topics. If you like what you see, please subscribe.