Cameron Bertuzzi has recently shared the following Argument for God from Contingency to the Capturing Christianity Facebook page:

(1) For any contingent thing or group of things, there is an explanation of the fact that it (or they) exists.

(2) Considering all the contingent things that exist, if there is an explanation for those things, then there is a necessary part of reality.

(3) There is a necessary part of reality (from 1, 2).

(4) If there is a necessary part of reality, then it has some perfections.

(5) If it has some perfections, then it has all perfections.

(6) If it has all perfections, then God exists.

(7) God exists (from 3, 4, 5, 6, 7).

This is a variation on a standard cosmological argument, the contingency argument, and has antecedents in Leibniz and Clarke. However, I think this argument is deeply flawed.

To start, premise (5) is too strong. While Cameron is not offering the ontological argument, Anselm appears to have adopted a similar principle when defending the ontological argument against Gaunilo. Anselm had pointed out that the idea of God appears to entail God’s existence. According to our idea of God, God is more perfect than any other being that can be conceived. We can conceive of a being that exists and one that does not; since God is more perfect than any other that can be conceived, God must exist. Therefore, Anselm concluded, God exists. Gaunilo famously objected to Anselm’s ontological argument by postulating the perfect island, that is, an island that is more perfect than any other island that can be conceived. Since any island would be more perfect if it actually existed than if it didn’t, if Anselm’s ontological argument for God’s existence succeeds, then so, too, Gaunilo’s argument for a perfect island succeeds. Yet Gaunilo’s perfect island argument is obviously absurd and so unsuccessful. Hence, Gaunilon concluded, Anselm’s argument must not succeed either. Anselm replied that if any x were a perfect island, then that x would be God. That is, an island would be more perfect if it had God’s other perfections and weren’t even an island. Hence, for Anselm, if some x has one perfection, then that x must have other perfections.

I’m not persuaded. Suppose that Persephone is not only an avid reader, but the best reader that there has ever been. She instantly understands any sentence she reads. She understands the relevant symbolism, the full context, and the way that the words hang together. Provided any conceivable reading comprehension task, Persephone instantly achieves the highest possible score. And so on. Yet nothing follows about whether Persephone is perfectly morally good or a perfect runner. In other words, from the fact that Persephone has one perfection, nothing follows about whether Persephone has other perfections. Hence, it’s not true that if x has some perfections, then x has all perfections.

Anselm/Cameron might object that Persephone would be more perfect if Persephone had all of the perfections. But I don’t see why that should matter. Persephone would be no better as a reader if Persephone had all of the perfections. In fact, for some items — like islands — having all of God’s perfections, in the way conceived of by theologians, is straightforwardly incompatible with being a perfect member of its kind. For example, anything that is an island is not also immaterial.

Perhaps Cameron can make do with a modified premise, e.g., that if some x has some perfections, then, probably, x has all perfections. This sort of premise might be argued for inductively, e.g., the fact that x has some perfections gives us reason to think that x has other perfections, or abductively, e.g., the fact that x has some perfections is best explained by x having all perfections. I am doubtful of both strategies.

On one hand, consider, as I’ve already discussed, that there are instances in which being a perfect member of a kind, e.g., the perfect island, is incompatible with having God’s perfections, e.g., being immaterial. In that case, the inductive argument fails because being a perfect member of a kind is not evidence for having all perfections and the abductive argument fails because being a perfect member of a kind is not best explained by having all perfections.

On the other hand, having one perfection, e.g., being a perfect reader, is not evidence for having infinitely many other perfections, e.g., being able to do infinitely many tasks perfectly. Having infinitely many perfections may be a simple explanation for having one perfection, but it’s also a maximally immodest explanation, since there are then infinitely many ways that an entity could fail to have a perfection. (Consider: perhaps the fact that x can complete one task provides evidence that x can complete other tasks. But surely the fact that x can complete one task is not good evidence that x can perform all other possible tasks.)

In the discussion below Cameron’s post, he indicates that he lifted premise (5) from Andrew Mooney’s (2019). In turn, Mooney takes the premise from T. Ryan Byerly’s (2019). However, Byerly does not argue that any being with some perfections has all perfections. Instead, Byerly argues that, in the context of a cosmological argument like Leibniz’s, once a necessarily existent being has been established, we can use an inference to the best explanation in order to infer that the necessary being is a perfect being. We can use an inference to the best explanation to infer that the necessary being is a perfect being because universal generalizations explain their instances; the necessarily existent thing having all perfections would explain why the entity has the perfection of necessary existence. This suggests compressing Cameron’s premises (4) and (5) into one premise: If there is a necessary part of reality then, probably, that part has all perfections.

Byerly’s argument is subject to many of the same objections I’ve previously articulated. For example, since an entity being a perfect member of its kind may be incompatible with the entity having some other perfections, we have reason to think that no entity can have all perfections. To save his argument, perhaps Byerly can identify a different reason that the entity’s being divine would be the best explanation of the entity’s necessary existence.

For example, Christian philosophers pursuing an Anselm-inspired strategy to articulating the nature of God do not typically say that God has all perfections. Instead, they pursue Perfect Being Theology, that is, roughly, that God is a perfect being means that God has the greatest compossible collection of properties (Morris 1991). In that case, God doesn’t have all perfections. But, in that case, it also becomes much less obvious to me how the abductive move from necessary existence to perfect being is supposed to work.

Even if Cameron did successfully make the alterations to his argument that my objections suggest, I think the combination of premises (1) and (2) is implausible. Premise (1) claims that any contingent thing or group of things has an explanation for its existence. Consider the group consisting of the towel hanging in my bathroom and the number 3. Presumably, each of these items has an explanation for its existence. But does the group have an explanation? I think many philosophers are going to be doubtful that just any arbitrary grouping has an explanation. In correspondence, Alex Malpass has told me that he doubts that just any arbitrary groupings have explanations. Setting aside the intuitive view that some arbitrary groupings do not have explanations, I can think of two ways that the grouping consisting of the towel and the number 2 (or any other arbitrary grouping) has an explanation; yet, as I show below, both entail that (2) is false. This suggests that the combination of (1) and (2) is implausible.

First, perhaps if we have an explanation for each member of a collection, then the explanation of the entire collection is just the conjunction of all of the individual explanations. For example, perhaps the towel can be explained by the actions taken by the towel manufacturer, the material from which the towel was constructed, and so on, while the existence of the number 3 is explained by the fact that the number 3 necessarily exists. And then the grouping of the towel and the number 3 can be explained by the combination of those two explanations. But surely that’s not what Cameron intends. For suppose that we have an explanation of each contingent thing in terms of some other contingent thing. In that case, we’d have an explanation for every contingent thing; so, to explain all of the contingent things, we just need to conjoin all of those individual explanations. In that case, there would be an explanation for all contingent things without a necessary thing, and so premise (2) would be false.

Second, according to a standard philosophical view, sets are explained by their members. The set {1, 2} exists because 1 and 2 each exist. So, perhaps arbitrary groups of things are explained by their members. For example, perhaps the existence of the group containing the towel hanging in my bathroom and the number 3 are explained by the fact that the towel and 3 each exist. And, in that case, we’d again have an explanation for all of the sets of contingent things in terms of their members without needing to invoke a necessary thing. So, premise (2) would be false.

So, on either of the two ways that any arbitrary grouping could have an explanation, premise (2) is false. There are, of course, other cosmological arguments that do not depend on the idea that any arbitrary grouping has an explanation. And so perhaps Cameron can modify his argument so that completely arbitrary groupings don’t have explanations. Nonetheless, I think there’s a stronger argument that premises (1) and (2) are not jointly consistent.

Premises (1) and (2) jointly entail that there is a necessarily existent part of reality and that this necessarily part of reality explains all of the contingent things. Philosophers standardly distinguish between explanans — roughly, the premises in an explanation — and the explanandum — roughly, whatever is being explained. There are two possibilities: either explanations entail the explanandum or they do not. As I will now show, either possibility, together with (1) and (2), entails a contradiction.

First, suppose that explanations do entail the explanandum. Any consistent collection of statements, all of which are necessarily true, have only necessarily true entailments. And now consider an explanation of all of the contingent things in terms of the necessarily existent part of reality. In any such explanation, the explanans would be necessarily true and the explanandum would be contingent. Yet given that the explanans are necessarily true, the explanandum is necessarily true. Hence, the explanandum is both contingent and not contingent. Contradiction.

Second, suppose that explanations need not entail the explanandum. In that case, the necessarily existent part of reality can explain, without entailing, the conjunction of all of the contingent truths. For example, perhaps — as Cameron ultimately intends — God is the necessarily existent part of reality. In that case, God might have libertarian free-will and exercise God’s libertarian free-will to create the universe. The universe could then be a contingently existent thing whose existence is explained by a necessarily existent thing. Nonetheless, it would be false that God would provide an explanation for all of the contingent things. The fact that God created this universe as opposed to some other would be a contingent fact without an explanation. So, premises (1) and (2) would entail both that some necessarily existent part of reality explains all of the contingent facts, yet would also entail that some contingent fact does not have an explanation. Contradiction.

Consequently, premises (1) and (2) entail a contradiction. For that reason, I think we have good reason to reject Cameron’s contingency argument.

Disclaimer: The posts on the Cosmotherium should never be taken as definitive and I am typically not completely convinced of what I post here. This is my place for working out my views without the pressure or rigor of publication.

References

T. Ryan Byerly, 2019, “From a necessary being to a perfect being”, Analysis 79(1), pp. 10-17.

Thomas V. Morris, 1991, Our Idea of God: An Introduction to Philosophical Theology. University of Notre Dame Press.

Justin Mooney, 2019, “From a cosmic fine-tuner to a perfect being”, Analysis 79(3), pp. 449-452.