The main theme in Peter van Inwagen’s Metaphysics is uncertainty. In most chapters the author enters with his refutation and exits with a tenor of inconclusiveness. The liberal use of modal logic in countering some of the arguments for God’s existence was a favorite tactic. I kept thinking the pea must be under the cup in possible-world three, but the author’s logical sleight-of-hand was too quick for me. My conclusion: There is a set of possible worlds in which van Inwagen’s use of modal logic leaves little doubt in the reader’s mind, but unfortunately our actual world is not in that set. I probably just need to brush up on modal logic – but in the meantime I want to turn to van Inwagen’s treatment of the design (or teleological) argument from the fine-tuning of the cosmos. This variation of the classical argument for the existence of God is one I personally find compelling. Here too the author leaves things unresolved and the counter-positions at par. We are told neither a rational designer nor brute materialism has the upper hand. The universe may find its ultimate origin in God or in some material realm beyond the boundary of our observable universe – take your pick. But did the author successfully make his case?

I agree with van Inwagen the Arche is either a Chaos or a Logos. The Greek word arche or origin represents the foundation of existence in which all things rest. According to the author, it is either grounded in meaningless Matter (Chaos) or purposeful Mind (Logos). The ultimate origin of our observable universe is either God[i] who created it ex nihilo or some unobservable hyper-reality which spawned it ex materia[ii]. I agree with van Inwagen there are no other alternatives worthy of consideration. The author does a good job dispelling the nonsense suggested by some pop-writers of an observable universe exploding into existence out of metaphysical nothingness. From nothing, nothing comes, plain and simple. All attempts to state otherwise completely miss the boat on what nothingness really is. True metaphysical nothingness is what rocks dream about – as Aristotle put it. When we talk about voids in space or the quantum vacuum, those things are emphatically something. Creation from these would be considered ex materia (from material). So we are on the same page; our observable universe began to exist, and the Arche is either a Chaos or a Logos. Now it’s easy to see how fine-tuning squares with a Logos (since there is an empirical correlation between fine-tuning and a designer), but how do we square it with a Chaos?  

Various astrophysical constants and parameters from the Standard Model, including their relationships, are narrowly just-so for the existence of a universe with conscious observers. Van Inwagen’s recognition of the overwhelming improbability raised by these known anthropic-coincidences is in line with what most experts say on the subject today (Penrose, Davies, Hawking et al.) Nontheistic cosmologists have been working for decades to get around the theological implication of the apparent fine-tuning of the universe by offering several hypotheses. However, van Inwagen doesn’t buy most these. The author rightly rejects the Weak Anthropic Principle using his short-straw analogy. He rejects the Strong Anthropic Principle and anything involving a life principle. He doubts there will be a theory-of-everything revealing how conscious life obtains necessarily. So where does van Inwagen put his money? How do we get from a Chaos to a fine-tuned universe? His answer is the multiverse and the observer selection effect. This is the only plausible alternative to the Logos according to van Inwagen and it does seem to be growing in popularity. 

Why does van Inwagen think the multiverse hypothesis is as good as a rational designer? He tells us here: 

An explanation [e.g. the multiverse] will be at least as good as that explanation [a rational designer] if it contains no element known on independent grounds to be false or improbable-for that (together with the fact that it does explain the observed phenomenon of the fine tuning of the cosmos) is really all that can be said in favor of the hypothesis of rational design.”[iii] 

So basically, the author considers both hypotheses to be little more than logically consistent proposals lacking any known defeaters. They both provide an explanation of apparent design, but that’s about it. All in all, they are on par according to van Inwagen – and this is where he and I diverge. On the one hand you have a Designer as the source of cosmic fine-tuning, and on the other, a hypothesis with ad hoc constraints and assumptions. 

The multiverse is basically an ensemble of independent universes, each one like our very own observable region. Van Inwagen uses the term cosmos to refer to one of these. He uses the term cosmoi to refer to a plurality of cosmos-siblings in the multiverse, each with its own randomized physical parameters (Physics.) So in our universe the ratio of the mass of a proton to an electron is about 1836 to one. But in another cosmos it might be two to one, or a million to one. This would apply to all of the constants and relationships found in the Standard Model as well as other cosmological values. The multiverse is then treated as a cosmoi generator where our unlikely cosmos is just one in the ensemble. And this generator must churn out more than a few cosmoi in order to overcome the magnitude of the improbability of a fine-tuned cosmos suitable for conscious observers to obtain by chance.

The author rightly recognizes the overwhelming mount improbable in the anthropic coincidencesHoweverhe fails to mention how the problem is compounded by the fact: a suitable abode is a necessary condition for life to exist but not a sufficient condition for life to obtain. It is no better than me saying: the conditions are right for a garden in my backyard, therefore a garden in my backyard will come into existence. The height of mount improbable is not just due to the right settings for a suitable abode on the cosmoi generator, but it is compounded by the odds of life obtaining (abiogenesis) from purely material causes. Once you factor that in, it’s no wonder why the author and cosmologists like Alexander Vilenkin posit an infinite number of cosmoi. Yet van Inwagen does not seem to have any problem with this approach: 

“It’s hard to think of a reason to suppose the number of actual cosmoi would have to be finite. If the number of cosmoi were infinite, it would certainly not be surprising that some of them were suitable abodes for life.” [iv] 

Positing a mind-boggling number of cosmoi is ad hoc enough[v]. But why not stop there? Well, because the only way to secure the multiverse hypothesis is to suggest an actual infinite set of cosmoi. However actual infinities turn the multiverse hypothesis into a bizarre concept. Under this thesis, there are not only an infinite number of identical clones of myself typing this sentence right now, but an infinite number of clones typing it backwards, yesterday, and on their head. Think about it, if we open Pandora’s Box and allow for actual infinites, we turn possible-world semantics into reality! In other words, every possible world that is actualizable [vi] has in fact been actualized in the infinite multiverse. Might this include a possible world with an omnipotent and omniscient being capable of transcending its cosmos into all sibling cosmoi? I don’t see why not. In order to deny this, one has to postulate an ad hoc constraint: Being cannot transcend its cosmos – even a maximally great being. In other words, demigods exist in some cosmoi, but they are stuck there with limited greatness. Assumptions do not get much more ad hoc than this, and if you remove it, then why wouldn’t an infinite number of what we think of as God exist?

But it gets worse; the Physics in each cosmos must be based on free parameters. If there were only, say ten parameters, allowing for only ten discrete values each, then you would have a mere ten billion possible Physics regardless of how many cosmoi are generated. Therefore, a much higher (i.e. infinite) degree-of-freedom in how cosmoi-physics is configured must be assumed. Here’s what van Inwagen says:

“[there is] the possibility that the cosmos might have arisen as a fluctuation in some pre-cosmic analogue of the quantum field…We suppose that the cosmoi that arise in Chaos do not resemble one another as closely as the bubbles in a pot of boiling porridge resemble one another. The differences among them-which, we must remember, are the products of chance-are, or can be, of the radical kind we should describe as differences in the laws of physics and large-scale cosmic structure. ”[vii] 

The problem with the boiling porridge analogy is it oversimplifies things and obfuscates an assumption. Boiling porridge creates lots of boring blobs of oats. They’re not very interesting. Not only must the physical parameters of the cosmoi be freely variable, as I just mentioned, but the law (or laws) of the multiverse, the Arche, must allow for the generation of randomized sets of Physics suitable for life. It would be impossible to get conscious observers from Chaos without a sufficiently creative potential in the Physics generator. Imagine trying to create a functioning computer from mere plastic Lego. Regardless of the variation of shapes, the number of pieces or the number of attempts; it’s impossible. Without conductor and semiconductor materials, plastic alone will never get you there.

A parsimonious view should reject a Chaos generating Physics with the richness and complexity necessary for life to ever obtain, even with an actual infinite number of randomized attempts. Why not an infinite number of cosmoi where each cosmos is merely empty space or contains nothing but a random number of elementary particles? It seems exceedingly convenient the kinds and number of dials on the cosmoi generator even allow for conscious observers under any configuration. Now this is where van Inwagen might introduce his observer selection effect by saying, yes, it does seem odd for conscious observers to be a random byproduct of Chaos, but that’s what it takes to recognize the oddness – we have to be here. But this would assume the very thing we are trying to prove and only makes sense once you violate a parsimonious view and include the assumption the multiverse must generate rich rather than frugal Physics. 

In conclusion, the choices for Arche are not on par. On the one hand you have actual design by the Logos and on the other, apparent design by a Chaos. The Chaos option holds only as long as we accept an unobservable actual infinite [viii] set of cosmoi in a multiverse; each with different Physics; some with demigods who cannot escape their cosmos; and others with Richard Dawkins as a television evangelist. Somehow a meaningless Chaos must have the potentiality in its cosmoi generator to create not just randomized Physics, but Physics sufficiently rich and complex such that given an infinite number of spins, conscious observers will obtain instead of nothing more interesting than a sea of quarks or globules of cooked oats. You be the judge which one takes more faith.

[i] I purposely leave out gods at this point or make any claims as to the nature of God

[ii] I leave out here all of the discussion around the Borde-Vilenkin-Guth theorem ( – theists like Dr. William Lane Craig say the BVG theorem proves an ultimate beginning for the multiverse as well. However Vilenkin himself seems to have some reluctance. Alan Guth is less reluctant and said that we do not have absolute certainty but it appears there must be an ultimate beginning for a multiverse as well. If true, warrant shifts substantially to the Logos option.

[iii] Kindle 4546 of 7762

[iv] Kindle 4522 of 7762

[v] Shots at this range from 1e500 to 1e10^150 – ridiculously large (but finite) cosmoi

[vi] Actualizable meaning: where there is neither logical contradiction, nor violation of the kinds of Physics created by the cosmoi generator, nor absurdities [like a prime minister being a prime number]

[vii] Kindle 4526-4528 of 7762

[viii] I agree with the great mathematician David Hilbert that an actual infinite set is nowhere to be found outside of a concept of the mind - "let us draw the conclusion from all our reflections on the infinite. The overall result is then: The infinite is nowhere realized. Neither is it present in nature nor is it admissible as a foundation of our rational thinking - a remarkable harmony between being and thinking. (David Hilbert)"