The Collapse of the Anti-Modal Collapse Objections

By Shannon Eugene Byrd


July 11, 2017


The doctrine of divine simplicity (DDS) has come under fire in recent years and shows no sign of cooling. Currently, one of the most interesting arguments against this doctrine is what has been referred to as the “modal collapse.” In this paper, I demonstrate that many of the objections raised against the argument are confused, logically contradictory, and inconsistent. I conclude that the modal collapse argument stands firm against these objections.


There has been a lot of discussion about the coherence of the doctrine of Divine Simplicity (DDS) in recent years. William Lane Craig and J. P. Moreland alleged that according to the doctrine, “everything about him (God) is essential to him” and in that case, “all modal distinctions collapse and everything becomes necessary.”[1] R. T. Mullins asks, “Does divine simplicity entail a modal collapse?”[2] He then answers, “Yes.” One might ask, “what does a modal collapse entail?” According to Moreland and Craig it entails “everything that happens does so, not with temporal necessity, but with logical necessity.”[3] Similarly, Mullins says, “Contingency and freedom completely vanish. Everything becomes necessary, so there is no such thing as free will. Second, it denied that God is free over creation. He cannot exist without it.”[4]

Recently, there has been an allegation that the modal collapse argument of Mullins is invalid. I will examine this allegation and demonstrate that on its face, the critique is patently false and that Mullins’ version of the modal collapse argument is not only valid, but also sound.

The Modal Collapse Argument

It is a fundamental claim of DDS that God is identical to his essence, will, knowledge, and so on. Divine Simplicity is perhaps Thomas Aquinas’ most controversial teaching.

For there is neither composition of quantitative parts in God, since He is not a body; nor composition of matter and form; nor does His nature differ from His “suppositum”; nor His essence from His existence; neither is there in Him composition of genus and difference, nor of subject and accident. Therefore, it is clear that God is nowise composite, but is altogether simple.[5]

Eleonore Stump says in order to maintain DDS several things need to be denied of God:

  • God cannot have any spatial or temporal parts.
  • God cannot have any intrinsic accidental properties.
  • There cannot be any real distinction between one essential property and another in God’s nature.
  • There cannot be a real distinction between essence and existence in God.[6]

Edward Feser elaborates on simplicity:

For Aquinas, God is “simple” in the sense of being in no way composed of parts (ST I.3). As has been said, he is incorporeal and immaterial, and thus cannot have any bodily parts nor be composed of form and matter. But neither does he have even any metaphysical parts. . . One famous implication of this doctrine is that though we distinguish in thought between God’s eternity, power, goodness, intellect, will, and so forth, in God himself there is no distinction between any of the divine attributes. God’s eternity is his power, which is his goodness, which is his intellect, which is his will, and so on. Indeed, God himself just is his power, his goodness, and so on, just as he just is his existence, and just is his essence.[7]

James Dolezal on God’s properties:

In addition to maintaining the real identity between God’s essence and existence, the traditional DDS also holds that all of God’s attributes are really identical in him. If God were a complex of really distinct attributes or properties then those various attributes would be more basic than the Godhead itself in explaining or accounting for what God is.[8]

DDS supporters often insist (even though they speak as if God has distinct properties), “the referent, God, does not possess them as really distinct properties.” Below you will see that modal collapse arguments focus greatly on the claim that God cannot have ontologically distinct properties. Generally the critic of simplicity thinks something along this line: If God doesn’t have distinct properties, and these “properties” are identical to God’s essence, then every property predicated of God is necessary as it is identical to God’s essence, which is necessary.

Before looking at Mullins version of the argument, recall that J.P. Moreland and William Lane Craig offered one:

. . . if God is identical with his essence, then God cannot know or do anything different from what he knows and does. He can have no contingent knowledge or action, for everything about him is essential to him. But in that case all modal distinctions collapse and everything becomes necessary. Since God knows that p is logically equivalent to p is true, the necessity of the former entails the necessity of the latter.[9]

Moreland and Craig’s version demonstrates that if God’s knowledge is identical to his essence, then his knowledge is purely necessary knowledge; there is no real ontological distinction in his knowledge. So everything God knows, he knows necessarily, thus, there is no contingency in his knowledge, if there were, God’s knowledge couldn’t be identical with God’s essence, because God’s essence is necessary.

The law of noncontradiction says that p and ~p cannot be identical; the two clearly contradict one another. If something is necessary, it exists in all possible worlds, whereas something contingent does not exist in all possible worlds. This signifies there is a world where p exists and ~p does not exist. Thus, they are not identical.

With all this in mind, lets look at Mullins argument and the critiques lodged against it.

“On divine simplicity God’s essence is identical to his existence. Also, God’s one simple act is identical to His essence/existence. God’s act of creation is identical to this one simple act, and so identical to God’s essence/existence. God exists of absolute necessity. So His act of creation is of absolute necessity since it is identical to His essence/existence.”[10]

If successful this argument demonstrates that divine simplicity does in fact lead to modal collapse, everything is absolutely necessary.

A recent objection to this argument has been over the supposed form of the argument:

Fortunately for the proponent of DDS, the argument from modal collapse is facially invalid, and indeed commits a fallacy that has been well-known since at least Quine (1953). It substitutes “God’s act of creation” for “God” into a modal context (within the scope of a necessity operator, to be exact), but as Quine teaches us, modal contexts are referentially opaque, which means that substitution into them does not generally preserve the truth of the sentence into which such a substitution has been made. To use Quine’s own well-known example, while it is necessarily true that 8 is greater than 7, and it is true that the number of planets = 8, it is false that the number of planets is necessarily greater than 7. So this argument would be invalid:

4. ◻︎(8 > 7)
5. Number of the planets = 8
6. ◻︎(Number of the planets > 7)

But note that this is precisely the same form as that of the argument from modal collapse![11]

Substitution in a modal context:

Firstly, Tomaszewski’s criticism seems muddled. It says that one cannot substitute in a modal context within the scope of a necessity operator, yet, the very premise in question, has no necessity operator in the translation of the argument. On its face, this is just a mischaracterization of Mullins’ modal collapse argument. That is to say, it is a straw man and an uncharitable translation of the argument. Additionally, if the objection is that one cannot substitute in a modal context for the reason that they do not generally preserve the truth value of the sentence the substitution has been made, then Tomaszewski’s argument is itself a non-sequitur. It is as though he is arguing the following:

1- Truth is not always preserved in modal contexts within the scope of a necessity operator.

2- The Modal Collapse Argument is given in a modal context and within the scope of a necessity operator.


3- The Modal collapse argument does not preserve the truth.

The conclusion doesn’t follow from the premises. If there are modal contexts in which truth is not preserved within the scope of a necessity operator, it doesn’t follow that this is always the case. As long as there is at least one truth statement preserved within the scope of a necessity operator, the criticism fails.

With the above in mind, consider the following argument:

1- Necessarily, 2 + 4 = 6

2- Necessarily, 5 + 1 = 6

3- Necessarily, 5 + 1 = 2 + 4

Now, is Tomaszewski or any defender of DDS ready to reject this argument as well? I would hope not! But, if one did, it would likely be due to a commitment that there are no a priori truths, and if that is the case, then one is holding to a self defeating concept, because the assertion that there are no a priori truths, is in itself an a priori truth claim, and if true, it is self-refuting, but if false, it is just plainly false and there is in fact a priori truths.

In my analogous argument above, note that the necessity operator is in every premise, just as the modal collapse argument, though it isn’t explicitly stated in every premise in Mullins’ version, it is obviously implied.[12] All three premises are a priori and substitute in a modal context, that is to say, they are analytical and do not require experiential knowledge.

All designators (terms that flank the identity symbol or “is”) are what Saul Kripke termed “rigid designators” that is, in every possible world, they refer to the same object in which it exists and in the parody argument first given, several mistakes are made on this point alone.[13] First, in the premise “8 > 7,” both terms are at minimum rigid designators, but in the second premise, “Number of planets = 8” only one designator “8” is rigid, whereas, “number of planets” is what is termed an accidental or non-rigid designator, that is to say, it does not refer to the same thing across all possible worlds, there is a possible world with a different number of planets, or no planets at all, but there is not a possible world where 8 is not 8.

A rigid designator cannot be identical to an accidental designator, one refers to the same object in all possible worlds in which it exists and the other does not. Trying to say the two are identical is a violation the law of noncontradiction, P is not ~P at the same time and in the same way. In order for both terms to co-refer, they have to share the same properties and be used the same way, in this case, they do not. What this seems to indicate is “Number of planets = 8” is either necessarily false due to their divergent properties (mixing necessity and contingency), or the parody is equivocating on the term “is.” Rather than being an “is” of identity, it is masquerading as one and in reality it’s an “is” of predication. A little thought demonstrates this: the collection of planets instantiates the number 8, 8 is not identical to the number of planets, necessarily, 8 = 8, but the number of planets is contingently 8. Thus, it seems fairly clear that the parody indeed equivocates an is of predication for an is of identity, that is to say, the parody is logically fallacious.

Second, the parody isn’t consistent with regard to giving a priori truth claims. “8 > 7” is an analytic or a priori truth, “8 = number of planets” is an a posteriori truth claim, no one can know this conceptually, it takes empirical evidence such as looking through a telescope to substantiate this. In Mullins argument, “◻︎God = Act of creation,” both designators that flank the identity sign are strongly rigid designators, which is to say they are both necessarily existent. Moreover, they are a priori truths, and not a posteriori truths. One cannot know from experience that, “God’s act is identical to his act of creation.” So again, Tomaszewski’s parody fails, and this indicates that he has failed to understand the subtle nuances within modal contexts and the related epistemological status of the designators. I think he also fails to understand the distinction between de re and de facto modalities. De re modalities assign a property to a particular designator and thus are open to substitution. De facto modalities assign a property to the entire proposition, on this last note we’ll see below, he thinks the entire identity statement “God = Act of creation” is a contingent proposition. But first we’ll look at Tomaszewski’s accusation that the modal collapse argument begs the question.

Does the Modal Collapse Beg the Question?

As with the other critiques, the issue is with the premise, “◻︎God = Act of creation.” Christopher Tomaszewski argued that it “overtly begs the question” in assuming God by necessity must create.[14] Keep in mind my analogous argument from the above, does it beg the question? I think not. Moreover, “◻︎God = Act of creation” if one designator is strongly rigid then both designators must be strongly rigid designators in order for the identity statement to be true. One isn’t presenting them both as strongly rigid just to make the argument work; rather, one is presenting them this way because DDS posits this. On this note, Randy Everist states that DDS supporters,

“have to realize the implications of stating that it is only a contingent matter that God is identical with his act of creation. This means in some worlds God is not identical with his act of creation. Such a view is not only problematic for identity, but also seems to introduce potentiality in God, the very thing they wish to reject.”

Ryan Mullins’ argument does not assume the truth of what he is trying to prove; it isn’t as if he is arguing, “creation exists necessarily” because “necessarily creation exists,” no, Mullins is arguing that God exists necessarily and his act of creation is identical to God’s existence and so his act of creation is necessary as well.

Edward Feser, a DDS supporter, notes:

Indeed, any identity statement in which the identity sign is flanked by rigid designators is, if true at all, going to be true in every possible world and thus a necessary truth. Such statements, when true, are thus taken to tell us about the essential properties of the things they refer to. . . On the other hand, if an identity statement in which the identity sign is flanked by rigid designators is false in any possible world, then (since again, such statements must be true in every possible world if true at all) it is not a necessary truth and so not true at all. Hence it is false in the actual world.[15]

Again, something contingent cannot be identical to something else that is absolutely necessary, that is a logical contradiction. In Tomaszewski’s paper, it is at this exact point where an appeal to mystery is made and this will be demonstrated below.

‘God’s act of creation’ designates that act, not how it is in itself, but by way of its contingent effects. That is, whether ‘God’s act of creation’ designates God’s act depends on the existence of a creation which is contingent, and so the designation is not rigid.[16]

To explain this plainly, his position is, “God’s act of creation” is a contingent truth that isn’t rigid and thus will not co-refer to “God’s act” across every possible world. There is a world or many worlds where God’s act, which is absolutely necessary (strongly rigid), is not “God’s act of creation,” which is contingent and thus accidentally rigid (at least according to his view).[17] At first, this may seem like a valid point, but I think it’s in fact invalid. Again, something contingent cannot possibly be identical to something absolutely necessary, the two are logically contradictory (P ≠ ~P). So, rather than the entire identity statement not being necessary (contingent) as he claims, in reality, it is necessarily false, there is no possible world in which a necessarily existent object is identical to a contingent object. If one of the designators is strongly rigid, then both have to be strongly rigid in order for the identity statement to be true.

Also, if his view is that there is a world where “God’s act” is identical to “God’s act of creation” as he seems to indicate by supposing the identity statement as a whole is contingent, then he is holding to some form of contingent identity, instead of this sort of identity statement being a necessary relation.

I suppose another possibility might be that the is, is an is of predication instead of an is of identity. If that is the case, then it would strongly imply “God’s act” instantiates “God’s act of creation” and is not in reality ontologically identical to it. Note also, that on his understanding that in order for “God’s act of creation” to co-refer to “God’s act” it is dependent on creation itself. This is incomprehensible, how would this not introduce contingency in God? If “God’s act,” which is absolutely necessary and thus a strongly rigid designator is identical to his essence and also identical with a contingent fact (“God’s act of creation”), then due to the transivity of identity, this would indicate that God’s essence, is contingent as well. I am sure any adherent to DDS would rightly protest that God’s essence is not contingent and it is at this point where mystery comes into play.

It should, of course, be admitted here that there is mystery in how the effects of God’s act can vary without His act varying. But mystery is not incoherence, and mystery doesn’t entail a modal collapse.

The oxford dictionary defines a mystery as: “Something that is difficult or impossible to understand or explain.” I posit what Tomaszewski offered here as a “mystery” is not the case. If “God’s act” is necessary, and as he assumes “God’s act of creation” is contingent and in some possible world both happen to co-refer, then he is positing a logical contradiction as opposed to a mystery. We can precisely understand that P ≠ ~P. Equally as bad, this seems like an outright taxicab fallacy in which one “rides” the logic until just before the logic reaches an unwanted conclusion/destination. Additionally, the notion that, “God’s act” is the same across all possible worlds and yet the effects are different is a strange notion. Obviously the reason cannot be found in the contingent effects themselves because the reason has to be explanatorily prior, yet, it cannot be in God, because “God’s act” is the same across all possible worlds.[18]

It must also be said, that he refers to the “conceptual resources available to the defender of DDS” in making the distinction between “God’s act of creation” and its effects. It must be said that conceptual distinctions do not rescue the defender of DDS from the incoherence of positing something contingent as identical with something else absolutely necessary. Conceptual distinctions are in the mind only and bear no resemblance to the ontological status in question, that is to say, they aren’t real in any sense of the word and appealing to them in any attempt to defend DDS doesn’t help. Thus, it seems clear to me that the modal collapse argument stands firm.


The first assertion was over the alleged fallacious form of the argument. As I demonstrated above, the parody used by the Tomaszewski was an uncharitable one; not all premises contained a necessity operator as Mullins argument does. So on its face, this accusation is factually incorrect.

Tomaszewski also asserted that the modal collapse argument commits a fallacy of substituting “God” and “Act of creation” in a modal context. As I showed in the above, substitution is allowed in modal contexts so long as both designators share the same properties and one doesn’t intermingle rigid and accidental designators with one another.

Tomaszewski’s parody argument also equivocates on the term “is,” rather than using an “is” of identity in the second premise, he uses an “is” of predication. So his parody itself is fallacious in a way he did not take note of.

He also alleged that if the modal collapse argument possessed the necessity operator in “God = Act of creation” then it plainly begs the question. It’s clear to see that the argument does not argue this way. It is due to God’s essence being identical to God’s act of creation, that creation is necessary, otherwise the two cannot be identical; something contingent cannot be identical to something else absolutely necessary, positing such is nonsensical.

Lastly, Tomaszewski posited that God’s act is the same across possible worlds and yet its effects can mysteriously differ in these worlds. I concluded that he is trying to masquerade a logical contradiction as a mystery, something contingent (creation) cannot be identical to something necessary (God’s act). Given the objections thus far, the modal collapse argument stands firm.[19]


1. J. P. Moreland & William Lane Craig, Philosophical Foundations for a Christian Worldview. (Downers Grove: Intervarsity Press, 2003), 525.
2.  R. T. Mullins, The End of the Timeless God, (Oxford: Oxford University Press, 2016), 138.
3.  Moreland & Craig, Ibid.
4.  Mullins, Ibid.
5.  Thomas Aquinas. Summa Theologicae I.3.7.
6.  Eleonore Stump, “Simplicity” in A Companion to Philosophy of Religion. (Malden: Blackwell Publishers, 2000), 250.
7.  Feser, Edward. Aquinas: A Beginner’s Guide (Beginner’s Guides) (Kindle Locations 2073-2076). Oneworld Publications. Kindle Edition.
8.  James Dolezal, God Without Parts: Divine Simplicity and the Metaphysics of God’s Absoluteness. (Eugene: Pickwick Publications, 2011), 125.
9.  Moreland & Craig, 525.
10.  Mullins, 138. 
11.  Christopher M. P. Tomaszewski, “Collapsing the Modal Collapse Argument” . 5., Retrieved from: 
12.  The necessity operator is transferred to the other identity claims due to the transivity of identity.
13.  Saul Kripke, Naming and Necessity. (Malden: Blackwell Publishing, 1981), 3-4.
14.  Tomaszewski, 7.
15.  Edward Feser, Scholastic Metaphysics: A Contemporary Introduction (Editiones Scholasticae, 2014), 238. ; Feser’s example of a necessarily true identity statement is “Water = H2O”
16.  Tomaszewski, 7. 
17. I think the Tomaszewski misunderstands necessity and contingency when it comes to which he refers to strongly rigid, rigid, and accidentally or non-rigid designators. A rigid designator is rigid because it refers to the same object in all possible worlds, but that isn’t to say it exists in every possible world, just that it refers to that object it designates in every world in which it exists. Here the he thinks that a contingent entity is not rigid and this is incorrect.
18. Moreland & Craig, 525.
19. Special thanks to Randy Everist for reviewing this article and suggesting its title. I appreciate all you do brother!
Tagged with:

About the Author

By Shannon Eugene Byrd