How the Absolutely Simple Creator Escapes a Modal Collapse

Christopher Tomaszewski
Belmont Abbey College

Abstract

Classical theism’s commitment to the doctrine of Divine simplicity is alleged to result in a modal collapse. But a simple
argument aimed at demonstrating that the doctrine of Divine simplicity entails a modal collapse is invalid. Nevertheless, a
fundamental worry that motivated this argument remains: the doctrine of Divine simplicity does entail that God is really,
intrinsically the same across all possible worlds. So how is it possible for a Cause that is really, intrinsically the same across
all possible worlds to create all the creatures in those possible worlds with all the differences in each one among those
creatures? In this paper, I show that just as with the original argument against the doctrine of Divine simplicity, this is a
difficulty that confronts virtually all theists and even some atheists, not just classical theists, and point to a solution to the
difficulty that can reconcile a deterministic causal relation between the Divine creative act and the effects thereof with the
doctrine of Divine simplicity.

1. Introduction
The doctrine of Divine simplicity has been the subject of recently renewed interest and vigorous
defense by those attracted to classical views of the Divine nature.1 (Barrett 2017, Brower 2008,
Cohoe 2017, Cohoe 2020, Crisp 2019, Dolezal 2011, Dolezal 2017, Duby 2016, Grant 2004,
Levering 2017, Long 2019, Minich & Kamel 2019, O’Connor 1999, Rogers 1996, Rogers 2020,
Stump 2016, Stump & Kretzmann 1985) But it has also faced serious criticism in the form of claims
that it is both intrinsically incoherent and incompatible with other commitments of classical theism,
some of which might be thought to be ultimately more fundamental to classical theism than the
doctrine of Divine simplicity (DDS) itself. In this paper, I wish to focus on one specific claim of
incompatibility that has been leveled against the DDS, namely the claim that it is incompatible with
the classical and ancient Christian doctrine that God created freely, and not out of any necessity
whatsoever.2 A number of philosophers and theologians have pressed this objection against the

1 As with most doctrines, the doctrine of Divine simplicity comes in several versions due to various authors. In this
paper, my focus will be on the Thomistic version, according to which there is no real distinction between God and His
attributes (and therefore between any attribute and any other), between God and His essence, between His essence and
His existence, and between God and anything in Him which, in creatures, would be accidental, including His acts. I
focus on the Thomistic version of the doctrine both because it is the most controversial popular version of the doctrine,
giving a background against which most of the debate surrounding the doctrine revolves, and also because it is the
strongest popular version, such that if the objection I discuss in this paper fails against the Thomistic version of the
doctrine, it likewise will fail against any weaker version of the doctrine.
2 That God created (and more generally, in all He does, acts) freely is widely attested in Scripture and the Christian
Tradition. “Whatsoever the Lord hath pleased he hath done, in heaven, in earth, in the sea, and in all the deeps.” (Ps.
cxxxiv: 6) And it is an explicit dogma at least for Catholics, as defined by the Fathers of the First Vatican Council: “This
sole true God by His goodness and ‘omnipotent power,’ not to increase His own beatitude, and not to add to, but to
manifest His perfection by the blessings which He bestows on creatures, with most free volition, ‘immediately from the
beginning of time fashioned each creature out of nothing, spiritual and corporeal, namely angelic and mundane; and then
the human creation, common as it were, composed of both spirit and body.’” (Denzinger §1783) And: “If anyone does
DDS (Moreland & Craig 2003: 525, Mullins 2013, Leftow 2015 and Mullins 2016: 137–43), and
recently I (Tomaszewski 2019) made a reply to the effect that one simple and straightforward way of
formalizing this objection in fact an invalid argument. While I stand by everything I said in that first
paper about that simple argument, in this chapter my primary goal is to show that an important
intuition connected to the invalid simple argument attacked in that first paper is nevertheless
plausible and can be used to construct a more powerful, valid argument against the DDS. I then go
on to show, however, that this argument, if successful, can be adapted to create a modal collapse for
virtually all theists, not just those who accept the DDS. I conclude by showing that the new
argument can be rejected as unsound if we also reject the principle that the effects of the Divine
creative act are causally determined by that act considered as really identical (in the Thomistic sense
of “really identical”) to the Divine essence.3
But before going to defend the DDS against the charge of modal collapse, it is important briefly to
rebut a common motivation among contemporary philosophers and analytic theologians for
attacking the DDS: namely, that the doctrine (and all the alleged problems that come with it) are an
effect of an alleged unfortunate influence of Hellenistic philosophy on post-Nicene and especially
medieval Christianity (see Boyd 2017: 670 for a recent example of this complaint). Part of this
accusation is that the doctrine of absolute Divine simplicity demanded by Saint Anselm, Saint
Thomas Aquinas, and others in the Scholastic tradition is simply not Biblical and was unknown to
the early Fathers of Church, even if they did hold a more modest doctrine of Divine simplicity that
ruled out only material or integral parts in the Divine Essence. But this is blatantly contradicted by
the very robust doctrines of Divine simplicity taught by early Fathers such as Saint Hilary of Poitiers:
God, Who is Life, is not a Being built up of various and lifeless portions; He is Power, and
not compact of feeble elements, Light, intermingled with no shades of darkness, Spirit, that
can harmonise with no incongruities. All that is within Him is One; what is Spirit is Light
and Power and Life, and what is Life is Light and Power and Spirit. He Who says, I am, and
I change not (Malachi 3:6), can suffer neither change in detail nor transformation in kind.
For these attributes, which I have named, are not attached to different portions of Him, but
meet and unite, entirely and perfectly, in the whole being of the living God. (Hilary 1994)
Here, already in the mid-4th century, even before Saint Augustine, is a very robust DDS which not
only identifies many of the Divine attributes in precisely the way that opponents of the doctrine
think is absurd, but importantly insists that all that is within Him is one. It is true, of course, that
Saint Thomas Aquinas found in his metaphysics two new things (i.e. essence and existence) that in
God must be one, but this represents a progression in fundamental metaphysics, not in the DDS itself,
which was left formally the same from Saint Hilary to Saint Thomas in its declaration that all within
God is one. This is probably why Saint Thomas himself quotes this passage from Saint Hilary in the

not confess that the world and all things which are contained in it, both spiritual and material, as regards their whole
substance, have been produced by God from nothing, or, shall have said that God created not by a volition free of all
necessity, but as necessarily as He necessarily loves Himself, or, shall have denied that the world was created to the glory
of God: let him be anathema.” (Denzinger §1805)
3 For reasons that I will explain below, it is crucial to understand that all that must be rejected is a determination of the
effects of the Divine creative act by that act considered as really identical with the temporally and modally immutable
Divine Essence. We need not (and should not, on pain of denying God’s perfect freedom in choosing whether and what
to create and His ongoing Providence over His created effects) reject a determination of the effects of the Divine
creative act by that act considered as virtually distinct from the Divine Essence.
most important article (i.e article 7) of question 3 of the Prima Pars of the Summa Theologiæ, claiming
that Saint Hilary “touches” his own argument for excluding all composition whatsoever in God in
saying (in the different translation of the Summa): “God, Who is strength, is not made up of things
that are weak; nor is He Who is light, composed of things that are dim.” (ST I.3.7, respondeo)
But even in the ante-Nicene period, Saint Irenæus taught in his Against Heresies:
But if [heretics] had known the Scriptures, and been taught by the truth, they would have
known, beyond doubt, that God is not as men are; and that His thoughts are not like the
thoughts of men. (Isaiah 55:8) For the Father of all is at a vast distance from those affections
and passions which operate among men. He is a simple, uncompounded Being, without
diverse members, and altogether like, and equal to Himself, since He is wholly
understanding, and wholly spirit, and wholly thought, and wholly intelligence, and wholly
reason, and wholly hearing, and wholly seeing, and wholly light, and the whole source of all
that is good—even as the religious and pious are wont to speak concerning God. (Irenæus
2007)
Of course, Saint Irenæus is only two generations removed from Saint John the Evangelist, and yet
we still find in his work the core of the DDS: that God is “simple, uncompounded Being” who is
wholly identical to His attributes. These two quotations alone ought to put to rest any claims that
the DDS is not to be found in the doctrine of the early Church Fathers. Let’s move on to the
problem of modal collapse.

2. Simply Invalid
Simple arguments from modal collapse typically employ two immediate premises: one stating the
necessary existence of God or some Divine attribute, and another stating the identity of God or that
Divine attribute with something that is at least prima facie contingent, such as God’s act of creation or
the God’s knowledge that P, where P is any contingently true proposition.
One very straightforward way of characterizing the argument from modal collapse against the DDS
by attending to the identification on the DDS of God with His act of creation is like so:
S1. Necessarily, God exists.
S2. God is identical to God’s act of creation.
S3. Necessarily, God’s act of creation exists.
I’ll call this “the simple argument”.4 Proponents of the DDS are certainly committed to (S1) and
(S2), but want to reject (S3), since (S3) together with the very plausible premise that necessarily, if
God’s act of creation exists, then a creation exists, entails the necessary existence of a creation.5 So
this argument had better be invalid if the DDS is to escape the conclusion.

4It is a formalization of the argument found in Mullins (2016: 138), though without the subargument found there for
premise (S2).
5It is important to see that God’s act of creation can be designated as specifically as one likes in this argument to make
necessary not just a creation, but the creation of the actual world, exactly as it is, down to every detail.
It is worth noting here that the old Thomistic distinction between absolute necessity and
suppositional necessity6 won’t help with the simple argument, for the whole point of the argument is
that if God’s act of creation is identical to God, then it is as absolutely necessary as He is, and not
necessary merely by the supposition that God wills to create. In this sense, a reply to the simple
argument relying only this distinction would be question-begging.
But, fortunately for the proponent of the DDS, and as I (Tomaszewski 2019) have shown, this
argument commits the famous formal fallacy of substituting a contingently co-referential term into
the scope of a modal operator. So the argument is simply invalid, for the reasons explained long ago
in Quine (1953: Ch. 8), who gives us the following counterexample:
C1. Necessarily, 8 is greater than 7.
C2. The number of the planets is identical to 8.7
C3. Necessarily, the number of the planets is greater than 7.
And it is a good thing for proponents of the simple modal collapse argument that it is invalid, since
many of them accept both of the premises of the following argument I call the “alternative
argument from modal collapse”:
A1. Necessarily, God exists.
A2. God is identical with the Creator.
A3. Necessarily, the Creator exists.
Both (A1) and (A2) are doctrines of all three Abrahamic religions, for example, and yet (A3) leads to
precisely the same trouble as (S3), since the necessarily existence of a Creator, together with the very
plausible premise that, necessarily, if there is a Creator, then there is a creation, entails the necessary
existence of a creation.8 So the form of the simple argument, if it were valid, would prove far more
than most of its proponents would like.
The problem arises because while the relation of identity is metaphysically necessary (that is, it is
metaphysically impossible for anything not to be identical with itself), identity statements (i.e. statements
asserting that an identity relation holds) are not always necessarily true, and it is truth that is relevant
to the validity of an argument. Identity statements are not always necessarily true, because, as Kripke
(1980) showed, the truth of an identity statement depends not just on whether an identity relation
holds between an entity designated in one way and that same entity designated in another way, but
also on the successful designation of that entity by those two designating terms. So, for example,

6 See ST I.19.3.
7Note that it would a serious mistake to reject Quine’s counterexample due to thinking that (C2) is false because the
number of the planets instantiates 8 rather than being identical to it. It is true, of course, that the planets instantiate the
number 8, but the number of the planets does not instantiate 8, because the number of the planets is a number, and no
number plausibly instantiates a number except for 1, which instantiates itself since it is just one thing, like every other
number.
8 And just as with the simple argument, the Creator can be designated as specifically as one likes, to make necessary not
just a creation, but the creation of the actual world, down to every last detail.
while Saint Louis IX is identical to the King of France in in 1250, and that relation of identity holds
necessarily between Saint Louis and the King of France in 1250, it is obviously contingent that “the
King of France in 1250” designates Saint Louis and not some other man. That contingency in the
relation of designation between terms appearing in an identity statement and the entities designated
by those terms can make identity statements contingently true even while the relation of identity
holds necessarily. Thus, while the relation of identity between God and God’s act of creation (or the
Creator, in the alternative argument) holds necessarily, just as the identity relation between anything
else and that same thing holds necessarily, the contingency creeps in by way of designating God by
the term “God’s act of creation,” since the relation of designation holding between this term and
God is contingent, not necessary. It is contingent because, in that world in which God does not
create, “God’s act of creation” does not designate God, since it designates nothing at all. So while
the identity relation between God and God’s act of creation holds necessarily, the identity statement
that God is identical with God’s act of creation is contingently true, and therefore cannot support
the conclusion that necessarily, God’s act of creation exists in a valid argument.
Moreover, as I explain in Tomaszewski (2019: 279), the most obvious way of validating the
argument (i.e. introducing a necessity operator in front of the second premise) involves introducing a
new premise into the argument to which proponents of the DDS are not committed (i.e. proponents
of DDS need not and should not concede that necessarily, God is identical to God’s act of creation),
and thereby validates the argument only at the cost of giving up on the claim that the DDS, by itself,
entails a modal collapse.
Nor can one escape fix the simple argument by attempting to eliminate the problem of the
contingency of designation by giving a de re reading of the simple argument. A natural way of doing
so would be to use Leibniz’ Law like so:
LL1. Necessarily, if God is identical to God’s act of creation, then God and God’s act of
creation have all the same properties.
LL2. God is identical to God’s act of creation.
LL3. God has the property of necessary existence.
LL4. Therefore, God’s act of creation has the property of necessary existence.
While this Leibniz’ Law version of the simple argument is valid (when “God’s act of creation” is
read de re), and even sound, the problem now is that (LL4) does not lead us to any modal collapse,
because while God’s act of creation does indeed have the property of necessary existence, and
therefore exists even in that world in which God does not create, it is not an act of creation in that
world, and therefore a creation need not exist in that world. By going to a de re reading of “God’s act
of creation,” the proponent of the simple argument eliminated the problem of invalidity, but only at
the tremendous cost of also eliminating that implication from the conclusion to an actual modal
collapse, since a modal collapse is implied by the necessary existence of God’s act of creation only
on a de dicto reading of “God’s act of creation” in (LL4).
To summarize, then: the simple argument is invalid, and even if it were not invalid, arguments of the
same form and appealing in no way to the DDS would prove the same disastrous conclusion as the
simple argument. This latter fact is quite important, because it shows that if the fundamental
intuition behind the simple argument were right (that God, being necessary, cannot be identical to
anything which is contingently what it is), it would cut just as much against any Abrahamic theism as
against classical theism!

3. Modal Indiscernibility and a New Argument

3.1 The DDS and Real Modal Indiscernibility
So, is the DDS out of the modal collapse woods? Not so fast! For the DDS is committed to the
something I’ll call “the modal indiscernibility thesis”:
MIT If the DDS is true, then God in any given possible world is really, intrinsically
indiscernible from God is any other possible world.
MIT tells us that, on the DDS, if we look just at how God is really and intrinsically in any possible
world, we could not distinguish Him on that basis from how He really is in any other possible
world.9 In other words, how God is really and intrinsically is invariant across all possible worlds. The
DDS entails MIT because it excludes the possibility of any real, contingent attributes in God, for the
DDS rules out any accidents in God, and God’s essence does not vary really from possible world to
possible world (an important fact that I’ll discuss more below). So, the DDS just doesn’t leave any
room for real, intrinsic modal variance in God: He is really absolutely the same in every respect in
every possible world.
The astute reader will have noticed that I included the qualification “really” in my statement of the
modal indiscernibility thesis. I do not mean by “really” in this content the colloquial sense of “really”
which is synonymous with “truly” and antonymous with “falsely.” Rather, I mean “really” as Saint
Thomas Aquinas and other Scholastic authors meant it, meaning “as the thing is in itself, prior to
any cognition of it.” It is antonymous to “logically”10 in this sense, which means “as thing is in the
mind, posterior to some cognition of it.” This is crucially important because at least the Thomistic
doctrine of the DDS (the one I am concerned to defend) does not claim that there are no
distinctions whatsoever in God. Rather, it claims that there are no real, absolute distinctions in God.11
The qualification “absolute” permits relative distinctions in God arising from relations of mutual
opposition such as Saint Thomas identifies with the Divine Persons in his doctrine of the Trinity
and doesn’t concern us here. The qualification “real” permits logical distinctions of both the virtual
and purely logical varieties in God.12 Virtual distinctions are distinctions with some foundation in the
thing itself, insofar as that thing lends itself to the distinction made by the mind, even though that
distinction is not really present in it, such as the distinction in men between their sentient life and

9 Thomists and other Aristotelians have occasionally complained that the Kripkean semantics of possible worlds is
inadequate to capture the Aristotelian theory of modality. I’m not sure about this complaint, but those who are needn’t
worry about my usage of the term “possible world” throughout this paper, as I mean to refer by it only to ersatz possible
worlds as a convenient shorthand for the global possibilities for the actual world. In my usage, therefore, Aristotle and
Aquinas could have easily spoken of “possible worlds.”
10 The term “conceptually” is also sometimes used, but I will use “logically.”
11As to the compatibility of virtual distinctions in God with the Thomistic DDS, see QPD IX.8 ad 4. As to the
compatibility of relative distinctions with the Thomistic DDS, see In Sent. d. 2, q. 1, a. 3.
12 See Feser (2014: §1.3) for an introductory treatment of the Scholastic theory of distinctions.
their rational life, which are really identical but lend themselves to distinction in the mind insofar as
we might say that the mind which distinguishes such things is carving the one real thing at its joints,
to use a loose but helpful metaphor. Purely logically distinctions are distinctions with no such
foundation in the thing itself, such as the distinction in Venus between the Morning Star and the
Evening Star. In such distinction the mind is reading the distinction into the thing rather than reading
it out, as in virtual distinctions.
All of the foregoing is important for two reasons. The first reason is that the distinctions between
God and His acts (including that of creation), between God and His attributes, and between God
and His ideas are all virtual. The second reason is that while the DDS entails the MIT, it does not
entail the stronger principle that would result from omitting the qualification “really.” Or put
another way: the DDS is consistent with virtual variance in God across possible worlds, which is to
say that the DDS is consistent with God being conceived (whether by us, or more importantly, by
God Himself) in different ways in different possible worlds. For example, the DDS is altogether
compatible with God’s conceiving of Himself as an act of creation in some worlds and not in
another world (and doing so truly in both cases), or with God’s conceiving of Himself (truly) as the
exemplar cause of lions in those worlds where He freely creates lions and not so conceiving of
Himself (again, truly) in those worlds where does not create lions.13 All of this will be important to
my solution below to the new modal collapse argument I am presently developing.
Before moving on, it is important to note briefly that the entailment from the DDS to the MIT is
not without at least some controversy, at least among Thomists. Eleonore Stump seems to deny it
(though she doesn’t clarify if the relevant modal discernibility is real or merely logical):
Consequently, on Aquinas’s interpretation of divine simplicity, not all God’s acts are
necessitated; as contemporary philosophers would put this point, God is not the same in all
possible worlds. On the contrary, on Aquinas’s interpretation of divine simplicity, it is in
fact right to say that there is contingency in God, in our sense of the term ‘contingency’.
But if so, then there is no problem about God’s having alternative possibilities open to him.
It is true that God is not changeable across time. At each and every time, God is one and
the same. But since even on the doctrine of simplicity, God can do other than he does, this
is sufficient for the claim that God has free will, that God has the power to choose among
alternative possibilities. (Stump 2016: 206)
But it’s not remotely clear how a rejection of the entailment from the DDS to the MIT could be
right, either as a matter of what Aquinas explicitly says concerning contingency in God, or as a
matter of what is consistent with Aquinas’ broader account of Divine simplicity. Concerning the
former, Aquinas tells us, “As regards the things which are in God himself, nothing can be described
as potential: all is naturally and absolutely necessary.” (QPD III.15 ad 11) And as to the latter,
Aquinas’ doctrine of Divine simplicity and general metaphysics is certainly committed to the
propositions that: (i) God is necessarily really identical to His essence (ST I.3.3) and (ii) whatever is
really essential to a thing is absolutely necessary to it.14 But these two propositions jointly entail the

13 See Doolan (2008) for a full treatment of God and His ideas as exemplar causes in Aquinas’ work.
14This latter claim is, of course, a principle that almost everybody who endorses essentialism about anything endorses
concerning essence. Aquinas endorses it explicitly: “Now whatever is absolutely necessary differs from the other types of
necessity, because absolute necessity belongs to a thing by reason of something that is intimately and closely connected
with it, whether it be the form or the matter or the very essence of a thing.” (In Meta. V.6.883)
MIT: if there were any real, intrinsic discernibility between God in one possible world and God in
any other possible world, this would be a discernibility in His essence by proposition (i), which is
impossible by proposition (ii). Nor, of course, should any theist (let alone a classical theist) be willing
to countenance the suggestion that God could differ essentially across possible worlds.15
But how does MIT lead to a modal collapse? Recall premise (S2) of the simple argument from
modal collapse above: God is identical to His act of creation. If God is identical to His act of
creation, and God is also really, intrinsically indiscernible across all possible worlds, as the DDS
entails, then although God may not be identical to an act of creation in that possible world in which
He does not create, He is nevertheless really, intrinsically indiscernible from one, because He is really,
intrinsically indiscernible from how He is in the actual world, and in the actual world, He is identical
to an act of creation.16 But how could God be really, intrinsically indiscernible from an act of
creation in those possible worlds in which He doesn’t create?

3.2 The Modal Indiscernibility Argument

We can turn this puzzle into a new, valid, stronger argument from modal collapse against the DDS:
MI1. God is absolutely simple.
MI2. If God is absolutely simple, then He is really, intrinsically indiscernible across all
possible worlds.
MI3. If God is really, intrinsically indiscernible across all possible worlds, then His act is
really, intrinsically indiscernible across all possible worlds.
MI4. If God’s act is really, intrinsically indiscernible across all possible worlds, then the
effects of this act do not really vary across all possible worlds.17
MI5. Therefore, the effects of God’s act do not really vary across all possible worlds.
I’ll call this “the modal indiscernibility argument” (or “MI argument” for short). The MI argument is
plainly valid, and (MI5) is a modal collapse.

15 Thanks to Eleonore Stump and Alexander Pruss for discussion of this issue.
16 Here, the advocate of the modal collapse worry might sense an opening: if God is really, intrinsically indiscernible
from an act of creation, then, by the identity of indiscernibles, He is an act of creation. And since He is so indiscernible
in every possible world, He is an act of creation in every possible world! But this is too fast for at least three reasons.
Firstly, as Max Black (1952) has famously argued, the identity of indiscernibles is very questionable in general. Secondly,
the identity of transworld indiscernibles, which is what this objection would require to cause problems, being strictly
logically stronger than an intraworld principle of the identity of indiscernibles, is even less plausible than it. Thirdly, God
is only really and intrinsically indiscernible across possible worlds. Crucially, He is logically and extrinsically discernible
across possible worlds, and His being an act of creation is, at least on classical theism, precisely one of God’s extrinsic
properties at least in the sense that His possession of it depends on at least one thing that is not God, namely the
existence of at least one creature.
17“Effects” here should be read broadly enough to include not just the relata of a causation relation, but also what is
grounded in a grounding relation, especially if one thinks that the relation between contingent attributes and essence (if
any) or Creator and creature is one of grounding. I take it that, at this point in the dialectic over the modal collapse
objection to the DDS, that nothing rides on whether the relation between either of these two sets of relata is that of
causation or that of grounding.
Does the proponent of the DDS have any reasonable reply? Since the argument is valid, the
proponent of the DDS must reject at least one premise. (MI1) just states the DDS, so obviously not
that one. (MI2) is a restatement of the MIT, which I’ve explained and defended above; rejection of
(MI2) is not consistent with the DDS, so that premise must stay as well. Likewise, (MI3) is entailed
by the DDS: insofar as God’s act is intrinsic to Him, it cannot vary across possible worlds any more
than anything else really intrinsic to Him, given the MIT.18 So it’s the rejection of (MI4) or bust for
the proponent of DDS. And indeed, this will be the premise I would counsel the proponent of the
DDS to reject.

4. What’s Simplicity Got to Do With It, Anyway?

But it’s not just the proponent of the DDS that ought to reject (MI4). Every theist who accepts that
God exists necessarily, accepts that God has at least a modal essence, and accepts the Principle of
Sufficient Reason ought to reject (MI4). To see why, consider this argument parallel to the MI
argument:
EMI1. God exists necessarily.
EMI2. If God exists necessarily, then God’s essence is really, intrinsically indiscernible across
all possible worlds.
EMI3. If God’s essence is really, intrinsically indiscernible across all possible worlds, then
the effects of God’s essence do not really vary across all possible worlds.
EMI4. If the effects of God’s essence do not vary across all possible worlds, then God’s
non-essential attributes do not really vary across all possible worlds.
EMI5. If God’s non-essential attributes do not really vary across all possible worlds, then
God’s act is really, intrinsically indiscernible across all possible worlds.
EMI6. If God’s act is really, intrinsically indiscernible across all possible worlds, then the
effects of this act do not really vary across all possible worlds.
EMI7. Therefore, the effects of God’s act do not really vary across all possible worlds.
I’ll call this “the essential modal indiscernibility argument” (or “EMI argument” for short).
Swinburne is the only well-known author who even remotely plausibly denies (EMI1), and even his
denial is plausibly only of God’s logically necessary existence. (Swinburne 2004: 79) I’ll set aside going
forward the thesis that God exists with only metaphysical contingency, especially because it seems
clear to me that such a claim is more absurd than even a modal collapse.
The second premise is a commitment of essentialism about God: if God has any essence
whatsoever19 and exists necessarily, this is just for that essence to be really, intrinsically indiscernible

18Of course, something extrinsic to God is ultimately necessary in order for His act to count as an act of creation, namely
a creation.
19The only sense of “essence” necessary for the EMI argument to work is that of modal essence, or a set of properties
possessed by a thing in every world in which it exists.
across all possible worlds. Or, to put it more bluntly, any possible God whose essence is really
discernible from the essence of the actual God would be a numerically distinct God.
The non-classical theist rejects (EMI3) only at the price of giving the classical theist adequate excuse
to reject (MI4), since (EMI3) is simply a restatement of (MI4) with for God’s essence rather than
God’s act. If what is caused or grounded by God’s act cannot vary across all possible worlds unless
God’s act is really discernible across all possible worlds, then what is caused or grounded by God’s
essence cannot vary across all possible worlds unless God’s essence is really discernible across all
possible worlds. At the very least, the proponent of the MI argument would owe us a very strong
reason to think that God’s act is relevantly different from God’s essence in this respect, and how so.
(EMI4) is true, in turn, because God’s non-essential attributes (if He has any) can only be caused or
grounded by either God’s essence or by His other non-essential attributes.20 Appeal to God’s other
non-essential attributes would lead to a regress that, in turn, could be terminated only by appeal to
God’s essence. So, ultimately, all non-essential Divine attributes must be caused or grounded by
God’s essence. Suppose, on the contrary, that the non-classical theist claimed that some of God’s
non-essential attributes are uncaused. Then they are either necessary or contingent. If they are
necessary, then the consequent of (EMI4) is still true. If they are contingent, then the non-classical
theist is positing contingent, uncaused entities, which violates the PSR. This is a serious problem not
only because the PSR is at the heart of many cosmological arguments for the existence of God, but
also because it opens the door for the classical theist to escape by the same means: if the non-
classical theist can posit contingent, uncaused Divine attributes, then the classical theist can posit
contingent, uncaused entities outside of God altogether, and so escape a modal collapse. And so, by
(EMI3), God’s non-essential attributes cannot vary across all possible worlds.
Since God’s act is one of God’s attributes, it is either essential or non-essential. If it is essential, then
it cannot vary across all possible worlds, and if it is non-essential, then by (EMI4) it cannot vary
across all possible worlds. Not to really vary across all possible worlds is, by definition, to be really,
intrinsically indiscernible at all possible worlds. Thus, (EMI5).
Finally, what is left is (EMI6), which is identical to (MI4). So, I think the non-classical theist ought
to reject (EMI6). But he does so only at the price of allowing the classical theist to reject (MI4) and
escape modal collapse as well! The only other option left to the non-classical theist is to accept
(EMI6) alongside the other premises of the EMI argument and therefore accept (EMI7), which is
the very modal collapse that he has been warning the classical theist is entailed by the DDS.
Allow me briefly to summarize the parallel between the MI argument and the EMI argument more
informally, but perhaps more insightfully. The central worry at the heart of the MI argument is just
this: if God is absolutely the same in every possible world, how could His created effects be
different across all possible worlds? But, crucially, the EMI argument shows us that this is not a
worry faced only by classical theists, for non-classical theists escape this worry only by positing non-
essential, contingent attributes in God that themselves must be caused or grounded by something
that is really, intrinsically indiscernible across all possible worlds, namely the Divine essence. That is,
by introducing composition into God, the non-classical theist only pushes back one step the central
worry animating the MI argument, rather than eliminating it. Thus, that central worry is a worry for

20Fruitful appeal to variance in either creatures or necessary but non-essential attributes across possible worlds cannot
be made here, because variance in creatures across possible worlds is the very phenomenon the argument concludes
against (and so appeal to it to reject (EMI4) in response to the EMI argument would be question-begging), and because
necessary attributes cannot vary across possible worlds despite their non-essentiality.
any theist who posits an essence for a necessarily existing God and accepts the PSR, not just for
classical theists who accept the DDS. In fact, we can go a step further and say that it is a worry for
anybody who accepts that all contingent things find their ultimate cause or ground in one or more
necessary things, which is a very broad class of people indeed! Arguably, it is everybody who accepts
a sufficiently robust version of the PSR.
What we have found, then, is that just as with the simple argument from modal collapse examined
above, the central worry behind the MI argument is a worry not just for classical theists, but for
most theists in general. And whatever non-classical theists might say to escape that worry is
something that can be translated rather readily into an equally viable escape from modal collapse for
the classical theist. This urges the question: does the modal collapse objection to the DDS (in any of
its formulations) really have anything to do with Divine simplicity? It seems not. If it does, its
proponents need to tell us exactly how the DDS is supposed to contribute to that collapse over and
above the contribution to it made by the fact that, on most theisms and even some atheisms, at
some point the contingent is derived from the necessary. We are all in that same boat.

5. Creative Determinism and Hyperintensionality

5.1 Creative Determinism
So, what is almost everybody who is confronted by this problem to do? As I said above, the classical
theist ought to reject (MI4). Those familiar with the literature on determinism will have recognized
(MI4) for what it is: an assertion of a deterministic causal relation between God’s act (considered
really and intrinsically) and God’s created effects. But what reason does the classical theist (or
anybody else, for that matter) have to accept such a determinism, especially in a context in which
there is so little evidence for it? Many claim that the success of the natural sciences since modernity
are evidence of physical determinism, but assuming that there is a God, what reason is there for
thinking that the relation between God’s real creative act and what that act causes is deterministic?
None whatsoever. Indeed, as Elizabeth Anscombe, in her critical and very relevant “Causation and
Determination,” argues against a broader determinism:
There is something to observe here, that lies under our noses. It is little attended to, and yet
still so obvious as to seem trite. It is this: causality consists in the derivativeness of an effect
from its causes. This is the core, the common feature, of causality in its various kinds.
Effects derive from, arise out of, come of, their causes. For example, everyone will grant
that physical parenthood is a causal relation. Here the derivation is material, by fission. Now
analysis in terms of necessity or universality does not tell us of this derivedness of the effect;
rather it forgets about that. For the necessity will be that of laws of nature; through it we
shall be able to derive knowledge of the effect from knowledge of the cause, or vice versa,
but that does not show us the cause as source of the effect. Causation, then, is not to be
identified with necessitation. (Anscombe 1981: 136)
And if there were any such reason, this would only return us to the parallel described above: if the
causal or grounding relation between the Divine real creative act and the effects of that act is
deterministic, why should we think that the causal or grounding relation between God’s essence and
His non-essential attributes is indeterministic?21
Importantly, one might think that rejection of such a deterministic causal relation between God’s
real creative act and His created effects will, in turn, require us to reject what Alexander Pruss calls
“contrastive PSR,” the principle that every proposition of the form P rather than Q has an
explanation.22 This is because, if we reject such a deterministic causal relation, how could we explain
why God created world W1 rather than world W2, where W1 and W2 are any two distinct possible
worlds?
Supposing for a moment that this is so, the classical theist (and other theists in the modal collapse
boat with him) should not fear. For such contrastive explanations are just as unavailable for
indeterministic free choices (necessary for free will as understood by libertarians) and possibly many
quantum phenomena (deterministic theories consistent with the experimental data exist but are
unpopular with scientists)23 as they are for the classical theist. The classical theist would therefore be
in broad and arguably good company in rejecting contrastive PSR.

5.2 The Hyperintensionality of Creative Causal Contexts

But the classical theist in fact need make no such concession. Nor need he, strictly speaking, reject a
deterministic causal relation between God’s act of creation and the created effects thereof. All of this
is because God’s act of creation is the action of an intelligent Agent, and thus the fruit, in part, of His
intellectual activity and Divine wisdom. That is to say that the Divine creative act is posterior to God’s
cognition of Himself, and is therefore is, while really identical to God, virtually distinct from Him.
And mental causation of the sort that God is engaged in when He creates is hyperintensional: even
necessarily co-referring terms cannot be substituted salva veritate into the context of a sentence
reporting an instance of mental causation.24 For a very pedestrian example, consider:
R1. Thinking that I saw Jack the Ripper caused me to run away.
R2. Thinking that I saw Francis Tumblety caused me to run away.
Both (R1) and (R2) report instances of mental causation, but (R1) does not entail (R2) even if we
grant that “Jack the Ripper” and “Francis Tumblety” necessarily co-refer because both are rigidly
designating names and Francis Tumblety was Jack the Ripper (which is very unlikely, but he was a
suspect for a time).
Now, compare:
Z1. God’s idea of zebras caused zebras to exist.

21I think there is, in general, a great deal more reason to accept grounding necessitation than causal determinism, but
that should be seen simply as reason to think that the relation between God’s essence and God’s non-essential attributes,
or God’s creative act and His created effects, is that of causation and not grounding.
22 See Pruss (2006: 148-155) for critical discussion of the principle.
23 For philosophical discussion of such a theory, see Goldstein (2021).
24 See Nolan (2014) for a broader discussion of hyperintensionality and its importance in metaphysics.
 Z2. God’s idea of unicorns caused zebras to exist.
Here, (Z1) is true while (Z2) is obviously false, even though God’s idea of zebras is necessarily really
identical to God’s idea of unicorns (and to the Divine essence).25
Thus, we can keep a deterministic causal relation between the Divine creative act and the effects
thereof, as well as contrastive PSR (at least as far as this one question goes), if we bear in mind that
the Cause in question is virtually distinct from the Divine essence (because it is posterior God’s
cognition of Himself) and that mental causation is hyperintensional.
One might wonder or object here: how can God’s intellectual activity, or the Divine ideas which
figure in His creative act, differ from possible world to possible world without any difference in how
God really is across possible worlds? This is simply a fundamental truth about the intentional order:
it is considerably and necessarily more fine-grained than the real order. Consider, for example, a
triangle: one and the same real thing, without any real difference, can be conceptualized either as
triangular or as trilateral. Or consider the number 2: one and the same real thing, without any real
difference, can be conceptualized either as the first prime or as half of 4. These concepts are obviously
distinct in the intentional order. But they also obviously necessarily represent one and the same real
thing. And such examples abound. So one cannot insist against the defender of the DDS that there
is a one-to-one correspondence between the real order and the intentional order, and therefore there
can be no decisive objection here against logical discernibility in God across possible worlds without
real discernibility in God across those same possible worlds.
It’s true, of course, that the situation is more complicated with God: my model requires logical
discernibility in God considered as intellectual Agent (that is, as One Who has different ideas) and as
intellectual activity (that is, as the cognition of different ideas) and not merely considered as
intentional object (that is, as the ideas themselves), as in my examples with the triangle and the
number 2. But no matter: once the possibility of a one-to-one correspondence between the
intentional order and the real order falls, it falls altogether. Or, at least, there is no reason to think
that one could reasonably insist on such a correspondence for intellectual agents or activities but not
for intentional objects. At a bare minimum, one pressing such an objection would owe us a
substantial account of how such a correspondence could exist for intellectual agents or activities
without being possible for intentional objects. A simplified version of the essence of my reply, then,
is this: if there is logical “room” in the real objects of cognition for logically distinct ideas, then there
is no obviously good objection to the claim that there is such logical “room” in the real intellectual
agents and activities that generate those logically distinct ideas, as well.
How, then, does God’s intellectual activity differ from possible world to possible world without any
real difference in Him? Simply in virtue of being intellectual activity. No further explanation is
necessary, and its possibility, at least to me, is doubtful at best.

6. A Conclusion to Modal Collapse

From a very plausible beginning, the modal collapse worry and the arguments pressing it have
collapsed: the simple argument is invalid and has a form that would lead to a modal collapse for
non-classical theists as well as classical theists. And the modal indiscernibility argument, along with

25 See ST I.15.2 ad 1.
the fundamental intuition behind it, similarly finds both classical and non-classical theists alike in the
boat of alleged modal collapse, because classical and non-classical theists alike are in the logically
fraught game of deriving the metaphysically contingent from the metaphysically necessary.
But it is actually the careful distinction drawn by Aquinas and other Scholastics between the real
distinction and the virtual distinction, together with a careful consideration of the role of the Divine
intellect in the creative act, that shows us how we can reconcile the noble aspiration of the classical
theist for a First Cause devoid of all composition and potentiality with God’s ability to determine
whether and what He creates.
References
Anscombe, E. 1981. “Causation and Determination,” in Metaphysics and the Philosophy of Mind: Collected
Philosophical Papers, Volume II, by Elizabeth Anscombe (Oxford: Basil Blackwell), 133–147.
Aquinas, T. 1948. Summa theologiæ, trans. Fathers of the English Dominican Province. New York:
Benzinger Bros.
Aquinas, T. 1952. Quaestiones disputatae de potentia Dei, trans. Fathers of the English Dominican
Province. Westminster, MD: The Newman Press.
Aquinas, T. 1961. In duodecim libros Metaphysicorum expositio, trans. John Rowan. Chicago: Henry
Regnery Company.
Aquinas, T. 2000. Scriptum super IV libros Sententiarum. Pamplona, Spain: University of Navarra. URL:
https://www.corpusthomisticum.org/iopera.html.
Barrett, J. 2017. Divine Simplicity: A Biblical and Trinitarian Account. Minneapolis: Fortress Press.
Black, M. 1952. The Identity of Indiscernibles. Mind 61(242): 153–164.
Boyd, G. 2017. The Crucifixion of the Warrior God: Interpreting the Old Testament’s Violent Portraits of God in
Light of the Cross, Volumes 1 & 2. Minneapolis: Fortress Press.
Brower, J. 2008. Making Sense of Divine Simplicity. Faith and Philosophy 25(1): 3–30.
Cohoe, C. 2017. Why the One Cannot Have Parts: Plotinus on Divine Simplicity, Ontological
Independence, and Perfect Being Theology. Philosophical Quarterly 67(269): 751–771.
Cohoe, C. 2020. Accounting for the Whole: Why Pantheism is on a Metaphysical Par with Complex
Theism. Faith and Philosophy 37(2): 202–219.
Crisp, O. 2019. A Parsimonious Model of Divine Simplicity. Modern Theology 35(3): 558–573.
Denzinger, H. 1957. The Sources of Catholic Dogma, trans. Roy Deferrari. Freiburg: Herder & Co.
Dolezal, J. 2011. God without Parts: Divine Simplicity and the Metaphysics of God’s Absoluteness. Eugene,
OR: Wipf and Stock.
Dolezal, J. 2017. All That Is in God: Evangelical Theology and the Challenge of Classical Christian Theism.
Grand Rapids, MI: Reformation Heritage Books.
Doolan, G. 2008. Aquinas on the Divine Ideas as Exemplar Causes. Washington, DC: Catholic University
of America Press.
Duby, S. 2016. Divine Simplicity: A Dogmatic Account. London: Bloomsbury.
Feser, E. 2014. Scholastic Metaphysics: A Contemporary Introduction. Heusenstamm, Germany: Editiones
Scholasticae.
Goldstein, S. 2021. Bohmian Mechanics, in The Stanford Encyclopedia of Philosophy (Fall 2021
Edition), ed. Edward N. Zalta. URL:
https://plato.stanford.edu/archives/fall2021/entries/qm-bohm/
Grant, W. 2004. Aquinas, Divine Simplicity, and Divine Freedom. Proceedings of the American Catholic
Philosophical Association 77: 129–144.
Saint Hilary of Poitiers. 1994. De Trinitate, in Nicene and Post-Nicene Fathers, Second Series, Vol. 9, eds.
Philip Schaff & Henry Wace, trans. E.W. Watson & L. Pullan (Peabody, MA: Hendrickson
Publishers).
Saint Irenænus. 2007. Against Heresies, in The Ante-Nicene Fathers, Vol. 1, eds. Alexander Roberts,
James Donaldson, & Arthur Coxe (New York: Cosimo Classics), 309–567.
Kripke, S. 1980. Naming and Necessity. Cambridge, MA: Harvard University Press.
Leftow, B. 2015. Divine Simplicity and Divine Freedom. Proceedings of the American Catholic
Philosophical Association 89: 45–56.
Levering, M. 2017. Engaging the Doctrine of Creation: Cosmos, Creatures, and the Wise and Good Creator.
Grand Rapids, MI: Baker Academic.
Long, D. 2019. Thomas Aquinas’ Divine Simplicity as Biblical Hermeneutic. Modern Theology 35(3):
496–507.
Minich, J. & O. Kamel. 2019. The Lord is One: Reclaiming Divine Simplicity. Landrum, SC: Davenant
Press.
Moreland, J. & Craig, W. 2003. Philosophical Foundations for a Christian Worldview. Downers Grove, IL:
InterVarsity Press.
Mullins, R. 2013. Simply Impossible: A Case against Divine Simplicity. Journal of Reformed Philosophy 7:
181–203.
Mullins, R. 2016. The End of the Timeless God. Oxford: Oxford University Press.
Nolan, D. 2014. Hyperintensional Metaphysics. Philosophical Studies 171(1): 149–160.
O’Connor, T. 1999. Simplicity and Creation. Faith and Philosophy 16(3): 405–412.
Pruss, A. 2006. The Principle of Sufficient Reason: A Reassessment Cambridge: Cambridge University
Press.
Quine, W.V.O. 1953. Reference and Modality, in From a Logical Point of View, by W.V.O. Quine
(Cambridge, MA: Harvard University Press), 139–159.
Rogers, K. 1996. The Traditional Doctrine of Divine Simplicity. Religious Studies 32: 165–182.
Rogers, K. 2020. An Anselmian Approach to Divine Simplicity. Faith and Philosophy 37(3): 308–322.
Stump, Eleonore. 2016. Simplicity and Aquinas’s Quantum Metaphysics, in Die “Metaphysik” des
Aristoteles Im Mittelalter: Rezeption Und Transformation, ed. Gerhard Krieger (Boston, MA: De
Gruyter), 191–210.
Stump, E. & Kretzmann, N. 1985. Absolute Simplicity. Faith and Philosophy 2(4): 353–382.
Swinburne, R. 2004. The Existence of God. New York: Oxford University Press.
Tomaszewski, C. 2019. Collapsing the Modal Collapse Argument: On an Invalid Argument Against
Divine Simplicity. Analysis 79(2): 275-284.