Response to Hsiao and Sanders on the Thomistic Contingency Argument, Part 2

“Okay, so…”

Graham Oppy, Monash University

So, today I continue with my appraisal of Hsiao’s and Sanders’ (henceforth, ‘H&S’) argument presented in this paper.

Half the matter in the universe was missing – we found it hiding in the  cosmos

0 Outline

So, here’s an outline of this post:

1 Quantifier shift
2 Necessity and eternality
3 Necessity and immutability
4 Uniqueness
5 Immateriality
6 Mindedness
7 Omniscience
8 Freedom of Will
9 Omnipotence
10 Perfect
11 Perfectly good
12 Simplicity and oneness

13 Conclusion

1 Quantifier shift

“From the previous premises it follows that a chain of sustained beings must terminate in an
independent being that is not itself sustained. We are left with an ultimate sustainer of all dependent beings. In the words of St. Paul, it is a being in which ‘we live and move and have our being.'”

But this is just a non-sequitur. All H&S have argued thus far is that for each chain of dependent beings, there is some first, independent member that imparts the relevant causal power or property to said chain. But it simply doesn’t follow that there is one first, independent member that imparts the relevant causal power or property to all chains. This is simply a quantifier shift: merely from the fact that for each Nobel laureate, there is some university with which they’re affiliated, it does not follow that there is some one university with which all Nobel laureates are affiliated.

(Indeed, we haven’t even concluded to a single independent being that imparts the causal power/property to a single chain. For there may be any number of of independent beings that jointly impart such power. The important point for now is that nothing H&S say justifies the inference to a single independent being that is the source of all else. They say it follows from their preceding premises, but that is a straightforwardly false claim.)

(But wouldn’t it be simpler to have one independent being rather than many? I address this in this video. In short, this response fails to take into account the distinction between quantitative simplicity and qualitative simplicity. Also: H&S claim that a single independent being follows from their premises. They did not say it follows from the premises in conjunction with a defeasible application of Ockham’s Razor.)

2 Necessity and eternality

“Something that is necessary must be eternal because there could never be a time at which it did not exist.”

I’ll let this slide, because I recognize that it isn’t implausible. So, I’ll be short in my response:

This claim is technically false. Necessity only entails existing or obtaining in all possible worlds; this neither means nor entails existing at all times in all possible worlds. (Note that I’m also not just making this up; this is a standard distinction in debates concerning necessity and eternality.)

3 Necessity and immutability

H&S then say:

“Something necessary must also be immutable because if it could change then its nature would cease to exist and become something else.”

But this is false. It’s simply wrong to say that “if it could change then its nature would cease to exist”. Consider that things change all the time without changing their nature: I can get a haircut, but my nature doesn’t cease to exist upon getting a haircut. This is because my hair length is not essential to me. Thus, it is simply false to say that if x could change, then x’s nature would cease to exist. This presupposes that the change in question would change something essential to the thing in question. But it could easily merely change in a non-essential way. Thus, H&S’s claim is wrong.

See this post for more details. Necessity doesn’t in any way, shape, or form entail immutability.

4 Uniqueness

H&S then say:

“We also know that it causes and sustains all dependent beings in existence. This means it is the
Uncaused Cause, the Creator, or the First Cause of all reality.”

But I’ve already shown why this is wrong. See Section 1.

5 Immateriality

H&S write: “The first problem with [the suggestion that the being in question is material] is that science tells us all matter began to exist. So this eternal being cannot be material.”

First, this inference doesn’t work, since H&S haven’t successfully inferred eternality. (Indeed, if matter begins to exist in every possible world, then it simply follows that matter necessarily exists. And so a beginning of existence is no mark against necessity.)

Second, science does not tell us this. At best, science only tells us that our local spacetime has an earliest temporal boundary. This is perfectly compatible with there being a non-temporal, causally initial segment of reality that is physical as adumbrated by numerous authors in the literature. (For a brief adumbration, see Oppy’s remarks about this in here.) It’s also compatible with time being non-fundamental and either realized by or emergent from fundamentally non-temporal physical things or stuff. Philosopher Daniel Linford has done work on this proposal in the context of the Kalam and the philosophy of physics. It’s also compatible with other spatiotemporal universes–perhaps abiding by their own physical laws, not the ones of our universe–pre-existing the earliest temporal boundary of our local spatiotemporal reality. And so on. The point is that science does not tell us that matter began to exist.

(I also say “at best”, because from what I’ve researched, we would need a theory of quantum gravity in order to determine what happened prior to the Planck era of the early period of universe expansion. But I’m not a physicist, so I have no qualifications to speak further on this matter.)

H&S also say:

“The second problem with this is that matter seems to be inherently dependent. That quantum field, simple particle/s, or even that universe making world could fail to exist. There is nothing in their natures that requires their existence.”

This is the very question at issue, though. This has merely asserted that they can fail to exist, and that they aren’t necessary.

Now, they attempt to provide independent justification for such claims, but the justification doesn’t work. Here’s what they say:

“I can conceive of their non-existence whereas I cannot conceive of necessary truths like 1 + 1 = 2 being false.”

But this is simply unconvincing. I can conceive of the non-existence of the God of classical theism. Does it somehow follow that the God of classical theism is possibly non-existent? It doesn’t. But yet this is precisely the inference that H&S need for their appeal to conceivability to work. (Alternatively, if they only mean for the appeal to conceivability to defeasibly justify the possibility claim, then this equally defeasibly justifies the denial of classical theism, and hence the symmetry is restored between our appeals to conceivability.)

Indeed, their appeal to conceivability undercuts their own argument here. For they either think conceivability entails possibility, or else only gives defeasible justification for it.

If they think conceivability entails possibility, then since I can easily conceive of the necessary existence of some foundational quantum field, say–and since the possible necessity of x entails the necessity of x (per system S5)–it follows that some foundational quantum field is, after all, necessary. And this contradicts their claim. So, they cannot say that conceivability entails possibility.

But if they only think that conceivability gives defeasible justification for possibility, then our defeasible justifications cancel out. This is because, again, I can conceive of the necessary existence of some foundational quantum field. And hence (by S5) I have defeasible justification for accepting the necessary existence of some foundational quantum field. But I can also–following H&S–conceive of the non-existence of such an entity, and hence I have defeasible justification for thinking that accepting its non-necessary existence. But notice that the appeals to conceivability here simply cancel one another out. The appeals are awash. And hence the appeal to conceivability doesn’t justify H&S’s claim over and above the denial of their claim. And hence their appeal to conceivability fails.

(In fact, there might even be an asymmetry in favor of my appeal (i.e. against theirs), since absences are notoriously hard to conceive. Cf. Pruss (2006; 2009).)

Note that I am not here claiming that some foundational physical thing exists of metaphysical necessity. I’m only addressing the unsuccessful arguments against this position by H&S.

H&S then argue:

“Furthermore, any simple particle/s could be replaced by other simple particle/s. There’s
nothing necessary about these simple particle/s being fundamental to all reality rather than some
other simple particle/s.”

This is the very question at issue, though. It’s the very question at issue whether such particles or fields or whatever can be replaced. This has merely asserted that they can be replaced, and hence that they aren’t necessary.

H&S then write:

“This leads us to a related problem: Suppose only one simple particle, field, or world was
necessary. Why just one? Why not two, seven, a hundred, or even zero? Picking any number as
necessary is as arbitrary as flipping a coin. Necessity cannot be arbitrary or else anything could be

First, the same problem will apply to their views. And a problem for everyone [in a dialectical context] is a problem for no one. Consider: “Suppose only one God is necessary. Why just one? Why not two, seven, a hundred, or even zero? Picking any number as necessary is as arbitrary as flipping a coin. Necessity cannot be arbitrary or else anything could be necessary.” Or, more poignantly: “Suppose only one three divine persons are necessary. Why exactly three? Why not two, four, one, a hundred, or even zero? Picking any number as necessary is as arbitrary as flipping a coin. Necessity cannot be arbitrary or else anything could be necessary.” [For some of my musings on Trinitarianism in the context of classical theism, see here.]

Second, there need be nothing arbitrary here. We have sophisticated, worked-out theories of intrinsic probability (cf. Draper and Swinburne’s work). And we can use such theories to give principled reasons for accepting one account of the number of foundational things as opposed to all other accounts. (E.g. one necessary foundational field will be more probable than other numbers of such fields.)

Then H&S say:

“We run into this problem of arbitrariness because all matter is multipliable or reducible
in principle.”

The very question at issue is whether all matter is multipliable or reducible in principle. In particular, the very question at issue is whether there could be more or less of the foundational things. One cannot merely assert that there could be more or less when the very proposal to be refuted is that it is necessary that there be this many foundational items.

6 Mindedness

H&S write:

“Now that we know that this necessary being cannot be material, we know by process of
elimination that it must be an immaterial being. This leaves us with only two possible candidates:
Either it is an abstract object or it is a mind.” They then argue that abstract objects are causally effete, and hence the entity in question must be a mind.

First, I’ve already argued that their inference to immateriality fails.

But, second, they have merely asserted without any justification that the only two possible candidates for x’s being immaterial is that x is either a mind or an abstract object. I address this claim further in this video. For now, I flag that this is a mere assertion. Moreover, I’ll simply connect this to what Feser says in his Neo-Platonic proof.

Feser argues in premise twenty-two of the Neo-Platonic proof that “[e]verything is either a mind, or a mental content, or a material entity, or an abstract entity” (2017, p. 81). From this it is concluded that the absolutely simple being must be a mind.

Curiously, though, Feser gives no justification for this premise but instead cites Vallicella (2002, p. 255)—who himself gives no justification for the claim. Indeed, it seems eminently possible at least in principle that there be a non-physical or non-material, non-mental concrete entity. Simply consider an impersonal principle from which all complexity and multiplicity derives—akin to (some understandings of) Brahman or to Plotinus’ One (which, as we will see below, is prior in being to Mind). In fact, we might plausibly take the Neo-Platonic proof (ignoring its other worries) as an argument for this different category of thing.

Plotinus himself decidedly rejected a view according to which the absolutely simple, radically independent One (which transcends all multiplicity, qualification, and differentiation) is mental. As Gavrilyuk notes, for Plotinus “the divine Mind (nous), as the repository of the eternal Forms, represents a perfectly unified plurality, rather than perfect simplicity. For this reason, the divine Nous must be the second hypostasis, which derives from and is ontologically subordinate to the One” (2019, p. 442).

7 Omniscience

They write:

“First, it is evident that in order to create anything, this mind must have some knowledge of what it creates.”

First, we’ve already seen why their inference to mindedness fails.

Second, this isn’t evident in the slightest. No justification is given for why (A) entails (B):

(A) x is a mind, and x causes y to exist

(B) x has some knowledge of y

This inference is NOT evident. Indeed, it’s a straightforward non-sequitur: merely from the fact that x is a mind and that x causes y to exist, nothing follows about x’s cognitive grasp of y. [Consider: I am (or have) a mind; I cause whole hosts of things; and yet it’s simply false that I thereby have some knowledge of them.]

“Not only does it have knowledge, it must have perfect knowledge. If it could learn, this being
would undergo change, which we know to be impossible. Hence this being must be omniscient.”

This is a straightforward non-sequitur on so many levels.

First, as I showed above, the inference that it has knowledge failed.

Second, as I showed above, the inference to the being’s immutability failed.

Third, this presupposes that changing in knowledge is incompatible with having perfect knowledge. But no justification is given for this presupposition. Indeed, it seems obviously false: God is not more or less perfect because he goes from not knowing that humans exist to knowing that humans exist (once humans came into existence). In this change, God has simply retained his value, his perfection. He has moved horizontally along the axis of perfection and value, not vertically (getting better or worse).

Fourth, even if it cannot change in knowledge, this doesn’t mean that its knowledge is perfect or all-encompassing. H&S have only argued that the being has some knowledge. [And even that argument failed.] They haven’t shown–anywhere–that the being’s knowledge is all-encompassing, i.e. embracing all truths and existent things. Indeed, we saw that their inference to the singularity or uniqueness of the being failed, and hence they cannot infer omniscience from the (unestablished claims) that (i) this being is the cause of all else and (ii) if x is a mind and is a cause of y, then x knows y.

Fifth, if the being in question cannot gain new knowledge–as H&S seem explicitly to endorse–then it seems that they’re gonna have to say goodbye to any non-eternalist view of time. In other words, it seems they’re gonna have to adopt eternalism. See my points here, the supplemental points here, and the popular exposition of the argument here.]

They then claim:

“If this being were not omniscient, there would be nothing to ground the fact that it is possible for things like unicorns to exist.”

This is incorrect. The possibility of unicorns could easily be grounded merely in the causal power of the being in question, or in the causal power of other beings. Or, it might be grounded in the knowledge of some other independent being (remember, their inference to uniqueness didn’t work). In fact, surely this being’s knowledge doesn’t ground the possibility of unicorns; surely it’s the other way around. The being can only know unicorns are possible if they are, in fact, possible to begin with!

They then claim:

“Possibility doesn’t exist out there as ‘stuff’ because all possibilities are abstractions; they must
exist in a mind. If it doesn’t exist in this mind, then it is impossible.”

First, this is a mere assertion. They did not present any justification for this claim.

Second, it seems a bit imprecise. What is a “possibility”? What is it for a possibility “to exist”? Is this not for a possibility just to be actual? Moreover, the Doctrine of Divine Simplicity (DDS) is explicit that anything in God is numerically identical with God. But in that case, since H&S say that all possibilities are in God, and since DDS explicitly says that anything in God is identical with God, it follows that all possibilities are identical with one another and with God. But this is (i) clearly absurd [since not all possibilities are identical; e.g. some are about unicorns, but some aren’t; and hence they aren’t identical], and (ii) it entails the absurdity that there’s only one possibility. Hence, either DDS is false, or else H&S’s claim above is false.

They then claim:

“So all possibilities must be known by this mind, which further confirms that this being must be omniscient.”

This is a non-sequitur, on many levels.

First, merely from the fact that a possibility exists in the mind, it doesn’t follow that the possibility is known by the mind. Nowhere do H&S justify this claim. (Indeed, it’s simply a non-sequitur: lots of things exist in my mind which are such that I don’t know them, such as various hidden essential properties of my mind of which I’m unaware, and other things of which I am aware but nevertheless don’t know.)

Second, even assuming that the being knows all possibilities, this is categorically insufficient for omniscience. For this says absolutely nothing about whether this being knows contingent actualities. Mere knowledge of possibilities is utterly insufficient for such knowledge of contingently obtaining actualities.

Third, once again, we have not concluded that all possibilities exist in this entity’s mind, for starters because their inference to uniqueness failed (and hence they haven’t even arrived at a single source of all which must pre-contain all possibilities).

8 Freedom of Will

They write:

“Second, it must have a will because out of all the possible beings it could create, there is nothing about the nature of any possible being that required their creation.”

This is a non-sequitur. All that follows is that the being indeterministically causes its effects. Nothing H&S says justifies the inference that the kind of indeterministic causation at play here is an intentional one under an agent’s control.

They also write:

“If instead we suppose that this mind’s nature compels it to create these possible beings from eternity, then its existence would depend on creating these possible beings to be what it is, which makes it dependent rather than independent.”

This is a non-sequitur. It doesn’t follow from the fact that x necessarily generates or causes or creates y that x is somehow ontologically dependent on y. In fact, the opposite is true: precisely because x is the cause or ground of y, it follows that y is ontologically dependent on x. And since ontological dependence is asymmetric, it follows that x is not ontologically dependent on y. (And recall that the relevant kind of dependence at play in H&S’s argument is, indeed, ontological dependence and independence, i.e. not depending on another to receive being or actuality).

9 Omnipotence

They write:

“Now if it has power, then it cannot change in its power. … An immutable being cannot lose or gain power whenever it creates because that would involve a change.”

This is wrong, since we have already seen that their inference to immutability fails.

They write:

“Something that can create vast worlds like ours without losing power must have unlimited power.”

This is wrong, since nowhere have they shown that this is the single independent source of all else. Moreover, even granting that this being is the single source of our cosmos, it is just a non-sequitur: all we can infer is that the being merely has the power to create this cosmos. We might inductively infer that it has more power. That’s fine. But that isn’t what H&S claimed; they claimed it must have unlimited power. This is false.

They then write:

“Another reason to think this is if it could only create a limited set of possible beings, the other set of possible beings could not be created. But if they cannot be created by the cause of all being then they are not possible beings.”

But this is wrong, since it presupposes that all possible beings could only be grounded in the power of this one, single unique being. But we’ve already seen why their inference to this claim fails.

10 Perfect

“Indeed, this being must have all perfections. Since its nature is to exist, it cannot possibly
be lacking in anything, since to lack something is to lack existence of some kind.”

So, I don’t know what ‘existence of some kind’ means, but I’ll set this aside. Their claim here is mistaken nonetheless. It is mistaken because nowhere did H&S show that the independent being’s nature is to have existence of every kind. They only argued that it necessarily exists, that it is built into its nature to exist. But nowhere did they justify the claim that existence of every kind is built into its nature, or that it necessarily has existence of every kind.

Consider, moroever, extension (i.e. taking up space). Surely the being in question is lacking in extension; it is supposed to be, after all, non-spatial (according to H&S). Thus, it’s lacking in something, contrary to their claim.

Now, they may say that extension isn’t a positive reality that the being lacks but is instead some absence, lack, or imperfection of some kind. But, firstly, this assumption would require justification; and, secondly, it seems intuitively false. Extension surely isn’t a negative property in the sense of a lack or privation or absence. After all, it reports something’s positively occupying space. It attributes some positive reality to something.

11 Perfectly good

They write:

“Another reason for thinking it has perfect goodness requires understanding what goodness is. Whenever we say someone is a good person or that something is a good whatever, we always judge it to be good by how well it matches the nature of that activity or thing. A person is good if his character matches with what is good for human nature. A pianist is good if he matches the nature of what it is to play the piano well. A hamster is good as a hamster if it matches the nature of what it is to be a hamster with little to no defect. Something’s nature determines what is good for it. Since a necessary being cannot fail to match its nature perfectly, it must be perfectly good for that reason.”

Well, first, no justification is presented for this account of goodness. But we can let that pass, since this might plausibly be understood to be beyond the scope of their paper. (If so, they should at least specify that instead of simply asserting the nature of goodness. By my lights, at least.)

Secondly, and more importantly, is that this sense of ‘perfect goodness’ is, by my lights at least, wholly uninteresting. H&S’s inference mirrors that in Feser (2017). Feser writes that the “sense of ‘good’ and ‘bad’ operative here is the one that is operative when we speak of a good or a bad specimen, a good or bad instance of a kind of thing” (2017, p. 217). By itself, then, this mode of argumentation is completely devoid of normative, evaluative, or axiological import. But we normally wish to ascribe moral goodness to God. H&S’s argument, however, is categorically unable to establish moral goodness. Indeed, their account of goodness would entail that an exact, geometric circle (not the approximations we draw or print) would be ‘perfectly good’, since it has no privations, fully realizes its essence, and cannot fail to match its nature (circularity) perfectly and fully and wholly. This is a far cry from omnibenevolence or moral perfection.

12 Simplicity and oneness

They write:

“Lastly, this necessary being must be simple, and there can only be one such being in principle. By simple we just mean that it is not composed of any physical or metaphysical parts. If it had parts, then the whole would be dependent on its parts.”

This is a non-sequitur. It could easily be the case that the whole grounds its parts, and hence the parts are dependent on the whole, NOT vice versa. This has already been made rather forcefully in the literature. (Cf. Baddorf (2016)). I’ve also already addressed this in numerous videos, including this one and 1:38:56 and onward in this one.

They then try to argue for oneness:

“Furthermore, there can only be one necessary being. If there were multiple necessary beings, they would all have the same features. How could we differentiate between two identical necessary beings? What makes this necessary being distinct from this other necessary being? If we had two identical pencils, we can say that what makes them distinct is that they got this bit of matter located in one space and this other bit of matter located in another space. But we cannot appeal to matter because these necessary beings are immaterial. So there must be something immaterial that makes them distinct, which only leaves us with their properties. But it cannot be their properties either because they have identical properties. Every necessary being has to have the properties mentioned above. You would need to introduce a new property that one necessary being has that the other does not. But why introduce that property and why does it only apply to one but not the other? There’s no non-arbitrary way to answer this. Therefore, only one necessary being is possible.”

First, this assumes the inference to simplicity, which we’ve already seen fails.

Second, this is incompatible with Trinitarianism. Consider:

“Furthermore, there can only be one divine person. If there were multiple divine persons, they would all have the same features. How could we differentiate between two identical divine persons? What makes this divine person distinct from this other divine person? If we had two identical pencils, we can say that what makes them distinct is that they got this bit of matter located in one space and this other bit of matter located in another space. But we cannot appeal to matter because these divine persons are immaterial. So there must be something immaterial that makes them distinct, which only leaves us with their properties. But it cannot be their properties either because they have identical properties [’tis the nature of an absolutely simple being!]. Every divine person has to have the properties mentioned above [because divine persons are necessary and simple, and H&S explicitly claim that such properties follow upon necessity and simplicity]. You would need to introduce a new property that one divine person has that the other does not. But why introduce that property and why does it only apply to one but not the other? There’s no non-arbitrary way to answer this. Therefore, only one divine person is possible.”

I expand on this line of argument more in Section 4 of this post, for those interested. (Actually, the other sections of the post are relevant, too, come to think of it.)

13 Conclusion

H&S claim that they “have just proved that this independent being who causes and sustains all dependent beings must be necessary, eternal, immutable, immaterial, omniscient, omnipotent, free, perfect, simple, and one.”

As I’ve argued, however, this is very, very, very far from reality. Not only did they fail to make headway against existential inertia [cf. this post], but they also failed in each of their inferences to eternality, immutability, omniscience, omnipotence, freedom, perfection, simplicity, and oneness. I have argued that nearly every claim they make in Section V (“Premise Five”) of their paper is either false, or a non-sequitur, or merely asserted without justification.

While I’ve been critical of their essay, I appreciate their engagement with my scholarly work as well as our unity together in pursuit of Truth. I may have sounded harsh in some of my criticisms. If so, I apologize. I do all of this in the service of our collective pursuit of the Ultimate.

Take care y’all.

“Friends, I love you so. It’s a matter of time before we leave the ones we know.” (Source)

Author: Joe



  1. Hey Joe! Nice post. Concerning Hsiao, I was wondering if you had any opinion on the ethics he espouses and am eager to hear your assessment of the natural law approach to moral questions.

    • Thank you!!!

      Re: natural law theory, I had a discussion on my channel b/n Brian Besong and Dustin Crummett on it. I am sympathetic with some of Dustin’s criticisms.

      In particular, I find the following two claims to be almost crystal clear (i.e. almost self-evidently true): (1) it is morally permissible to lie about how many cookies I ate last night in order to avoid both myself and my family being tortured to death; and (2) it is morally permissible (perhaps even obligatory, though I won’t make this stronger claim) to lie to the Nazis at your door about the Jews you’re hiding in your basement.

      And yet the natural law theory defended by Besong, Skalko, Feser, Hsiao, and company entail that both (1) and (2) are false. (They themselves explicitly admit as much, from what I recall.) I recognize that justification is person-based. For me, then, I take this to be a reductio of their view. But I guess one’s modus tollens is another’s modus ponens.

      I’ve also been told that Hsiao defends factory farming? I sure hope that he doesn’t. (I’m not terribly acquainted with his work, since I don’t do much research in ethics)

  2. “I can conceive of the non-existence of the God of classical theism. Does it somehow follow that the God of classical theism is possibly non-existent?”

    You can conceive of the non-existence of existence itself? That’s impressive.

    Jk, I agree with most of your post. Well done.

Leave a Comment

Your email address will not be published. Required fields are marked *