Before delving into my response, I want to emphasize how truly grateful I am to each of the individuals for being such wonderful interlocutors and human beings. Their responses are so incalculably valuable to our collective pursuit of truth. We are, fundamentally, on the same team. As dialectical partners (not dialectical opponents), we share the same goals: (i) greater understanding of the issues, and (ii) discovering treasures of truth. It’s a mutual effort. We are on the same path towards a common destination. Explorers on a journey.
Before reading through this post, here is a fair warning: it’s long. But that’s for good measure. I’m dealing with twelve-to-thirteen (yes, a whole dozen!) problems all at once for one of the most formidable research programs in human history. That requires being thorough. And it would be remiss of me not to give a thorough treatment for my interlocutors who have given such thorough and gracious criticisms of my post. If it helps, you can break your reading of this post up into sessions, since there are natural breaks in the article.
 Why didn’t I just split this up into a series of posts? Well, I knew people would only selectively pay attention to some problems and not others, and I want to avoid that. Some problems build off or compliment others. And also, I think we can learn from all the objections, since there is value in all of them (and in my responses to them). Don’t be picky, young padawan!
So, sit back, relax, grab a cup of coffee or apple juice or a glass of wine, and let’s do some philosophy!
Problem 1: Absolute Simplicity and Theistic Conceptualism (response by Juliano)
Let’s see what the original argument was:
(1) For any proposition P, P is identical to God’s thought (that P). (Theistic Conceptualism as applied to propositions)
Another way to put this is that for any universal form or pattern F, F is identical to God’s conceptual thought (about F).
(2) Necessarily, if x and y are identical, then whatever is true of x is true of y.
(3) The proposition P*, ‘there are dogs’, is about dogs.
(4) The proposition P**, ‘there are particles’, is not about dogs.
(5) So, P* and P** are not identical. (From (2), (3), and (4))
(6) So, God’s thoughts are not identical. (From (5) and (1))
(7) If God’s thoughts are not identical, then God is not absolutely simple.
(8) So, under divine conceptualism, God is not absolutely simple. (From (6) and (7))
Here is the first rejoinder to my argument:
Firstly, as the author is well aware, the classical theist could deny that God’s thoughts/ideas are propositional. Even if we were to swap out knowledge of propositions for knowledge of universal forms/patterns, the classical theist need not think that the multiplicity of forms entails that there is a one to one correspondence between propositions/forms and God’s act of thinking them/knowing them. For Aquinas, God knows/thinks all things by “one intuition.” So that God knowing a plurality of ideas/concepts does not entail a plurality of thoughts within God’s mind. As Michelle Panchuk notes, “…[God] has no need to ‘divide’ up divine knowledge into concept-size bits.”  Thus, the classical theist could plausibly deny premise 1.
In response, note that the argument is not an argument against classical theism per se, but an argument that theistic conceptualism and classical theism are incompatible. Although there are a number of versions of theistic conceptualism, I explicitly targeted the version according to which propositions, universals, and other abstracta are identical to God’s thought(s). If the classical theist denies this version of theistic conceptualism, then he or she hasn’t rebutted or undermined the argument; on the contrary, he or she has simply accepted the very conclusion I sought in the first place (namely, that if one is to be a classical theist, one must reject theistic conceptualism so defined).
Second, the rejoinder notes that “God knowing a plurality of ideas/concepts does not entail a plurality of thoughts within God’s mind.” Very well, but that doesn’t contravene anything in the argument. The argument never claimed something along the lines of ‘a variability in the things God knows is or entails a variability in God himself’. Rather, the argument is aimed at divine conceptualism as applied to propositions, according to which propositions are utterly identical to divine thoughts. This doesn’t, then, concern God’s knowledge of variability; it has to do with the ontological status of propositions.
The point, then, seems to remain (and again, I’m not saying it’s decisive; it’s a prima facie problem): if indeed propositions are identical to divine thoughts, then divine simplicity is false. For there’s more than one proposition. So, if propositions are identical to God’s thoughts, there is more than one divine thought. And that’s not allowed under divine simplicity. On the flip side, if there is only one divine thought, and propositions are identical to God’s thought, then there is only one proposition. But that ain’t true — there are many, many propositions.
The main thing to realize, then, is that we are focusing on the ontological status of propositions and other abstracta, not God’s (mode of) knowledge of a multiplicity of things.
The rejoinder also makes an excellent point here: “the classical theist need not think that the multiplicity of forms entails that there is a one to one correspondence between propositions/forms and God’s act of thinking them/knowing them.”
Rob Koons echoes a similar criticism:
The basic problem with your argument is an equivocation on the phrase “God’s thoughts”. We have to distinguish the act of God’s thinking from the objects of thought. God’s thinking has multiple objects but a single act. Many of the objects lack real existence–they have only “intentional existence” or “inexistence” (to use Brentano’s phrase). The intentional existence of the objects is dependent on God, and only God has real existence necessarily, so His aseity is preserved.
In my estimation, however, this distinction was already taken into account. For if there is only a single act of thought, then the act of thought clearly cannot be identical to a multiplicity of propositions. And that was the sole intent of the argument: propositions are not identical to God’s act of thought. Instead, there will be a multiplicity of propositions — even if God comprehends or grasps them all — independently of God’s act of thought. The divine conceptualist in this case seems either forced into anti-realism or Platonism.
The objects of thoughts are (we are supposing in the debate about the ontological status of propositions and universals) propositions and universals. These either exist or they don’t. If they don’t exist, we have anti-realism (implausible in my view). If they exist, then they cannot be identical to God’s act of thought. So, they must be something other than God’s act of thought.
Now, to be sure, I need to study intentional existence further. Prima facie, the suggestion seems eerily close to nominalism. In my estimation, abstracta either exist in mind-independent reality or they don’t exist in mind-independent reality. “Well, what about the divine mind?” As we have seen, however, the abstracta cannot be identical to God’s act of thought. Such abstracta, it seems, could only be the intentional objects of God’s act of thought. But then we need to inquire about the ontological status of those objects. If they don’t exist, we seem cast into nominalism. If they do, then since they aren’t identical to God or God’s act of thought, we seem cast into platonism.
Once again, though, I need to study these matters in greater depth for a more informed and considered view.
In correspondence, Dwight Stanislaw responded to these last points as follows:
The very via media you’re leaving out is moderate realism and the view that ‘existence’ and ‘being’ are said in many ways. We aren’t left with the dichotomy of Nominalism and Platonism as you suggest. It isn’t enough to say that by not having concrete, mind-independent existence we are left with no option but Nominalism. … [I]ntentional existence, beings of reason, and so forth really do exist, though they are not mind-independent. Something needn’t be construed as only one or the other, and this is precisely why such debates developed and views were refined following Boethius, Porphyry, and others during the inception of the medieval period. Realism needn’t be robust Platonism.
I don’t think I’ve neglected moderate realism as a middle way, though; it’s just that absolute divine simplicity doesn’t seem (prima facie, that is) to jibe well with it. Allow me to explain.
As Feser explicates moderate realism, abstracta exist in the things themselves (qua particularized or individualized) or in the minds which do the abstracting (as in, the universal of triangularity purportedly exists in my mind when I grasp triangularity or when I abstract away the particularizing, individuating features of the multifarious triangles of my experience).
 I say ‘purportedly’ because the notion of ‘existing in’ is rather unclear. More importantly, though, many philosophers are skeptical that the universal triangularity itself exists in our minds; instead, a representation of said universal exists in our minds. I leave this debate off to the side, however.
But an apparent problem arises for moderate realism: some abstracta are necessarily existent. For instance, the proposition ‘if there are philosophers, then there are philosophers’ is necessarily true. Call this proposition P. Now, it seems clear that if x doesn’t exist, x doesn’t have any properties. In other words, if x has properties, then x exists. Having properties seems to presuppose that the thing exists in the first in order to bear such properties. But P has the properties of ‘being necessarily true’, ‘necessarily corresponding to reality’, and so on. So, P exists. Now, P either exists contingently or necessarily. But if P exists contingently, then P can fail to exist. And if P can fail to exist, P can fail to be true (since having the property being true presupposes existence). But P cannot fail to be true — it is a necessary truth. So, P cannot fail to exist. So, P necessarily exists.
 For an elaboration and defense of an argument along these lines, see Rasmussen, “From Necessary Truth to Necessary Existence“
How to solve this problem? Well, it seems that no particular triangle could necessarily exist. After all, triangles on buildings, paper, and so on will (presumably) eventually cease in the heat death of the universe in billions of years time (or the ‘cold death’, where everything is so spread out and scattered that no organized complexity exists and, a fortiori, no triangular representations in the universe). But the only option left, then, is for such abstracta to exist in Plato’s heaven or as thoughts in a necessarily existent mind.
 It’s a bit more complex. For instance, it could be a mixture of the two — some exist in the necessary intellect, others in Plato’s heaven. Or, some (or all) abstracta could be grounded in the necessary intellect not in virtue of its thoughts but perhaps in virtue of its nature. In fact, it could be the case (in principle, that is — apart from additional considerations in favor of a necessary mind) that all minds are contingent, but that, necessarily, there is some mind or other such that necessarily existent abstracta are grounded in it (or them). In this case, though each mind is contingent, we couldn’t infer that all the minds (together) are contingent. That’s a scope distinction fallacy (also known as a quantifier shift). Though, we can set these complications to the side, since they don’t impact the force of the argument at hand.
But under absolute divine simplicity, we have seen that God’s act of thought cannot be identical to abstracta, for there is a unicity of the former but a multiplicity of the latter. And that means that we can’t identify the multiplicity of necessary abstracta with God’s act of thought. Koons proposed that they are the intentional objects of God’s act of thought. But now we still have the question: what are such objects? What is their ontological status? We are thus cast back into the very debate (viz. that concrning the ontological status of abstracta) for which theistic conceptualism sought a solution. Such (intentional) objects cannot be intrinsic to God or God’s act of thought, for God (and God’s act of thought) is identical with that which is intrinsic to him. So, it seems that we therefore have necessarily existent abstracta extrinsic to God. And that seems to be Platonism.
We could accept the the conjunction of moderate realism and divine conceptualism as I’ve defined it, but (it seems) only at the cost of absolute divine simplicity. So, I don’t fail to take into account moderate realism; it’s just that moderate realism seems to militate against absolute divine simplicity.
Finally, if we deny God’s thoughts are propositional, or even are (or correspond to) universals, then once again we have simply denied divine conceptualism and hence accepted the conclusion I sought — that divine conceptualism and classical theism are not compatible. For now we are cast back into debate concerning the ontological status of propositions and universals without recourse to God’s thought(s) or knowledge. If God’s thought, knowledge, or what have you is non-propositional, then clearly we haven’t provided an account of the ontological status of propositions by appeal to God’s thought (knowledge). Similar considerations apply to the multiplicity of non-propositional abstracta.
Fundamentally, then, Juliano’s first remark doesn’t seem to rebut my argument but seems to accept the denial of divine conceptualism demanded by classical theism.
Let’s continue our appraisal of Juliano’s response by considering his next point(s):
Further, it is not entirely clear to me that the classical theist couldn’t also deny premise 7. In drawing on the work of Gregory Doolan, Matthew Levering notes that there can be a multiplicity of divine ideas in a certain sense.  Levering (a la Doolan) draws the distinction between logical and ontological multiplicity. For Levering, that there is a multiplicity of divine ideas can be chalked up to logical multiplicity without entailing ontological multiplicity within the being of God.  So, if the multiplicity of divine ideas is understood to be a logical multiplicity, this would not entail that the divine ideas would be absolutely identical (i.e. without any sort of distinction). Yet the consequent of premise 7 (that God is not simple) would not follow because the logical multiplicity of divine ideas would not entail an ontological multiplicity within God. Thus, the classical theist has potential grounds for rejecting premise 7 as well.
Wonderful point again. However, once again, it seems to me that it fails to take into account the dialectical context at hand.
The dialectical context at hand is one wherein the debate concerning the ontological status of propositions, universals, and other abstracta is front and center. One proposal for grounding (or identifying) such abstracta is the divine mind. A particular brand of this is theistic conceptualism. A particular brand of theistic conceptualism (as I’ve defined it) states that abstracta exist in virtue of being identical to God’s thought(s). Fundamentally, then, the dialectical context at hand is one wherein the ontological status of abstracta is on the line.
But with this insight in mind, we can start to see what is inadequate about Juliano’s response. For abstracta are really distinct from one another, not merely logically distinct. Triangles are really, ontologically distinct from squares. The pythagorean theorem is really distinct from the fundamental theorem of calculus. The proposition
is really distinct from the proposition , which is really distinct from the proposition .
But that means that any adequate ontological grounding in terms of identity (e.g. the identity of abstracta with divine thought(s)) of one such abstract object O1 cannot be merely logically a(and hence not really) distinct from an adequate ontological grounding in terms of identity of another such abstract object O2, where O1 is really distinct from O2. And my argument simply follows from this: divine conceptualism construed in terms of identity — if it is to be an adequate ontological ground of really distinct abstracta– entails really distinct thoughts in the divine intellect. And that ain’t allowed under classical theism.
The appeal to logical (as opposed to real or ontological or metaphysical) multiplicity, then, cannot save the view according to which divine conceptualism and classical theism are compatible.
 Attentive readers will notice the similarity here between the rejoinder at hand and the Scholastic theory of distinctions. Because it may come up later, it is worth understanding this theory from our good pal Ed Feser:
“Scholastics define a real distinction as one that reflects a difference in extra-mental reality and a logical distinction (or “distinction of reason”) as one that reflects only a difference in ways of thinking about extra-mental reality. A logical distinction can be either purely logical or virtual. It is purely logical when it is merely verbal, without any foundation in reality.
It is virtual when it has some foundation in reality. For example, a man’s nature as a rational animal is (given the Thomistic account of essence, to be discussed in chapter 4) in reality one thing, not two. But we can view it either under the aspect of rationality or under the aspect of animality, for we know of instances when animality exists apart from rationality.”
Virtual distinctions, then, are such that in extramental reality there is only one, single, numerically identical thing. However, we can view such a single reality under different aspects. This might become important later, but it might not. We shall see.
I conclude, then, that the first argument retains its force.
Problem 2: Absolute Simplicity and Trinitarianism (response by Chutikorn)
Let’s see what the original argument was:
(1) The Father, Son, and Holy Spirit are either distinct or not.
(2) If they are not distinct, then they are identical.
(3) If they are identical, then there are not three persons in one God but rather one person – in which case, Christianity (according to which Trinitarianism is essential) is false.
(4) If they are distinct, then they possess distinct attributes – in which case, God is not absolutely simple.
(5) So, either Christianity is false, or ADS is false.
Here is the first rejoinder to my argument:
1. The Father, Son, and Holy Spirit are either distinct or not.
It is a teaching of Christianity that the persons of the Trinity are distinct. However, it is imperative to point out that this cannot be an absolute distinction (one of essence) since this would entail the heresy of tritheism, which is repugnant to Christianity. According to Aquinas, the persons of the Trinity are relatively distinct. In other words, they are distinct according to relative opposition. What this means is that the principle of distinction amongst the persons is origin which entails a real opposition between each person as to where they come from within the interior (ad intra) operations of God himself. The Father is distinct from the Son insofar as he is the principle (the generator) of the trinitarian processions, the Son is distinct from the Father insofar as he proceeds by way of intellect from the Father as knowledge of himself, and the Holy Spirit is distinct from both Father and Son insofar as he proceeds from both by way of will as love.
So, these are mainly clarificatory remarks from Chutikorn. I’ll add some such remarks as well.
The relevant notion of distinction at play is real distinction, i.e. distinction (non-identity) among the things in themselves — as they are in extramental reality, ontologically or metaphysically (as it were). The sense of identity at play can be captured by the phrase ‘Leibnizian identity’ and obeys what’s known as Leibniz’s Law (the indiscernibility of identicals):
Leibniz’s Law (LL): For all x and for all y, if x = y, then whatever is true of x is true of y and whatever is true of y is true of x.
It is important to note that my argument does not require or presuppose that the persons of the trinity are ‘absolutely’ distinct in the sense of possessing different essences. My argument leaves the precise nature of the distinctness open (apart from, of course, the requirement of distinct attributes/features, which we will come to in due time).
Now, it’s important to note that nothing thus far has been leveled as a rebutting or undercutting defeater against the argument. Chutikorn has been stage-setting (as it were).
I will, however,make a few points with respect to relational opposition. Chutikorn writes that “the principle of distinction amongst the persons is origin which entails a real opposition between each person as to where they come from within the interior (ad intra) operations of God himself.”
Now, this seems to support premise (4) — for if the principle of individuation or distinction amongst the persons is origin, then each divine person will possess distinct attributes or features (such as the feature of ‘being spirated from the mutual love of the Father and the Son’ or ‘being eternally begotten of the Father’ or what have you). And that’s precisely what (4) infers from the real distinction among persons of the trinity — they have really distinct attributes/features. And (so the argument goes) this is not compatible with absolute divine simplicity according to which all of God’s intrinsic attributes or features are identical (i.e. not really distinct) with one another and with God himself. Thus far, then, the argument seems on good footing. Let’s turn, then, to the rest of Chutikorn’s criticism.
2. If they are not distinct, then they are identical.
This premise is ambiguous and does not capture the different kinds of “distinction” and “identity” used to explain a real distinction of persons in the Trinity. Flowing from my explanation in response to premise 1, one can see that a) the persons are not absolutely distinct and thus are identical in essence, and b) that they are relatively distinct and thus not identical with respect to origin.
Once again, though, the argument doesn’t require or presuppose ‘absolute’ distinction. All it requires, it seems, is ‘real distinction’, i.e. non-identity captured (in part) by LL.
Here is another way to see where the argument is coming from. Relations presuppose relata. In other words, there can’t be a relation without relata to stand in the relation. Now, the relations amongst the persons in the trinity are not merely reflexive relations. For instance, presumably nothing can ‘beget’, ‘generate’, or ‘spirate’ itself, for that would seem to require itself to be both prior and posterior to itself, which is absurd. So, the relations are not reflexive. But if relations presuppose relata, and the relata aren’t reflexive, then there are three, non-identical, distinct things standing in these various relations. And that’s precisely (part of) what the argument sought to establish. Given there are three, distinct, non-identical things, the only way they could be distinct is if they each have (one or more) features/attributes that the others lack. And that, again, is incompatible with absolute divine simplicity, since any (non-Cambridge) attributes/features of God are identical not only with one another but with God himself.
But couldn’t the different relations be the individuating feature/attribute?
First, even if they could, the relations will be (i) intrinsic to God, not extrinsic to God, but (ii) NOT identical to God himself (since there are distinct, not identical relations among the three divine persons, whereas there is only one, numerically identical God). But that seems to entail composition in God, for then there is something intrinsic to God that is not identical to God.
But second, it seems that x and y standing in a non-reflexive relation already presupposes that x and y are non-identical in the first place, for if x and y were identical, then clearly x and y couldn’t stand in a non-reflexive relation. But if that’s the case, then the non-identity (distinctness) of x and y cannot be constituted by their standing in a non-reflexive relation. There must ‘already’ be individuating attributes or features between them, in which case the distinction won’t be accounted for in terms of different relations but instead in terms of different attributes/features. And again, that’s precisely the conclusion of my original argument.
 Moreover, relations seem in some sense less ontologically fundamental than their relata. Put another way, relata seem more ontologically fundamental than the relations in which they stand. Relations depend on their relata, not the other way around. But if that’s the case, then the non-identity between x and y cannot be constituted by a non-reflexive relation in which they stand, since their standing in such a relation would already presuppose the more fundamental reality of the distinct x and y. It must be the case that x and y are ‘already’ distinct in order for them to stand in a non-reflexive relation, since relations are in some sense dependent upon and hence posterior to their relata.
3. If they are identical, then there are not three persons in one God but rather one person – in which case, Christianity (according to which Trinitarianism is essential) is false.
This antecedent is ambiguous since being identical in essence does not necessitate that the persons are identical with regard to where they proceed from. Being that each divine procession is a real one (truly proceeding from intellect and will), there exists unique re relativa which are really distinct in their terms while remaining one in the divine essence. In fact, it is precisely because of absolute divine simplicity that the divine relations are three distinct modes of subsistence since any relations in a simple God cannot be accidental. Each of the persons (or subsistent modes) are of the same nature but can still differ (without contradiction) in their origin within the order of eternal processions.
Once again, however, my argument neither requires nor presupposes difference in essences. All it requires is at least two really distinct, non-Cambridge attributes of God. That, it seems, suffices for the argument against absolute simplicity.
Though, this does seem to spawn a different, related argument.
Now, either the Son is one in essence with God’s essence or not. If not, then trinitarianism is false. But (we are supposing) trinitarianism is true. So, the Son is one in essence with God’s essence. Now, God is identical to God’s essence. So, the Son is one in essence with God. But the Son has no accidents that effect composition with the Son’s essence, so the Son is his essence. But the Son’s essence is identical to (i.e. one with) God’s essence, which is identical to (one with) God. So, the Son is identical with God (in the robust, Leibnizian sense). So, whatever is true of the Son is true of God (and vice versa). But God himself is not begotten from anything; there is nothing apart from God from which God could proceed or ‘be begotten’. (Or so it seems). But the Son is begotten from something really distinct from him (even if that real distinction is cast in terms of a relational opposition). So, something is true of the Son that is not true of God. So, the Son both is and is not identical to God.
 What about the relations the Son stands in? Well, those are either essential or not. If they are not essential, then the Son could have failed to stand in those relations. But that seems absurd — the Godhead is necessarily trinitarian, not merely contingently Trinitarian. If the Son failed to stand in the relation of being begotten (say), it seems that the Son wouldn’t be who he is. But that isn’t possible — the Son necessarily exists. So, it seems the relations are essential to the Son. But since the Son’s essence is one with God’s essence (and hence God himself), it follows that the relations are essential to God himself as well.
Alternatively, we could reason that Jesus is one with God himself, and the Father is one with God himself (since both share the same essence with God, and God is identical to God’s essence). But Jesus is begotten while the Father is not. So, God is both begotten and not begotten.
 Now, a non-classical theist could seemingly avoid such difficulties, since the non-classical theist can hold that each divine person has various accidents that allows them to remain of one essence with one another but still really distinct from one another. But such an option is not open to the classical theist.
Once again, these considerations deserve further thought. Prima facie, though, they seem plausible.
4. If they are distinct, then they possess distinct attributes – in which case, God is not absolutely simple.
Since the divine relations are subsistent, those relations just are the divine essence. That is to say, the divine essence and the divine persons differ only in their mode of intelligibility – there is a logical distinction between persons and essence, while there is a real distinction between the subsisting persons. Now, one might think the Father possesses the unique attribute of “being the generator” and the Son possesses the unique attribute of “being filiated” and so argue that the antecedent of the premise is true. However, the trinitarian monotheist can deny that such “attributes” are constitutive of the nature of the Father and Son. They are not additional attributes of the divine essence, like omnipotence, omniscience, etc., but rather they are unique relational properties that can be constitutive of the persons without being constitutive of their shared divine nature.
I want to focus on two claims here in particular:
(1) Chutokorn states that “there is a logical distinction between persons and essence, while there is a real distinction between the subsisting persons.”
But a logical distinction is not a real distinction. When there is a logical distinction, that is a distinction of reason; in extramental reality, there is one, single, numerically identical thing. So if the persons are only logically distinct from the Divine essence, then they are really identical to the divine essence. And since real identity is transitive, it follows they are really identical to one another. It follows, then, that there is not “a real distinction between the subsisting persons.” It seems we would have to deny the transitivity of real identity to hold that.
(2) “Now, one might think the Father possesses the unique attribute of “being the generator” and the Son possesses the unique attribute of “being filiated” and so argue that the antecedent of the premise is true. However, the trinitarian monotheist can deny that such “attributes” are constitutive of the nature of the Father and Son. … [R]ather they are unique relational properties that can be constitutive of the persons without being constitutive of their shared divine nature.”
But once again, my argument doesn’t require or need the distinct features or attributes to be constitutive of the shared Divine nature. All the argument requires is that there is a multiplicity of really distinct non-Canbridge attributes in God, regardless of whether such really distinct non-Cambridge attributes are constitutive of the shared nature. This alone seems enough to threaten absolute simplicity.
Other elements of this last rejoinder of Chutikorn’s are such that I already attended to them above. I conclude, then, that the second argument retains its force.
Problem 3: Absolute simplicity and Christ (response by DeRosa)
The original argument roughly went like this. Christ has two natures, one divine and one human. But if that’s true, then God is part of Jesus. But if classical theism is true, then God cannot be part of anything. So, (the dominant account in) Christology is incompatible with classical theism.
 See my original post for a slightly more detailed and refined exposition.
[The post’s] problems can be avoided if the divine nature is not a part of Christ. Let’s explore that briefly.
Ambiguity arises regarding what “Christ” refers to. Prior to the Incarnation, we can speak of “the Son” or “the second person of the Trinity” as a divine person. Now, this does not entail that there are two really distinct things: 1) the Son and 2) his nature. Rather, “the Son” refers to one reality: a divine person. After the Incarnation, we can speak of “Christ” and “the Son” interchangeably as subjects of various sentences. Here, “Christ” refers to the divine person that has assumed a human nature. It makes sense to say, “Christ is human” and, “Christ is divine” since “Christ” is the divine person possessing a human nature in virtue of which both sentences are true.
Schmid says, “Christ himself is not completely and utterly identical to the divine nature. After all, Christ is truly human, whereas the divine nature is not human.” Based on what I explained, we can deny this. “Christ himself” refers to the divine person that assumed a human nature, and so “Christ himself” is identical to the divine nature in the sense that all of the divine persons are identical to the divine nature. It’s just that “Christ himself” also possesses a human nature.
The trouble, though, is with LL. If Christ also possesses a human nature but (say) the Father or the Spirit don’t, then the Son is not identical to the Father or the Spirit. Similarly, if Christ, C, is such that C also possesses a human nature, but the divine nature is such that the divine nature does not also possess a human nature, then C is not identical to the divine nature. And since the divine nature is not extrinsic to C but yet not identical to C, the divine nature is a part of C. Or so the argument goes. DeRosa’s response does not, therefore, render the argument inefficacious.
 Natures do not possess natures; they are natures. The divine nature does not possess a human nature; for whatever possesses a human nature is corporeal, and the divine nature is not corporeal.
The fundamental thing to realize, it seems, is that something is true of Christ that is not true of the divine nature: Christ has two natures (the divine nature and the human nature), whereas the divine nature doesn’t ‘have’ any natures at all. The divine nature doesn’t ‘have’ the divine nature — it is the divine nature. Similarly, the divine nature doesn’t ‘have’ a human nature. So, something is true of Christ that is not true of the divine nature, and hence Christ is not completely and utterly identical to the divine nature. The proffered rejoinder does not sidestep the argument.
 Examples, it seems, can be multiplied. Christ died for our sins, but the divine nature didn’t die for our sins. The divine nature can’t die, or cease, or undergo any process at all. Christ digested bread, but the divine nature can’t digest anything since the divine nature lacks a digestive system, stomach, gastric acid, and so on.
I conclude, then, that the third argument retains its force.
Problem 4: Presentism and Pure Act (response by Stanislaw)
Let’s see what the original argument was:
(1) If some times are not actual, then for some x, the proposition that ‘x is actual’ genuinely changes in truth value.
(2) God knows all true (knowable) propositions.
(3) If ‘x is actual’ genuinely changes in truth value, and if God knows all true (knowable) propositions, then God’s knowledge changes.
(4) If God’s knowledge changes, then God isn’t purely actual.
(5) So, if some times are not actual, then God isn’t purely actual.
(6) So, if God is purely actual, then all times are actual.
(7) If (6) is true, then God’s being purely actual entails eternalism.
(8) So, God’s being purely actual entails eternalism.
The easiest and most straightforward response is to reject (2) as is given its ambiguity and to modify it, because while God knows all true propositions, he knows them according to his mode of knowing, not ours. What is missing is the qualification of the premise accordingly: (2)* God knows all true (knowable) propositions tenselessly in his eternity. Once we’ve modified (2) and replaced it with (2)*, (3) is false, because while it is true that genuinely changes in truth value, and that God knows all true (knowable) propositions tenselessly in his eternity, it is false that God’s knowledge changes. Here, then, our two problematic conclusions, (4) and (5), no longer follow, as God’s knowledge does not change and some times in themselves can be unactualized without worry that God will no longer be Pure Act.
The justification provided for (4) is acceptable enough if we are talking about human knowers, but even if knowledge essentially involves the intrinsic features listed and God did indeed possess them, we needn’t accept that these features change within God, for if each feature is possessed eternally, then they, along with God, would remain unchanging.
As is often the case, the problem begins at the outset due to an error and allowing it to infect the entire argument. In this instance, the error is taking God to be a knower in the same sense that human knowers are, and articulating divine knowledge and time after our manner of speaking without important qualifications. As Aquinas wisely notes, “The difficulty in this matter arises from the fact that we can describe the divine knowledge only after the manner of our own…” (De Veritate, q. 2, a. 12).
With all due respect to Stanislaw (we have interacted on many occasions, and I have nothing but admiration for him and our friendship), this response is puzzling. My central rejoinder to it is that it is simply incoherent to tenselessly know truths which are such that they are objectively and essentially tensed (i.e. objectively true at some times but then become objectively false at other times). Let’s cash this out by having a deeper look at the premises of my original argument.
Premise (1) seems straightforwardly true. For if some times that did or will occur are not actual, then they either were actual or will be actual. And if they were actual but are not actual, or if they are not actual but will be actual, then the temporal contents of those times (i) changed from being actual to non-actual, or (ii) will change from being non-actual to actual. And if the temporal contents change in such a manner, then the propositions reporting their actuality genuinely change from true to false (in the first case) or from false to true (in the second case).
Premise (3) follows from the fact that one can know a proposition only if the proposition is true. So, if proposition P changes (say) from true to false, then one’s knowledge that P cannot remain unchanged (for then P would remain unchanged in respect of being true). So, if God knows all true (knowable) propositions, and if some such propositions genuinely change in truth value, God’s knowledge must likewise change.
 Some may suggest that God’s knowledge is non-propositional. But this doesn’t seem to avoid the argument. All the argument requires is that God knows x is actual in some form or other. If he knows it in some form or other (even non-propositionally), and if x genuinely changes from being actual to non-actual, then God’s non-propositional knowledge must change as well. Suppose, for instance, that God knows temporal facts by acquaintance, not propositionally. Even so, one can only be acquainted with F if x exists. If x doesn’t exist — if x is precisely nothing — one cannot be acquainted with x.
Another way to see this is through the lens of succession. In timeless, tenseless eternity, there cannot be succession. So, if God knows P tenselessly, God unalteringly — without succession or alteration — knows P. But knowledge presupposes the truth of the thing known. So, if God unalteringly knows P, P is unalteringly true. For if P is not unalteringly true, then either P genuinely goes from being true to being false, or from being false to being true. But in that case, anything which knows P cannot know P when P is genuinely false. So, anything that knows P must go from not knowing P to knowing P (or vice versa in the case where P goes from being true to false). And so if P is not unalteringly true, anything which knows P cannot unalteringly know P.
Premise (4) also seems true. For if God’s knowledge changes, then God has various potencies: the potential to gain and lose knowledge, the potential to gain and lose beliefs and other propositional attitudes, and so on. If God knows P but then doesn’t know P (or vice versa), then there is some sort of succession in God. God would go or transition from knowing P to not knowing P — and that’s a sort of succession. And it seems that there cannot be succession in timeless eternity. Indeed, for the Thomist, God is his knowledge, power, and so on – in which case, if God’s knowledge can change, then God himself can change. And if God himself can change, then God is not purely actual.
 For instance, presumably God desires to have true beliefs. And if propositions change in truth value, then his desires must change in accordance with the ‘new’ truth values (lest he desire a falsehood).
Note, also, that I will cover the Cambridge-relation-account of God’s knowledge later in this post.
It is useful to distinguish between tensed propositions and tenseless propositions. Even if God knows the tenseless proposition ‘Donald Trump is president in 2020’, by itself that says nothing about whether God knows the tensed proposition ‘Donald Trump is currently president’. God can timelessly and tenselessly know the former, but he cannot tenselessly know the latter. And the reason is because the latter is not tenselessly true. It objectively and genuinely changes in truth value. The tensed proposition ‘Trump is currently president’ is genuinely and objectively false in 2012 but genuinely and objectively true in 2020. God cannot even tenselessly know the proposition ‘Trump is currently president’ in 2012, since the proposition is false, full-stop.
As we can see, the question is not about assuming human models of knowledge apply to God, thereby violating the doctrine of analogy so central to classical theism. On the contrary, it has to do with an analysis of knowledge as presupposing truth — and that’s the case regardless of the model of knowledge we take.
One might object:
From the point of view of God’s timeless eternity, the past, present, and future are all actual. However, this does not entail that they are actual simpliciter. From the point of view of temporal reality, only the present is actual; the past was actual and the future will be actual. So, God’s knowledge of that which is actual doesn’t change, but that is compatible with a dynamic view of time.
There are, however, two central problems with this. First, it embraces a view according to which whether something actually exists is relative to reference frame. In other words, some things that actually exist from God’s reference frame do not actually exist from our reference frame. Needless to say, this is a significant ontological price to pay. Intuitively, whether or not something actually exists is independent of perspective or reference frame. Just as it is absurd to suggest that something can be ‘true for me’ but ‘not true for you’, it seems equally absurd to say something can be ‘true for God’ but ‘not true for temporal beings’. Something either actually exists, or it doesn’t – there is a determinate fact of the matter regardless of one’s perspective.
The second problem is that it doesn’t actually avoid the entailment that God’s knowledge changes. For there are still propositions to the effect of ‘from the perspective of temporal reality, x is actual’. Moreover, unless we accept eternalism, such propositions will genuinely change in truth value. Hence, God’s knowledge of temporal actualities, where such propositions are indexed to the temporal reality’s point of view, will still require change.
Indeed, some of Stanislaw’s comments elsewhere seem eerily close to this frame-relative existence approach:
[K]nowing them [truths concerning temporal actualities] tenselessly means knowing them apart from being future to us, present in themselves, and past to us. A thing needn’t be actual in itself to be known by God, as God knows himself as the very cause of all that was, is, and will be in time.
But under presentism, there is no such thing as future to us and past to us. Future things don’t exist, simpliciter. Past things don’t exist simpliciter. This proposal then amounts to the claim that certain things don’t exist to us while other things do exist to us. But existence isn’t perspectival or frame-relative. There is no such thing as ‘existing to us’ or ‘not existing to us’. Things either exist or they don’t exist — there is no perspective under which x exists but a perspective under which x does exist. And if classical theism requires us to accept frame-relative existence (many thinkers have, on the basis of knowledge of temporal actualities and the truth of presentism, concluded just that), I would see that as a reductio of classical theism.
 Though, one’s modus tollens is another’s modus ponens, I guess.
Furthermore, appealing to God’s causality merely pushes the problem back a step, since if x doesn’t exist or isn’t actual, then God still can’t know that x exists or is actual, even if God is such that God timelessly bestows being to x wherever x is on the temporal timeline. For even if God’s act of creation is timeless, classical theists will still want to maintain that some times are not actual — they genuinely come into and pass out of being. And that’s all the argument requires, since then there will be truths concerning such temporal actualities which genuinely and objectively become true (or false).
Stanislaw and Rasmussen exchange on these points is illuminating and echoes a number of my rejoinders to Stanislaw:
Stanislaw: [B]ut the very issue is between types of knowing—God’s mode and our own—so it doesn’t seem to me problematic to hold both that presentism is true and that God can know all things in his eternity that will become actual in themselves relative to us.
Rasmussen: But on presentism, there isn’t anything in eternity that will become actual in themselves relative to us.
As Rasmussen points out, there simply is nothing in timeless eternity except for the present, for only the present exists. So, there cannot be anything in timeless eternity which is such that it will become actual in itself relative to us, for that would require future things already to exist in timeless eternity — and that’s not compatible with presentism.
Stanislaw provides an analogy to illumine his view:
Imagine my creation of a concerto that has all notes present to me and I conduct an orchestra to play it. The final note, future to the beginning of the piece, is known by me from its inception and known fully. I see no reason why we can’t hold both that I know the concerto in its entirety qua composer while individual notes are known as actual in themselves qua played when they in fact occur.
But this analogy, far from supporting Stanislaw’s point, actually undermines it. For you can see the written note on the page and know it in that sense. But that neither means nor entails that you know the note qua a presently existent sound. In fact, precisely because the final note — the sound itself — has genuinely and objectively not become actual yet, you cannot know that it is actual yet.
And this actually helps illustrate a distinction I made earlier. The notes on the sheet are like the B-series, tenseless facts. One can know those in their entirety regardless of the A-series, tensed facts. One can know that ‘Donald Trump is president in 2020’, ‘Obama is president in 2012’, and ‘Person X is president in 2064’ without having a single clue what the present president is. The tenseless, B-theoretic facts, related tenselessly under the ‘earlier than, simultaneous with, and later than’ relations, underdetermine the tensed, A-theoretic facts. And that’s precisely why you can know all the notes on the sheet (the tenseless, B-theoretic facts) while not knowing what notes are actually being played at present (the tensed, A-theoretic facts).
The point to see for now is that the analogy actually illustrates the deficiency in classical theism, for one cannot know that the final note is actual (the tensed, A-theoretic facts) until it is genuinely actual (since knowledge requires truth). And no amount of knowledge of the sheet music (the tenseless, B-theoretic facts) will grant you such knowledge.
Let’s end with a forceful, formal reconstruction of Rasmussen’s:
[I]f there is no x such that Fx (e.g., there is no iPhone 99), then there is no x such that Fx is known (by any means). The *method of knowledge* is not at issue. The form of inference is:
1. ~Ǝx Gx
2. So: ~Ǝx (Gx & Fx)
The method of knowledge is irrelevant with respect to this inference.
Assume there are many, many ways of knowing (whatever ways you wish). None apply to nothing.
This is, fundamentally, capturing the same insight of the omniscience problem I leveled. I conclude, then, that the fourth argument retains its force.
Problem 5: Knowledge of Contingents and Pure Act (response by Juliano)
Let’s see what the original argument was:
(1) There is some true proposition P describing the world such that ~P could have been true.
(2) Necessarily, God knows every true proposition.
(3) So, God knows P but could have known ~P. (1,2)
(4) If subject S could have different knowledge, then some intrinsic feature of S could have been different.
(5) So, some intrinsic feature of God could have been different. (4,5)
(6) If God is purely actual, then no intrinsic feature of God could have been different.
(7) So, God is not purely actual. (6,7)
The potential issue with problem 5 is the conception of knowledge that is employed therein. The argument seems to assume that knowledge (even for God) entails a passivity in the knower. While it generally true that intrinsic features of the knower—in at least some way—depend on that which is known (e.g. as is the case with human persons), such a view of knowledge necessitates that at least some intrinsic feature of the knower has the potency to be affected by what is known. This sort of potency would be classified as “Passive potency.” But such a mode of knowing is not the mode by which God knows things.
Juliano’s point here misses the mark, but it’s completely understandable given how underdeveloped (intentionally, mind you) my justification for premise (4) was. The argument is not that the objects of knowledge causally impinge in some manner on the knower; rather, it’s just that intrinsic potency of some sort (not necessarily potency caused or actualized by the object known) exists in the knowing subject. I’ll expound on this more below, but let’s hear Juliano out.
Proponents of the view that God is actus purus would hold that God’s act of knowing does not involve the realization of some passive potency within God. In fact, God as actus purus cannot have any passive potency. But if the contingency of God’s knowledge would not entail some passive potency in God, then it would not be the case that some intrinsic feature of God could have been otherwise. But how are we to understand the mode of God’s knowledge? The Thomist need not hold to a view of God’s knowledge that would introduce some passive potency within God. One reason for this would be that God’s knowledge is something that is active rather than passive.
For example, Aquinas holds that God’s knowledge is causative. And, as John Wright notes in characterizing Aquinas’s position, “…the divine intelligible goodness as actually being communicated to another is at once both God’s causality of that other and His knowledge of it. His knowledge is causative and his causality cognitive. He ‘knows things into being.’ In causing them to be, He knows them.” Thus, God’s knowledge must be understood as (or at least like) an exercise of God’s active power/potency. But that God’s active power/potency is not fully realized and/or might have been realized in other ways does not risk introducing passive potency in God.
The first thing to note is that active potencies are not, strictly speaking, potencies at all but are instead a kind of actuality. But if God’s knowledge of contingents is constituted by God’s active potency, and if God’s active potency is an actuality, then it seems God’s knowledge of contingents involves no potency and, a fortiori, no potency to possess different knowledge.
Now, here is the fundamental question: from whence does the possibility of having different knowledge come? In other words, what accounts for or grounds how God can have different knowledge?
If we cite active potency, it seems we haven’t given an adequate answer. For active potencies are a kind of actuality — an actual causal power or capacity. But if an active potency is an actuality, the question seems to just resurface; where is this possibility of having different, non-actual knowledge coming from? Citing something that is actual seems unable to account for how that very thing could be otherwise. Or, at least, if it is purely actual, then it has no potencies to be otherwise. Citing an active potency amounts to citing an actuality, and indeed an actuality (an actual causal power) whose manifestation is purely actual, devoid of potency. If the causal power itself is actual, and the being has no potency for the causal power not to manifest, it’s (at the very least) very unclear where or how a possibility of knowing otherwise comes into play.
The most fundamental point I want to make, however, concerns the justification for premise (4), since expounding on that will alleviate many of the concerns Juliano raises. For it seems that we need not appeal to the passive potency to be causally affected as the only way for knowledge of contingents to entail potency in God.
In justifying premise (4), we need to allay one immediate worry: God’s knowledge is merely extrinsically relational. According to this view, God’s knowledge may change, but this only constitutes a Cambridge change (i.e. not a ‘real’ change in/to God but only a difference in his extrinsic relation to something else that does really change). There is no change in God himself with such changes in his knowledge. Hence, God can remain purely actual even if his knowledge ‘changes’, as his knowledge is relational and extrinsic. Premise (4) is therefore false (or at least undercut).
 Indeed, this may be at the core of Juliano’s criticism, since God’s causing the world to exist (something Juliano is keen to identify with God’s knowing the world) is generally taken by classical theists not involve a change in or to God but rather a mere extrinsic, relational, ‘Cambridge change’.
I offer at least (once again, prima facie) problems for this Cambridge-relational view of God’s knowledge. First, knowledge is essentially a kind of mental state, and as such it essentially involves features intrinsic to the knowing subject. Mental states are by their very nature states of the conscious subject; they cannot consist in mere extrinsic or Cambridge relations to things outside or external to the knowing subject. But if that is true, it seems that (i) changes in knowledge entail changes in mental state, and (ii) changes in mental state entail changes in intrinsic features. By transitivity, changes in knowledge entail changes in intrinsic features. Hence, the appeal to mere extrinsic or relational (Cambridge) changes seems to fail.
Second, it seems one can only know P if one has justification for believing that P. Hence, God’s knowing that P entails God has justification and beliefs. But surely beliefs, qua propositional attitudes, are intrinsic to their bearers. What’s more, beliefs are mental states, and surely mental states are internal or intrinsic features of the subject to whom they belong. It also seems plausible that S’s justification must at least partly consist in intrinsic features of the knowing subject. For instance, surely it is partly constitutive of S’s having justification that (i) S is internally consciously aware of one or more reasons upon which S bases S’s belief. It seems, then, that we have reason to hold that changes in S’s knowledge entail changes in S’s intrinsic features (since S would have different beliefs, different reasons or grounds upon which S bases S’s beliefs, different conscious awareness of such grounds, and so on).
 After all, if one genuinely has no clue on what basis one believes a proposition, then for all one knows, one’s belief could easily have resulted from a random neural misfiring that induced a conviction of the proposition’s being true. Note that the condition in the text is compatible with externalism, since externalism (as I understand it) only holds that factors external to the subject can and do confer justification; but that neither means nor entails that the only factors required for knowledge are ones extrinsic to the subject.
Third, a difference or change in S’s knowledge seems to constitute a real difference or change in S as opposed to a mere Cambridge change or difference. For to suppose that it constitutes a Cambridge change or difference contravenes the evidence of our phenomenological experience of knowledge. When we know x but come to know something different, we seem to feel that something internal to us is different or has changed.
 Granted, God doesn’t come to know things in a manner involving temporal becoming. My point, however, isn’t essentially tied to temporal becoming, since it is the difference in knowledge that accounts for our feeling that something internal to us is different – not the temporal becoming involved in acquiring such a difference (although that, too, will involve phenomenological experience of change). For we feel the difference in our phenomenology even long after our acquiring the difference in the first place.
Additionally, there seem to be clear, relevant differences between changes in knowledge and mere Cambridge changes. Prima facie, Cambridge changes to conscious subject S are such that S is not necessarily aware of them, whereas it seems that knowledge differences require that S be aware of the difference. Take, for instance, a textbook case of Cambridge change to a subject: Smith ‘changes’ insofar as he transitions from being taller than Jones to being shorter than Jones – but only because Jones grows. Clearly, Smith can ‘change’ in this manner without necessarily being aware of having done so. In fact, it is precisely because Cambridge changes to a subject S have no causal influence on S that he or she need not be aware of them. In order for S to be aware of x, it seems x must have some influence on S. But by definition, Cambridge changes to S consist in influences and changes wholly extrinsic to S. Because changes in knowledge essentially involve awareness but Cambridge changes do not, it follows that changes in knowledge are not Cambridge changes. This is a fourth reason favoring premise (4).
Indeed, it’s precisely because knowledge essentially involves factors intrinsic to the knowing subject that the distinction between intensional and extensional contexts is meaningful. Take, for instance, the propositions and . Extensionally, these propositions are flatly contradictory, since the Morning Star is identical to the evening star. Intensionally, however, both propositions can be true due to factors internal to Smith (such as the mental clusters of descriptions Smith associates with expressions). This constitutes a fifth reason in favor of (4), since (4) essentially affirms that knowledge involves factors intrinsic to the knowing subject (such that a change in knowledge entails a change in intrinsic features). If knowledge didn’t involve such intrinsic features, we wouldn’t be able to differentiate between intensional and extensional contexts.
 Among other things, intensional contexts are characterized by the failure of intersubstitutivity of co-referring expressions salva veritate.  This is not to say that the proposition has two truth values (true and false); rather, its determinate truth value contextually depends on the construal – whether extensional or intensional – of the expressions ‘Morning Star’ and ‘Evening Star’.
Sixth, (4) seems intuitively true. It seems to be the case that changes in knowledge don’t solely involve factors wholly extrinsic to the knowing subject and instead involve changes in the subject him or herself. And, plausibly, its seeming to be the case that P provides defeasible evidence for P’s being true. Hence, this seeming provides defeasible justification for (4)’s being true.
 This follows from phenomenal conservatism. For a defense of phenomenal conservatism, see Huemer (2001). The strength of the seeming may depend on your position on the epistemic landscape. I invite you to consider the intuitiveness of (4) for yourself. Speaking from my epistemic situation, my intuitive sense of (4)’s truth is strong.
Seventh, even if changes in God’s knowledge constitute mere Cambridge changes, it’s plausible that (i) true propositions are such that God desires to know them, and (ii) a change in S’s desires is (or entails) an intrinsic change in S. For desires are essentially mental attitudes, inclinations, or tendencies toward some intentional object, and hence they seem to be intrinsic features of the subject. A change in desires therefore seems to constitute a change in intrinsic features of the subject. And it is plausible that God desires to know true propositions, as possessing knowledge is intrinsically valuable. Both (i) and (ii) therefore follow.
 Here is another reason why desires cannot consist merely in Cambridge relations. Ontologically prior to God’s creative act, God surely had desires (e.g. a desire to create). But in such a state, there was nothing external to God to which God could stand in a Cambridge relation. Hence, God had desires without standing in any Cambridge relations, and thus desires cannot consist merely in Cambridge relations.  Moreover, if God could know the true proposition but doesn’t, then he has the potential to possess more value (in virtue of the intrinsic value of knowledge).
Now, one might think God doesn’t have desires. This, however, is deeply problematic. God would not desire the flourishing of human beings; he would not even desire to create anything at all; he would not desire anyone’s salvation; he would not desire eternal union with any creature; and so on. A personal being without desires/intentions seems unintelligible.
 Ed Feser — who of course isn’t representative of Thomism or classical theism as such — agrees God has something like desires: “We tend or incline toward something because we rationally apprehend it as worth pursuing, or incline away from something because we rationally apprehend it as not worth pursuing. We possess rational appetite… [And] there is in God something like… rational appetite” (2017, p. 224).
Alternatively, one might think God has desires only analogously. But the argument only requires that God has desires, whether univocally or analogically. For even if God has desires only analogously, he nevertheless desires to know true propositions. And hence if the truth value of (certain) propositions genuinely changes, his desires – whether analogical or univocal – change. Similarly, even if God has knowledge analogically, that’s compatible with the argument. All the argument requires is that God knows the propositions concerning temporal objects’ actuality – regardless of whether he knows them analogically or univocally. If his knowledge is to track the truth, and if the truth values genuinely change, then his knowledge must change as well.
 Unfortunately, while I’ve known about Matthews (2012) for some time, I haven’t gotten around to reading it. Before I finish my paper(s) on omniscience and classical theism, however, I have that one as essential reading for the paper. So I certainly won’t neglect it.
Finally, Juliano argues:
Further, just to say a bit in response to the author’s defense of premise 4 (which is actually offered in problem 4), the problem with God’s changing knowledge does not lie merely in the fact that God could have known otherwise. Rather, I would assert that for God’s knowledge to change precisely in the way that Feser describes is problematic insofar as such fluctuating knowledge presupposes that God has a real relationship with creatures and, hence, God has some passive potency. Such a relationship cannot be had if God is pure act. Thus, that God’s knowledge could have been otherwise tout court does not risk undermining God’s status as actus purus.
Evaluating the view that God has no real relations with creatures will unfortunately take us too far afield.
 For what it’s worth, I have to confess something: it seems deeply, deeply flawed to me, especially with respect to Christian theism. But we shan’t get into it here.
Suffice it to recognize for now, however, that the problem concerning knowledge of contingents seems to remain alive upon consideration of Juliano’s criticisms, for it seems (i) that the appeal to active potencies doesn’t ground the possibility of being (and hence knowing) otherwise, (ii) having passive potencies to be causally affected is not the only way for God to have (real, intrinsic) potencies, and (iii) we seem to have good reason to hold that God’s knowledge cannot be a mere Cambridge relation (and hence cannot be identical to God’s causing the world to exist, which is (so Classical theists (generally) hold) a mere Cambridge relation). I conclude, then, that the fifth argument retains its force.
Problem 6: Difference Principle (response by Nemes)
Here was the argument:
Difference Principle (DP): A difference in the effect presupposes a difference in the total cause (such that the difference in the total cause accounts for the difference in the effect).
(1) If classical theism is true, then DP is not true.
(2) DP is true.
(3) So, classical theism is false.
I’m actually not a huge fan of the difference principle — not because I think it’s false, but because the use of the term ’cause’ is accompanied by far too much conceptual baggage. In particular, one might think DP is obviously false given (even the possibility of) quantum mechanics and libertarian freedom. For under (some interpretations of) quantum mechanics, even under utterly identical initial conditions (it is claimed), a different end result can occur (say, going through the slit on the right as opposed to the slit on the left). Similarly, even under identical desires, brain states, reasons, and so on (it is claimed), an agent can exercise his or her causal power to do x as opposed to y (or vice versa).
So, let’s ditch DP for the Explanatory Difference Principle, EDP.
EDP: A difference in the explanandum presupposes a difference in explanatorily efficacious factors in the total explanans, such that the difference in the total explanans accounts for the difference in the explanandum.
Suppose the critics of DP are right: the various causal factors at play in the initial conditions can be fixed in the quantum realm (on the one hand) or the realm of free choice (on the other hand) while still allowing for different end results to obtain (an electron going through one slit as opposed to the other, or choosing chocolate ice cream as opposed to vanilla).
No fret for EDPist, however. Let’s take the case wherein I choose chocolate ice cream.
Now, why did I choose chocolate ice cream? There must be an explanation, lest we violate the PSR and threaten most arguments for a distinctively classical theism. Now, our explanation need not be contrastive — we need not explain why I chose chocolate rather than or as opposed to vanilla. But we still need an explanation of why I chose chocolate.
The explanation will presumably be like: (i) I desired chocolate ice cream, (ii) the means to fulfill my desire for chocolate ice cream are readily actualizable, (iii) I have some good reasons to eat chocolate ice cream (say, because I really enjoy chocolate ice cream, I haven’t had chocolate ice cream for a month, and I just biked 10 miles so I feel like I deserve a treat as a reward), (iv) I didn’t have stronger countervailing desires or reasons not to eat chocolate ice cream, etc.
These are all the elements, factors, or features within the explanans that are explanatorily efficacious with respect to the explanandum (my choosing chocolate ice cream). And we are carving nature at its joints when we cite these explanatory factors — their status as an explanation of the explanandum is real, objective, ‘out there’.
But suppose, with all the same causal conditions fixed, I chose vanilla. This is (we can grant) possible given libertarianism. Why did I choose vanilla in this situation?
Suppose Jones responds with the following explanation:
(i) Joe desired chocolate ice cream, (ii) the means to fulfill Joe’s desire for chocolate ice cream are readily actualizable, (iii) Joe has some good reasons to eat chocolate ice cream (say, because Joe really enjoys chocolate ice cream, Joe hasn’t had chocolate ice cream for a month, and Joe just biked 10 miles so Joe feels like he deserves a treat as a reward), (iv) Joe didn’t have stronger countervailing desires or reasons not to eat chocolate ice cream, etc.
What are you talking about, Jones? Seriously, what are you on about? That has nothing to do with my choosing vanilla. To cite the same explanatory facts as before is manifestly inadequate. We leave the explanandum unexplained. Instead, we need to cite my desires and reasons for choosing vanilla.
We can see, then, that the factors of the total explanans which are explanatorily efficacious with respect to the difference in explanandum (vanilla as opposed to chocolate) must be different in the two cases. You can’t cite reasons and desires for chocolate in explaining why I chose vanilla — it’s flatly irrelevant.
Similarly, suppose we shoot an electron at a slit and it goes through the right slit. Why?
Well, presumably, the electron has an indeterministic, probabilistic tendency or propensity of 0.5 (that is, 50%) for going through the right slit. And this tendency, though not a contrastive explanation, is explanatorily efficacious nevertheless. Now, suppose we re-wind the clock with all the same initial causal conditions, etc., but this time, the electron went left.
Jones wants to redeem himself, so he tries to explain why the electron went left:
Well, the electron has an indeterministic, probabilistic tendency or propensity of 0.5 (that is, 50%) for going through the right slit. And this tendency, though not a contrastive explanation, is explanatorily efficacious nevertheless.
Whelp, Jones failed, and he should probably be ashamed of himself. Again, a probabilistic tendency to go right has nothing to do with the electron’s going left. To cite the same explanatory facts as before is manifestly inadequate. We leave the explanandum unexplained. Instead, we need to cite the probabilistic tendency for the electron to go left.
So, I aver that, even if quantum mechanics and libertarian freedom undermine DP, EDP remains in tact (and, I would argue, actually supported by libertarian freedom and indeterministic quantum phenomena).
Why believe EDP?
Suppose you work in the ‘red division’ of a Play-Doh factory – the part dedicated to creating, forming, and packaging red Play-Doh. All you see every day is red Play-Doh coming out of the machines.
One day, something strange happens: one of the machines in your division produces blue Play-Doh. Your co-workers panic as they contact the division manager. The manager brings in a mechanic who, after examining the machine in detail for a few hours, reports on the situation.
“Well,” states the mechanic, “nothing is wrong with your machine. The mechanisms by which it operates and produces the blue Play-Doh are identical to the other machines in this division that reliably produce red Play-Doh. There are no differences in the circuits, wiring, tubes, input and output operations, the program, or anything in the machines themselves.”
“Good,” you respond relieved. “That must mean the machine is simply being fed different starting materials!”
“Not so fast,” replies the mechanic to your dismay. “I checked that too. The materials out of which the Play-Doh is made are also identical in this machine as in others. The same pigments, flour, and water are present in the same amounts and proportions. Every part of the machine, as well as every operation of the machine, as well as all the inputs to the machine — including Play-Doh raw materials — are identical in this machine as in others. In fact, I checked every environmental condition of this machine and compared it to the other machines – lighting, temperature, air pressure, and so on. The environmental conditions are identical.”
You start to wonder about the credentials and expertise of the mechanic. Surely, you reason, he must have missed something in the process. If absolutely nothing changed, the machine would spit out red Play-Doh just like the other machines in the division.
Why do we conclude that the mechanic’s inspection was inadequate instead of agreeing with the mechanic’s assessment? What undergirds this judgment?
I aver that EDP does. After all, if the total explanations in the two cases are utterly and completely (qualitatively) identical, there could be no ground or reason as to why the explananda differ. The difference in the explananda would be an inexplicable, brute occurrence.
In fact, there is a further sense in which the blue Play-Doh is inexplicable. If there was genuinely nothing in the total explanation that changed between the defective machine (and its inputs and environment) and the other machines (and their inputs and environments), then why was blue produced instead of orange? Blue is completely arbitrary. By the very nature of the scenario, blue’s appearance couldn’t be accounted for in terms of some different explanatory factor at play in the total explanation. There is therefore an additional level of arbitrariness and inexplicability: why would blue be formed rather than any other color? What could you point to as an explanation?
We can see, then, that EDP seems to enjoy abundant intuitive support. What else, though, can be said on its behalf? Below I offer at least seven reasons in its favor.
First, our universal experience attests to EDP’s being true. In particular, our universal experience of different explanada attests to their total explantia being different, and we have universal experiential support of things’ having identical explanda when their explanations are identical. For instance, qualitatively identical plants are produced when (i) their seeds are qualitatively (including genetically) identical, and (ii) their causally efficacious environmental conditions (such as amount and intensity of sunlight, soil conditions, water intake, etc.) are identical.
 Two notes. First, of course not everything in the two cases is identical. The two plants may be located in different portions of space, for instance. But this does not count against EDP, since merely being located in a different portion of space, at least in the case of plants, is not an explanatory factor contributing to their growth and development. Second, I will use ‘identical’ interchangeably with ‘qualitatively identical’.
Our phenomenological experience also seems to attest to EDP’s being true. Take, for instance, two successive days in the summer where you have the choice to snack on vanilla or chocolate ice cream. We can suppose that the two successive days are near copies of one another – almost identical meals, interactions, commute to and from work, and so on. Nevertheless, on the former day you chose vanilla whereas on the latter you chose chocolate.
The explanatory factors at play, then, must have been different (according to EDP). And this is precisely what we find: the difference in the explanandum (chocolate as opposed to vanilla) is accounted for in terms of different desires’ (and/or reasons’) being efficacious – in conjunction with the efficacy of your free will – with respect to the outcome. Alternatively, even if the choice is completely arbitrary (i.e. not accounted for in terms of specific desires and/or reasons), the difference in the explanandum is still accounted for in terms of a difference in the explanans: a different exertion of willed, agent-causal power.
Second, not only does EDP enjoy abundant inductive support, but it is also the best explanation for why there is such inductive support in the first place. It is a simple principle with explanatory breadth and depth. If EDP were false, for example, a puzzle arises: why don’t (say) plants that are genetically identical and that experience identical environmental conditions turn out wildly different? Why don’t we experience all around us a kind of chaos wherein wildly different effects or explananda are produced from the same ordinary causes and explanations with which we are familiar? The hypothesis that such wild chaos is simply impossible best explains why we never see it obtain in actuality. If it were genuinely possible, we get a puzzle: why does it never happen?
To make this concrete, take the striking of a match in normal living room conditions. Curiously, the struck match always produces heat and flame rather than frost, cold, a smell of lilacs, a kitten, Jones, or a banana. Why?
 For example, there is not an utter lack of oxygen, it isn’t raining, and so on.
Here is a simple answer: such differences in the explanandum couldn’t occur unless there was a difference in the total explanans. Because the total explanans is the same in each case (the match, the match’s being struck, and the environmental conditions), the explanandum is the same in each case (heat and flame).
By contrast, if this answer were false, we get a mystery. For then it would genuinely be possible that such differences in the explanadum could occur (on each occasion) despite there being no differences in the total explanans. It is therefore a mystery as to why it is always the case that no such differences in the explanandum occur.
Indeed, there is a powerful Bayesian argument for EDP. For simplicity, let’s focus on EDP as applied to matches. Now, billions of matches are produced and used each year. Let’s conservatively suppose that in all of industrial history, one billion such matches have been struck in normal living room conditions (such that the total explanation in each such case is roughly identical), all of which have (as far as I’m aware) produced heat and flame as opposed to frost and cold (or countless other things). On the hypothesis of EDP, the probabilistic expectation that all of these billion matches uniformly and reliably produce heat and flame is one (since, according to EDP, they could only differ (i.e. be non-uniform) in their explananda only if there is a difference in their total explanantia, and we have supposed arguendo that their total explanantia (the match, the striking of the match, and normal environmental conditions) are identical).
By contrast, take the hypothesis (~EDP) that there can be a difference in the explanandum without there being a difference in the total explanans. Per the principle of indifference, the epistemic probability (conditional on H) that any one such striking of the match will have the same explanandum as another will be 0.5. This, moreover, applies to each one of the billion matches. Considering the explanandum E of one match M, then, the conditional expectation that all of the other 999,999,999 matches uniformly produce the same E as M is 0.5999,999,999 – a staggeringly small number. Per the odds form of Bayes’ Theorem, the evidence of the uniformity of the explananda of the matches (given the same total explanantia) significantly confirms EDP over ~EDP.
 Why? Because there are two options: either the effect will be the same, or it will not. The hypothesis in question states that the effect genuinely can be either way regardless of the sameness of the total cause. The principle of indifference then advises us to distribute our credences equally among alternative possibilities (absent any special reason a priori to favor one possibility over the other).  One might worry about other, alternative hypotheses, like EDP’s truth for secondary causation/explanation but falsehood for primary causation/explanation. Arguably, though, this will just make the Bayesian argument re-arise at the level of primary causation, for the world is reliably uniform in a number of ways — and indeed persists in being that way. This is deeply surprising if there genuinely can be a difference in the explanadum (persistence in accordance with regularity) at any moment is possible with no difference in the explanans (God). I could formalize this, but it will look almost the same as the case with the matches.
A third reason favoring EDP stems from the Principle of Relevant Differences (PRD). Suppose we want to treat x and y differently in some moral way. We may want to privilege y but not x in some way, or we may want to preserve or protect y while disregarding x. However, it is only morally permissible to treat x and y differently if there is some relevant difference between x and y that justifies or accounts for their differing moral status. If there is no relevant difference between them such that it justifies differential treatment, then the differential treatment is simply immoral and unjustified.
This is one reason why racism is wrong: mere differences in skin color are irrelevant with respect to moral status, and hence to treat individuals morally differently merely in virtue of their skin color is to act wrongly. This is also why we can justify climbing, stepping on, and otherwise abusing rocks but cannot justify climbing, stepping on, and otherwise abusing (say) puppies. There is a relevant difference between the two that justifies such differential treatment: the capacities to suffer and to flourish.
But the same rational justification behind the PRD (as applied to ethics) seems to apply to EDP. It seems, in other words, that one cannot consistently accept and apply PRD in ethics without also accepting EDP. This is because the same sort of demand for a relevant difference-making feature applies to differences in explananda: the explananda can only differ if there is some relevant difference in their total explanations, just as our treatment of x and y can only (justifiably) morally differ if there is some morally relevant difference between them. The motivation for a relevant, underlying difference that grounds the discrepancy seems the same in both cases.
Fourth, the Principle of Sufficient Reason (PSR) lends support to EDP. For if the total explanation C of explanandum E is identical to the total explanation C* of explanadum E*, but E and E* are different, then the difference between E and E* is an inexplicable brute fact. This is because there is nothing in the explanans which accounts for the difference. This, in turn, seems to violate the PSR.
 The PSR states that necessarily, every fact (else: every contingent fact) has an explanation. See Pruss (2006) for a contemporary defense of the PSR and Feser (2017) for a survey of arguments for it. In this dialectical context wherein classical theists tend to accept the PSR, an extended justification of its rational grounds is unnecessary.
This argument for EDP, however, need not rest on the PSR. We only need some principle to the effect that, generally, some explanation is more probable than no explanation. There is, in other words, a defeasible presumption of the existence of an explanation absent any special reasons to the contrary. This alone provides evidence for EDP. Moreover, it is a theoretical vice to multiply inexplicability and arbitrariness. As we saw in the opening Play-Doh scenario, EDP minimizes both.
 This more limited principle helps provide a foundation for rational inquiry itself: if there being no explanation was just as likely as there being some explanation, then for any object of scientific or rational investigation O, deeper explanations that illuminate the nature, origin, development, or function of O would be just as likely as O’s having no explanation whatsoever.
Fifth, EDP helps form the bedrock of scientific investigation, as scientific investigation implicitly presupposes it. When scientists find discrepancies in the effects of their experiments, they assume that there must have been something different in the experimental conditions (such as one group being the placebo and the other being the treatment group). This drives their quest to understand the world and its diverse explanations and explanada.
Sixth, EDP helps ground the rationality and reliability of inductive inferences. If EDP is false, by contrast, it is unclear how we can avoid inductive skepticism. The inductive skeptic claims, for instance, that I do not know that a stone dropped from my waist tomorrow will fall. With EDP in hand, however, we can resist this claim. For the stone’s failing to fall tomorrow would constitute a difference in the explanandum (failing to fall as opposed to falling). But in order for such a difference to obtain, there would have to be – per EDP – some difference in the total explanation. Hence, if we are rationally justified in holding that no such difference in the total explanation obtains (or will obtain), we are rationally justified in holding that the explanandum will uniformly and reliably persist in accordance with the thus-far-established regularity. But if EDP is false, then it is genuinely possible – despite having identical explanations – that the explananda will radically differ in the unexamined cases.
 I mean both ontological and epistemological grounding. Ontologically, EDP provides an objective account of the dependence of the character of effects/explanans on the character of their causes/explanations (in particular, differences in the former depending on differences in the latter). Epistemologically, EDP can help us reason from knowledge of identical causes/explanations to identical effects/explananda or from knowledge of different effects/explananda to different causes/explanations.  The total explanation is something like (i) my dropping the stone from my waist, (ii) earth’s gravitational pull, and (iii) the absence of significant contributions from non-gravitational forces like air drag.
Seventh, EDP seems self-evidently true. The aim of the Play-Doh scenario was partly to illustrate EDP’s self-evidence. Having confirmed that the mechanic checked all relevant causal factors, you didn’t conclude the falsity of EDP. Instead, you concluded that the mechanic’s procedure must have missed some difference. You didn’t even stop to consider that the mechanic was correct.
 For a defense of arguments from self-evidence in the face of expert disagreement, see Pruss (2006), ch. 11.
Okay, but how does EDP relate to classical theism?
C1 and C2 refer to the total explanations of effects E1 and E2, respectively. The ‘non-God concrete world’ is the part of reality that, under classical theism, God creates and sustains (i.e. every concrete thing apart from God). The Contingency Thesis states that there is at least one thing in the non-God concrete world that is contingent (i.e. could have been different or otherwise).
 ‘Thing’ refers to one or more of: properties, events, relations, objects, substances, obtaining states of affairs.
Here, then, is the argument:
- A difference between E1 and E2 presupposes a difference between C1 and C2 such that the difference between C1 and C2 accounts for the difference in the explanandum. (EDP)
- There is at least one x such that (i) x is part of the non-God concrete world, and (ii) x could have been different. (Contingency Thesis)
- If there is at least one x such that (i) x is part of the non-God concrete world, and (ii) x could have been different, then God’s explanandum could have been different (where God is the explanation of God’s explanandum).
- So, God’s explanandum could have been different (where God is the explanation of God’s explanandum). (2,3)
- So, God could have been different such that the difference accounts for the difference in the explanandum. (1,4)
- If God could have been different such that the difference accounts for the difference in the explanandum, then either (i) God’s extrinsic relation(s) to something external to God could have been different, (ii) God has accidents (if the difference lies in accidental features), or (iii) God is contingent (if the difference lies in essential features).
- God (qua purely actual) can neither be contingent nor have accidents.
- So, God’s extrinsic relations to something external to God could have been different, such that the difference accounts for the difference in the effect. (5,6,7)
- But (8) is absurd.
- But if God (qua purely actual) exists, then (8) follows.
- So, God (qua purely actual) does not exist. (9,10) [Note that this is NOT concluding that God simpliciter doesn’t exist]
Why believe (9)?
Here is the key: ontologically prior to God’s act of creation, there was nothing external to God and hence no extrinsic relations in which he could stand. It follows, then, that extrinsic relations cannot be that which grounds (as a necessary condition) the difference in the effect (i.e. explanandum) of God’s act, since there being extrinsic relations in the first place already presupposes as a necessary condition the existence of the effect. It’s viciously circular.
If the existence of a different explanandum presupposes as a necessary condition a difference in the explanans (EDP), and the only candidate difference in God is a relation to external things (per (8)), but there being any relations to external things (in the first place) presupposes as a necessary condition the existence of God’s explanandum, then vicious circularity ensues. God’s standing in an extrinsic relation presupposes the existence of the effect/explanandum of God’s act; but per EDP, the very existence of such an effect/explanandum presupposes a difference in the cause, which (per (8)) could only be an extrinsic relation. Under classical theism, we therefore have vicious circularity: God’s extrinsic relation presupposes the effect/explanandum, but the effect/explanandum presupposes the extrinsic relation. Premise (9), therefore, is justified.
Whew! That was necessary to properly understand the argument and to properly respond to Nemes, as we will see.
In experience we notice the reoccurrence of certain things in nature. If this happens often enough, we can begin to think that we have come across some genuinely universal attribute of nature as such. At this point, we think not only that this attribute characterizes all of our past and present experiences of beings, but that it characterizes all possible experiences of beings as such. Sometimes this is true, but other times it is not. For some time Europeans believed that all swans are white, but later black swans were discovered in newfound Australia. What happened was that people forgot or were otherwise unaware of the intrinsically relative and perspectival nature of experience. Of course this does not mean that it is never actually possible to attain to the intuition of a genuine universal, but only that the process of eidetic intuition (as it is called) is fallible and subject to correction. For this reason, it is important to pay close attention to the origins of our knowledge and ideas — not because certain origins necessarily undermine or compromise their truth, but because such an inquiry can help us to gain a clearer understanding of the limits of our knowledge.
Nothing to object to here! Though, I see where you’re going…
This insight can be helpfully applied to the question of the nature of divine causality on a classical theistic scheme.
If God is absolutely simple as classical theism affirms, then He cannot be understood in the same way as finite beings. One thing we learn about finite beings in our experience is that they operate as causes by taking on various accidental modifications of their being. For example, Pat Metheny plays the guitar in virtue of various movements of his body and acts of will, from which he is distinct. Moreover, he produces different sounds as he plays because the position of his body is different in each case. This is called the “difference principle.” A difference in effect presupposes a difference in the cause. But because God is not a finite being like Pat Metheny, because He has no accidents and cannot be accidentally modified, He cannot be understood to function as a cause in that way. The “creaturely” concept of causation is no longer applicable. A different understanding must be developed, one which is appropriate for its object. And if classical theism is true, then God must be capable of acting as a cause apart from any accidental modification of His being. And if He can be a cause of something contingent, it must be possible for Him to produce His effect (or not) without sustaining any intrinsic change. So the “difference principle” cannot apply to Him.
There are lots, lots of things to say in response. First, note that EDP doesn’t require or presuppose accidental modifications to the explanans in the circumstances wherein a different explanandum exists. For instance, EDP would rest content with God having differential, essential tendencies to produce different worlds. Granted, this isn’t compatible with absolute simplicity, but it shows that appealing to accidental modifications as a symmetry breaker between God’s causality and the causality of the world, by itself, is insufficient.
Nor does the EDP require modifications of being. The EDP would rest content with a scenario as follows:
In all possible worlds, God has desires (or sets of desires) D1, D2, … Dn and reasons (or sets of reasons) R1, R2, … Rn to perform various possible (but incompatible) actions A1, A2, … An. These do not necessitate God to act as he does (thereby preserving God’s freedom amongst the range of Ai). Nor do they alter, modify, or actualize God in some way (since he tenselessly and timelessly possesses these, without actualization by anything, in all possible worlds).
 The set of desires D1 and the set of reasons R1 correspond to the possible action A1.
Now, suppose God did A1. Why did God do A1? Well, the non-contrastive explanation will (at least in large part) be in terms of D1 and R1. Say, why did God create cats (Ai)? Because he had a set of cat-relating desires Di and cat-relating reasons Ri. But, as we’ve seen with the EDP, had God created a world not containing cats, there must be some difference in explanatorily efficacious features of the cause (God) to account for this even if, in such a counterfactual world, God still possesses the cat-relating desires and reasons Di and Ri. It’s just that, in such a world, such desires and reasons are not explanatorily efficacious.
And this, in turn, entails that God cannot be absolutely simple (for there is a multiplicity of really distinct possibly explanatorily efficacious factors within God).
My point for including this is as follows: EDP doesn’t even require modifications of God’s being. We’ve seen, then, that it requires neither accidents nor modifications. All it requires is differential explanatorily efficacious elements, features, or factors.
Here are some final, brief remarks concerning the way of avoiding the argument from EDP by appeal to primary/secondary causality:
(1) As a necessary condition for avoiding the argument, it seems we need some relevant difference between primary causality and secondary causality which accounts for why EDP applies to the latter but not the former. We’ve seen that neither accidents nor modifications can serve that role.
 And (not that Steven did this) we can’t merely say that the relevant difference between primary causality and secondary causality (which accounts for why EDP applies to the latter but not the former) is that EDP, when applied to the former but not the latter, would entail the falsity of classical theism. For the truth or falsity of classical theism is precisely what’s on the line.
This is a sad example, but if someone were to demand a morally relevant difference between colored people and white people which could account for why it’s okay to treat the former worse than the latter, it obviously wouldn’t amount to a relevant difference to say ‘well, if there weren’t such a morally relevant difference between them, then that would entail that racism is wrong, whereas if there were such a morally relevant difference, no such entailment would arise.’ Yes — that’s the whole point. It’s precisely because there’s no morally relevant difference between them that racism is wrong. You can’t appeal to this entailment as a relevant difference between the cases, since this entailment simply follows immediately upon there being no relevant difference between the cases. It’s a subtle form of question-begging.
Similarly, if someone were to demand a relevant difference between primary and secondary causality which could account for why EDP applies to the latter but not the former, it obviously wouldn’t amount to a relevant difference to say ‘well, if there weren’t such a relevant difference between primary and secondary causality, then that would entail that classical theism is false, whereas if there weren’t such a relevant difference, no such entailment would arise.’ Yes — that’s the whole point. It’s precisely because there’s no relevant difference between them that classical theism is false (so the argument goes, that is). You can’t appeal to this entailment as a relevant difference, since this entailment simply follows immediately upon there being no relevant difference between the cases. It’s a subtle form of question-begging.
Indeed, proponents of primary causality can deny classical theism, so there seems to be nothing intrinsic to primary causality as such that demands the falsity of EDP.*
* Primary causality is (inter alia) a radical bestowal of being to another ex nihilo. By itself, a theistic personalist could accept that God’s form of causality is primary while the secondary causes operative in nature are unlike God’s causality.
(2) Internal probabilistic tension
Suppose you appeal to a variety of reasons to think that every wholly contingent state of affairs has a cause. If this is right, we can infer the existence of a necessary being: simply consider the maximal wholly contingent state of affairs (the state of affairs which includes all other wholly contingent states of affairs in the world). This maximal state of affairs has a cause (per our causal principle), but the cause cannot be a state of affairs involving only contingent beings (since then it would be included within the very thing in need of a causal explanation, which is viciously circular). So, the cause must include at least one necessary being.
But suppose I respond as follows: well, I only accept that every non-maximal wholly contingent state of affairs has a cause. And this disallows the inference from maximality to a necessary being.
One problem with my response is that there is internal probabilistic tension within my framework, for the applicability and universality of the causal principle within the domain of non-maximal wholly contingent states of affairs raises the probability (i.e. gives us a greater degree of expectation) that the same causal principle applies to maximal wholly contingent states of affairs. The internal proviso makes the proposal less probable, since the applicability of the principle in a whole domain gives us reason to expect its applicability to another domain.
 Of course, the reason is defeasible, but it is a reason.
What follows is that if EDP applies universally in the domain of secondary causality (indeed, in the realm of all non-divine explanations), then we have reason to expect (i.e. it is more probable, all else being equal) that EDP likewise applies in the domain of divine explanations. To claim otherwise is to adopt a view according to which there is internal probabilistic tension (and hence a lower intrinsic or prior probability).
 This is ignoring, of course, the variety of independent reasons given for EDP’s universal truth simpliciter.
(3) The move threatens to collapse cosmological arguments that serve as the very foundation upon which classical theism’s merits rest.
Cosmological arguments almost unanimously rest on some general/universal causal principle inferred from universal experience, intuition, abduction, and so on. The thing is, it seems that any rational support for such universal causal principles will likewise support EDP. Intuition? Abduction? Bayesian arguments? Induction? Demand for explanation? Whatever it may be, it seems that any justification mounted for the causal principles in cosmological arguments can likewise be mounted in favor of EDP. But if that’s the case, then there seems to be no principled, non-arbitrary, non-question-begging way to affirm such causal principles while denying EDP.
Another way to see the same point is through the lens of the proposed defeater of EDP on the basis of primary/secondary causality. The classical theist might claim: EDP applies to secondary causality, but not primary causality. But then what principled, non-arbitrary, non-question-begging reason do we have for ruling out the atheist’s response to cosmological arguments like this:
“That general causal principle you adduce? That applies only to secondary causality, not primary causality. But in order to infer God’s existence from the general causal principle, you apply it to God’s causality (of the world, or of some sub-set of it). So, the general causal principle cannot be used to infer God’s existence.”
It seems, then, that denying EDP’s universal applicability may come at the cost (though, there are many costs, as I’ve outlined) of undermining the very rational foundation for classical theism (cosmological arguments).
I conclude, then, that the sixth argument retains its force.
Problem 7: Existential Inertia (response by DeRosa)
Here was the original, unrefined argument (this is not how it is presented in my paper on existential inertia, of course):
(1) If classical theism is true, then both TST is true and EIT is false.
(2) We don’t have adequate reason for thinking both TST is true and EIT is false.
(3) So, we don’t have adequate reason for thinking classical theism is true.
The logical form of this problem is as follows:
- If X is true, then Y is true and Z is false.
- We don’t have adequate reasons for thinking both Y is true and Z is false.
- So, we don’t have adequate reasons for thinking X is true.
A parody might go like this:
- If it is raining, then Bob is using an umbrella and not wearing a hat.
- We don’t have adequate reasons for thinking Bob is using an umbrella and not wearing a hat.
- So, we don’t have adequate reasons for thinking it is raining.
(3) does not follow from (1) and (2) since there could be other grounds for thinking it is raining apart from whether we have adequate reasons for thinking Bob is using an umbrella and not wearing a hat. Perhaps we actually see that it is raining outside, but have not seen or heard from Bob in days and cannot judge anything based on his attire.
Let’s extend the principle of charity, though. I thought it was reasonably clear that I was invoking a version of the epistemic closure principle. More particularly, my aim was to reason as follows:
- We know that if classical theism is true, then both TST is true and EIT is false.
- Suppose (for reductio) we know classical theism is true.
- Well, if we know that classical theism’s truth entails (TST and ~EIT), and we know classical theism is true, then surely we are (in a position to be) justified in holding (TST and ~EIT).
- But we are not (in a position to be) justified in holding (TST and ~EIT).
- So, we don’t know classical theism is true (1-4).
This argument is perfectly valid. Now, we may object to the closure principle in (3), but I aver that (3)’s truth is deeply intuitive. If I know that Socrates is a man, and I know that whatever is a man is mortal, then surely I am (in a position to be) justified in holding that Socrates is mortal.
 Admittedly, it is my fault for not talking about or making explicit the epistemic closure principle. Though, keep in mind that I intentionally made my first post short and underdeveloped — that was baked in from the start.  See the note in my original post for a better interpretation of this premise that takes into account the fact that justification is person-specific.
What’s more, the parody argument isn’t a good one and actually serves to illustrate rather than undermine my point. For we must recognize that we know that ‘if it is raining, then Bob uses an umbrella and doesn’t wear a hat’. And so if we know that it is, in fact raining, then we would simply be improperly attentive to our beliefs and knowledge if we didn’t thereby know that Bob is using an umbrella and not wearing a hat. It doesn’t matter if there are other grounds for thinking it’s raining apart from Bob’s using an umbrella. What matter is that we know that if it’s raining, then Bob is, in fact, using an umbrella. So even if we come to know that it’s raining by looking outside, that doesn’t at all mean we don’t know Bob is using an umbrella. On the contrary, it is precisely because we know that ‘it’s raining entails that Bob is using an umbrella’ that we do know that Bob is using an umbrella.
Note, moreover, that the argument really has nothing to do with knowing something by other grounds. When we know ‘P implies Q’, we don’t commit ourselves to saying that the only grounds for knowing P is knowing Q. Indeed, that would be a sort of ‘epistemic affirming the consequent’. So, DeRosa’s claim that “there could be other grounds for thinking it is raining apart from whether we have adequate reasons for thinking Bob is using an umbrella and not wearing a hat” is simply irrelevant to the inference. The inference does not rely on the claim that the only way we could know whether it’s raining is if we know Bob’s using an umbrella. That’s like affirming the consequent. Rather, the argument is a form of denying the consequent. Sure, there may be other grounds for thinking it’s raining, but that doesn’t stop us from reasoning in the slightest like:
“Well, I know that if it’s raining, Bob is wearing an umbrella. And I know that Bob isn’t wearing an umbrella. But surely if I knew it was raining, I WOULD thereby know that Bob is wearing an umbrella provided I was properly attentive to my beliefs and knowledge. Given that I’m a good epistemologist, I am properly attentive to my beliefs. So, it follows that I don’t know it’s raining. This is simply modus tollens.”
What’s worse, DeRosa states this: “Perhaps we actually see that it is raining outside, but have not seen or heard from Bob in days and cannot judge anything based on his attire.”
But if that’s true, then we don’t actually know that premise (1) in the parody argument is true. But we have supposed from the get-go that we actually know that rain is a sufficient condition for Bob’s using an umbrella. Given that we know it is such a sufficient condition, then even if we come to know that it’s raining by means of looking out the window, we do (or at least are in a position to) know that Bob is using an umbrella. To deny that is to deny that we knew in the first place that rain is a sufficient condition for Bob’s using an umbrella.
So, the parody is no good.
Similarly, one can have other grounds for holding to classical theism independent of adequate reasons for thinking TST is true and EIT is false. One might hold to classical theism based on philosophical arguments that do not depend on the falsity of existential inertia. For example, one might endorse a contingency argument for an absolutely necessary being and then conclude that such a reality must be absolutely simple.
Or, one might hold that classical theism is the best explanation of a wide array of biblical data. Or, one might hold to classical theism based on ecclesiastical authority that one has come to trust. Or, one might hold belief in the God of classical theism is properly basic. Or, one might endorse classical theism based on a combination of those considerations. So, classical theists can point to the non-sequitur as a resolution to the apparent problem.
Sure, but this doesn’t target the inference, since this is again interpreting my argument as a species of affirming the consequent when, in actuality, it is a species of denying the consequent (modus tollens).
 For those interested, there seems to be no good inference from ‘necessary being’ to ‘purely actual’ or ‘absolutely simple’ being. For a defense of this, see this post.
DeRosa cites Feser for further reading on the topic:
For further reading on this topic, one may consult:
Feser, E. (2011). Existential Inertia and the Five Ways. American Catholic Philosophical Quarterly, 85(2), 237-267. This article opposes the existential inertia thesis.
Let’s evaluate Feser’s two central arguments that directly target existential inertia from that paper.
Feser’s first central argument against EIT derives from the Principle of Proportionate Causality (PPC), according to which a total cause cannot give to an effect what it does not have to give in the first place. More precisely, the PPC holds that whatever exists in an effect E must exist in the total cause of E in some manner (formally, virtually, or eminently). With this principle in hand, Feser argues:
- A cause cannot give what it does not have to give.
- A material substance is a composite of prime matter and substantial form.
- Something has existential inertia if and only if it has of itself a tendency to persist in existence once it exists.
- But prime matter by itself and apart from substantial form is pure potency, and thus has of itself no tendency to persist in existence.
- And substantial form by itself and apart from prime matter is a mere abstraction, and thus of itself also has no tendency to persist in existence.
- So neither prime matter as the material cause of a material substance, nor substantial form as its formal cause, can impart to the material substance they compose a tendency to persist in existence.
- But there are no other internal principles from which such a substance might derive such a tendency.
- So no material substance has a tendency of itself to persist in existence once it exists.
- So no material substance has existential inertia (Feser, 2011, p. 258).
There are at least four worries for this argument. First, it presupposes a controversial metaphysical account of the nature of substances. Indeed, the argument – if successful – establishes that “whether it [hylomorphism] is correct depends in part on whether things have existential inertia in the first place,” for if they have existential inertia, then (per Feser’s argument) hylomorphism is false (Audi, 2019, p. 7).
Second, consider chemical reactions in which two reactant species are each (individually) necessarily and essentially colorless, but yet when mixed together produce a vibrant red color. Although each individual thing within the total cause of the vibrant red color is essentially colorless, the combination of the individual things within the total cause nevertheless produces a vibrant red color.
This is not a proposed counter-example to the PPC. Instead, it reveals that there are ways that features can be present in total causes that Feser’s argument neglects. In particular, features can be present within total causes in a way I shall term conditional potencies. O possesses a conditional potency for F provided that O, when conjoined with some other condition or thing O*, gives rise to a system (O-O*) that manifests F. The red was not actually or formally present in the total cause (the two chemical species), but it was nevertheless present in the total cause as a conditional potency of each reactant species. The first reactant species had the conditional potency, when combined with the second species, to produce red; likewise with the second species.
We can now apply this to Feser’s argument. Specifically, merely from the fact that neither prime matter nor substantial form (of themselves) can have a tendency to persist in existence, it does not follow that their composition cannot have a tendency to persist in existence – any more than the fact that neither of the two reactant species can (of themselves) manifest redness entails that their composition cannot manifest redness. In the case of the chemical species, their composition can manifest redness precisely because each component has the conditional potency to manifest – when combined with the other – redness. Similarly, Feser’s argument neglects the fact that a form-matter composition may be able to manifest a tendency to persist in existence because each component has the conditional potency to, when combined with the other component, manifest such a tendency. By illegitimately presupposing that neither form nor matter could have such a conditional potency, Feser’s argument does not succeed.
Third, depending on how we understand ‘principle’, premise seven is arguably question-begging in this dialectical context. For whether or not there is an additional ‘principle’ of material substances (namely, existential inertia or a tendency to persist in existence) is precisely what is at issue. It is precisely the question at hand whether or not form and matter are the sole principles of material substances, since it is precisely the question at hand whether there is an additional principle (existential inertia) which accounts for the persistence of things in existence. Hence, Feser’s argument seems question-begging.
 Unfortunately, Feser does not define or explicate the notion (indeed, it seems to be a conceptual primitive in the Aristotelian framework).
Fourth, the argument (if successful) does excessive damage, as it entails the falsity of EET. Consider the following parallel argument. The only two principles of material substances are form and matter. But by itself, matter is pure potentiality and so doesn’t actually exist, in which case it cannot – of itself – impart any tendency to expire. And by itself, form is a mere abstraction and so doesn’t actually exist, in which case it cannot – of itself – impart any tendency to expire. From this, it follows (per Feser’s understanding of PPC) that no material substance has a tendency to expire. But such a tendency is precisely what Feser’s Aristotelian proof requires!
Feser may respond that EET just is the absence of a tendency to persist, and hence establishing that things lack such a tendency vindicates EET. There are two problems with this response. First, this is false — not possessing a tendency to persist neither means nor entails possessing a tendency not to persist (which is EET as defined in my previous post). Second, an exactly parallel reply can be given on behalf of EIT. In particular, the parallel argument showed that no material thing possesses a tendency to expire. So if Feser claims EET is vindicated by the absence of a tendency to persist, it will follow that EIT is equally vindicated by the parallel argument that things have no tendency to expire.
Feser’s second main criticism of EIT is form-matter interdependence. “For since in purely material substances matter depends on form and form depends on matter,” writes Feser, “we would have a vicious explanatory circle unless there was something outside the form/matter composite which accounts for its existence” (Feser, 2011, pp. 247-248).
First, even if it is true that, because they depend on one another for their actual existence at each time t at which they are conjoined, form and matter at t require an explanation outside themselves for their actual compositional existence, this is compatible with my accounts of the metaphysics of existential inertia. Consider the account according to which O-at-t-1 explains the existence of O-at-t. In this case, we avoid vicious explanatory circularity, since we are not explaining the form of O-at-t by the matter of O-at-t (or vice versa); instead, we are explaining O-at-t by O-at-t-1, which amounts neither to self-causation, nor to self-explanation, nor to vicious explanatory circularity.
Finally, consider again conditional potencies. Arguably, although form and matter may interdepend with respect to the beginning of a substance’s existence (i.e. with respect to the origination of the form-matter composition), it does not follow that they thereby interdepend at every moment at which they exist. This is because – for all Feser has shown – there may very well be a conditional potency within each that accounts for why, when combined with the other, they are able to manifest some further feature (namely, a tendency to persist). Consider again the case of the chemical species. The first species will not manifest redness unless the second is present, while the second will not manifest redness unless the first is present. But all this demands is an explanation for why they combined in the first place, since that original composition is what actualized the conditional potencies to transition into a state of actuality. And once the composition’s components have their conditional potencies actualized in the first place, they remain in a state of actuality unless separated by (say) some chemical or physical process.
“But,” one may object, “surely that is the very question at issue – namely, whether conditional potencies, once actualized, remain in a state of actuality with respect to one another.” This is true. But this shows that we cannot (in a non-question-begging manner) assume from the get-go an answer either way. In particular, we would beg the question if we assumed from the get-go that conditional potencies, once actualized, do not remain in a state of actuality. But such a presupposition is precisely what Feser needs for his form-matter interdependence argument to succeed. For if form and matter interdepend but also (individually) have the conditional potency to persist in existence once combined, and if conditional potencies, once actualized, remain in a state of actuality, then it is simply false that vicious circularity ensues in our explanation of the present existence of some substance. This is because the explanation of the present existence would not be in terms of form’s dependence on matter and matter’s corresponding dependence on form. Rather, the explanation would be in terms of (i) the cause of the origination into existence of the substance (and thereby the composition of the matter and form), (ii) the actualization of the requisite conditional potency within form and matter, (iii) the nature of conditional potencies (namely, to remain in a state of actuality once actualized), and (iv) there being no sufficiently destructive causal factors operative.
What about essence-existence distinction?
Merely from the fact that X’s essence and existence at t are distinct does not entail that X requires a per se, concurrent, efficient, sustaining existential cause at t. At best, all it requires is that X at t requires an explanation for why X exists. And that could be in terms of the state and existence of X immediately temporally prior to t in conjunction with no sufficiently destructive causal factors operative, or it could be in terms of that which brought X into existence in conjunction with an existential inertial tendency, or what have you.
Juarez (2018) gives a variant of this argument:
- In X essence and existence are distinct.
- A being in which essence and existence are distinct is a being whose existence is accidental to its essence.
- A being whose existence is accidental to its essence depends upon another for its existence.
- So X depends upon another for its existence.
- X’s dependence upon another for its existence obtains at any moment at which its existence is accidental to its essence.
- So, X depends upon another for its existence at any moment at which X exists (2018, p. 26).
There are a number of worries for this argument, some of which we have already canvassed. First, it presupposes existence is an additional, accidental feature or property of things. But should we accept this account of existence? Why think existence is a feature or property? No justification has been given for this presupposition.
Second, the consequent of (2) simply doesn’t follow from the antecedent. Merely from the fact that Y and Z are distinct, it doesn’t follow that one is accidental to the other. Consider, for instance, the properties having a circumference and having a diameter. These properties are clearly distinct; however, it is impossible to have one without the other – it is simply false that one is accidental to another. The fact that Y and Z are distinct, then, neither means nor entails any of the following: (i) Y doesn’t entail Z, (ii) Z is accidental to Y, (iii) Z is not necessarily ‘conjoined’ with Y, and so on. Hence, merely from the fact that a being is such that its essence is distinct from existence, we cannot infer that existence is accidental to it, or that it can fail to exist, or that it requires a concurrent cause to combine its essence and existence.
For existential inertia
It might surprise people, moreover, that there are actually arguments in favor of EIT. I present a Bayesian argument in my paper (and briefly in this video), but here is a separate argument for EIT.
I aver that the very nature of per se chains gives us reason to favor existential inertia.
To see why this is the case, let’s consider one of Feser’s examples of a per se causal chain. In explaining why a coffee cup is three feet above the floor, Feser writes that “it is sitting there at that moment only because the desk is holding it up at that moment, and the desk is holding it up at that moment only because it is in turn being held up, at that same moment, by the floor” (2017, p. 21). Other examples of per se causal chains include a lamp being held aloft by chains, in turn being held aloft by the ceiling, and so on; the stone moved by the stick, in turn moved by the hand, and so on; and gear one being turned by gear two, gear two being turned by gear three, and so on.
Notice, though, that in each of these chains, the only reason concurrent causal sustenance is required is that, absent such sustenance, there is some ‘net force’ or ‘net causal factor’ that is causally contributing to a single, definite outcome. In other words, the causal operation of the sustaining, per se cause C is required precisely because C acts against what would otherwise be a net causal factor towards some different outcome.
 I don’t mean net force in an expressly mechanistic or physical way (although such forces are sub-categories of what I mean). Instead, I just mean a causal factor or group of causal factors whose overall causal contribution is like a vector quantity insofar as it contributes toward a definite end state or outcome and is not counterbalanced by some other (group of) causal factor(s).
To see why this is the case, consider again Feser’s example. It is precisely because – absent the table’s existence – the cup would revert to the ground that it requires causal sustenance to remain in the air; and this, in turn, is because there is a net causal factor (namely, gravity) operating on the cup that the table is actively preventing from achieving its definite causal outcome (in this case, attraction toward the center of Earth’s mass).
 The same applies to the other examples of per se chains. For instance, the stone has net causal factors operating on it so as to keep it stationary (friction, gravitational and normal forces, and so on). A concurrent sustaining cause of the stone’s motion is required precisely because such a cause contravenes the causal activity of the friction, gravity, etc. toward the definite outcome of stationary spatial position.
There is inherent to per se chains, then, a requirement of either (i) a net causal force that causally inclines things toward outcome ~O that is contrary to the outcome O produced by the causally sustaining intervention, or (ii) a natural tendency, inclination, or disposition of a thing toward ~O that is actively being suppressed by the causal sustainer in order to maintain O.
With the requisite groundwork laid, we can level the following argument:
- A per se, sustaining cause C is required for substance S’s being in condition or outcome O only if (i) there is some causal factor or force F – intrinsic or extrinsic to S – acting on S to bring S toward some condition or outcome ~O; (ii) F is a net factor or force in the absence of C’s causal operation; and (iii) S (or some state of affairs involving S) is in condition or outcome O distinct from ~O.
- So, a per se, sustaining cause C is required for S’s actual existence only if (i) there is some F acting on S to bring S toward non-existence; (ii) F is a net factor or force in the absence of C’s existential sustenance; and (iii) S actually exists such that actual existence is distinct from the condition or outcome of S’s non-existence. (1)
- But there seems to be no adequate justification for holding (i) and (ii).
- So, understood in terms of epistemic closure, there seems to be no adequate justification for holding that a per se, sustaining cause is required for S’s actual existence. (2,3)
 An intrinsic causal factor or force would be something like a natural tendency or disposition inherent to a thing; an extrinsic one would be something like the effect of gravity, friction, and so on.
One preliminary worry for the argument concerns premise one. In particular, a classicl theist might hold that it is not the presence of some internal or external net causal factor inclining S towards some ~O distinct from the O in which S actually finds itself that spawns the requirement of a concurrent sustaining cause. Rather, the relevant feature of S (or state of affairs involving S) that spawns the need for a per se sustaining cause is that S by itself has no capacity or tendency to be in outcome or condition O and hence requires a concurrent cause to keep S in O.
Arguably, though, this inadequately describes the scenario, since it ignores the fact that (e.g.) the cup – absent any gravitational force pulling it down – simply retains its spatial location without causal sustenance. Consider astronauts on the ISS who, upon placing a cup in location L, observe the cup remain in L without any causal factor sustaining it there. Indeed, this illustrates rather than undermines my point concerning the nature of per se chains. For the only reason the cup would fail to remain three feet above ground (on Earth) is because there is a net causal force acting on it to pull it toward a given outcome. And the reason it is in fact able to remain three feet high even in the presence of (what would otherwise be) a net causal factor is because some C is actively concurrently preventing the net factor from eliciting its characteristic outcome. C accomplishes this by providing a causal force or factor towards an outcome contrary to that dominant in C’s absence.
 C would therefore be like a vector quantity that counterbalances in the opposite ‘direction’ of what would otherwise be a net causal factor/force.
But absent a tendency or causally inclining factor toward either O or ~O, S will simply remain in the condition or state in which it is in, as there would be no reason or explanation as to why it deviated away from the outcome in which it actually finds itself. And this is precisely what the cup on the ISS reveals: although the cup by itself has no capacity or causal inclination to be in any particular location, it will nevertheless remain in the actual location L in which it finds itself without requiring an external causal factor keeping it there. And again, this is precisely because (i) the cup is presently in L; (ii) any deviation from L would be inexplicable in the absence of a tendency or causal factor inclining the cup away from L; and (iii) there is no such tendency or causally inclining factor operative.
What the foregoing analysis reveals, then, is that a given substance S could lack a tendency either way (neither towards persistent existence nor existential expiration/annihilation) – and hence once placed in condition O (i.e. once brought into actual existence), S will simply remain in O as a form of stasis, rather than change. Its continually occupying such an outcome or state will simply not involve a reduction of potency to act but will instead simply be a persisting state of stasis or actuality. And this follows simply upon the nature of per se chains and the nature of explanation: any deviation from the actual state, condition, or outcome O in which S finds itself would be inexplicable in the absence of (i) a tendency toward ~O or (ii) a net causal factor inclining S towards ~O.
Feser, though, and many other classical theists have given us little to no justification as to why there exists either (i) a tendency of things to expire or annihilate, or (ii) a net causal factor ‘pulling’ or ‘inclining’ things toward non-existence at any moment at which they exist. And as we have seen, this is precisely what needs justification in order to affirm that S’s existence requires a per se, sustaining cause.
I conclude, then, that the seventh argument retains its force.
Problem 8: Presentism and Causal Sustenance (response by Tomaszewski)
For an elaboration of the original problem, see the original post.
It seems to me that there are actually two problems being presented here. The first one is the objection considered by Lombard, according to which God is where He was not before when new creatures come into existence, and therefore changes. The second one is Mullins’ own objection from presentism to the conjunction of the claims that God is absolutely simple and that He conserves every being in existence every moment that it exists. As far as I can tell, these problems are distinct, and it’s not clear to me why Mullins believes that the passage he quotes from Aquinas is supposed to be responsive to the problem considered by Lombard.
I am fine with this assessment, my dude!
In reply to the first objection, considered by Lombard, it is sufficient to note that God’s coming to be present in a creature constitutes a change in Him if and only if such presence is intrinsic to Him. But why should the Thomist (or anybody for that matter) believe that God’s presence in His creatures is an intrinsic feature of Him? It is hard to determine a good test for intrinsicality, but arguably, a property is intrinsic only if (but not if) a subject could have that property without anything else existing. But clearly, being present in one’s creatures is not a property God can have without anything else existing. I think this solution is in line with what Lombard would have said if he had given a fuller exposition of his solution.
I’m in agreement with Tomaszewski here — Lombard’s argument can be met. The prima facie problem in my post was mainly that proffered by Mullins (though Tomaszewski has certainly brought value to the discussion in considering Lombard).
The second objection, proposed by Mullins, insists that God “cannot act at non-existent times”. It is of course perfectly true that God cannot act at non-existence times. Indeed, He cannot act at any time, because His act is H
is essence, and His essence is eternal. But God’s conservation of His creatures in time is an effect of His intrinsic act. And Mullins gives us no reason to disbelieve what Aquinas says in Summa contra Gentiles II.35: “Nothing, therefore, prevents our saying that God’s action existed from all eternity, whereas its effect was not present from eternity, but existed at that time when, from all eternity, He ordained it.” This, and not the passage quoted by Mullins, would be Aquinas’ response to the objection given by Mullins.
It’s unclear that this response works, though (however, it’s also unclear that it doesn’t work).
I think there are two primary ways we can understand Mullins’ argument.
One way is like this (my formulation):
1. God’s causing x to exist is sufficient for x’s existence, and x only exists if God causes x to exist.
2. If (1), then whenever the condition ‘God causes x to exist’ is met, x exists or obtains.
3. So, whenever the condition ‘God causes x to exist’ is met, x exists or obtains.
3. But the condition ‘God causes x to exist’ is met timelessly and eternally — God has a single, timeless act of creation whereby he brings about his effects.
4. If it is the case that (i) whenever the condition ‘God causes x to exist’ is met, x exists or obtains, and (ii) the condition ‘God causes x to exist’ is met timelessly and eternally in a single act, then God’s temporal effects exist or obtain eternally and timelessly.
5. So, God’s temporal effects exist or obtain eternally and timelessly.
6. If (5), then eternalism is true.
7. So, eternalism is true.
I’m not entirely sure what to make of this argument. I’m not sure it works, but I’m also not sure it doesn’t work. Perhaps there are subtle questions begged against the classical theist that I am (at present) missing.
For instance, while God’s causing x to exist is sufficient for x’s existence, perhaps God causes x to exist qua past, present, or future in a dynamic timeline. So, God’s act of causing the past, present, and future to exist — while all being a single timeless act — is only causally sufficient for them to exist qua the particular manner or mode in which he wills them — and he wills them qua under different ‘manners’ or ‘modes’ of past, present, or future.
But then again, perhaps this misses the point — the point is that God’s causing them qua past, present, or future has ‘already’ happened (as it were). It would seem that this could only account for a different, B-theoretic ordering of the events, not for a genuine dynamism in their existence. Again, God causes them qua their temporal positions in an eternal, timeless, single act — and it seems that such an act is clearly sufficient for the effects to obtain.
Once again, though, I’m undecided on this interpretation.
Perhaps a distinction at play here is that between:
A: at t, God causes x to exist.
B: God causes x to exist at t.
The former entails a temporal conception of God; God acts in time to bring about the effects. The latter however does not entail this; all it entails is that the effect is in time.
So, maybe the classical theist would reply that D does not follow from C:
C: timelessly, God causes x to exist.
D: God causes x to exist timelessly.
Perhaps. I’m undecided as to whether that this adequately targets the argument, though, since the argument more-so focuses on causal sufficiency and the identity of God’s causing the past and future to exist. But perhaps it illicitly rests on an inference from C to D. Food for thought.
A second way to understand the argument is an appeal to intuition and an invitation to us to consider the force of the intuition. The following conjunction (so Mullins argues) seems intuitively strange: (i) God causes me to presently exist, (ii) God also causes me to exist tomorrow, (iii) but I do presently exist but don’t exist tomorrow, (iv) but yet somehow the act whereby God causes these two things is utterly identical.
Once again, though, all this serves to reinforce the fact that it’s unclear whether Tomazewski’s response succeeds, since there seem to be interpretations of the argument that might not succumb to the worries he expresses. In that case, though, it’s also unclear that the original argument succeeds.
I conclude, then, that whether the eighth argument retains its force is very unclear.
Problem 9: Identity and Potency (response by Tomaszewski)
Here was the original argument:
(1) If God is free, then God can create or not create.
(2) If God can create or not create, then God can be such that he has a creative act or he does not have a creative act.
(3) But if God has a creative act, God is identical to his act of creation.
(4) If God does not have a creative act, then he cannot be identical to his act of creation (for he wouldn’t have an act of creation with which to be identical!).
(5) So, if God is free, then God can be such that (i) he is identical to his act of creation, or (ii) he is not identical to his act of creation.
(6) If x can be such that (i) x is identical to y, or (ii) x is not identical to y, then x has (intrinsic) potencies.
(7) So, if God is free, God has (intrinsic) potencies.
(8) So, it cannot be the case that God is both free and purely actual.
(9) If classical theism is true, then God is both free and purely actual.
(10) So, classical theism is false.
This argument is unsound because premise (6) is false. Here is premise (6): “If x can be such that (i) x is identical to y, or (ii) x is not identical to y, then x has (intrinsic) potencies.” And here is a counterexample to (6): Aquinas is such that he could have not been identical to the Angelic Doctor, and this would not require any intrinsic potencies in him (though, of course, Aquinas had intrinsic potencies). For what makes Aquinas identical to the Angelic Doctor is just the fact that we use “the Angelic Doctor” to refer to Aquinas, and that is a fact about us, not about Aquinas. Had the Church not bestowed this lofty sobriquet upon Aquinas, it would be false that Aquinas is identical to the Angelic Doctor. In neither scenario do we have any need to appeal to any of Aquinas’ intrinsic potencies.
I actually agree with Tomaszewski that the example provided is indeed a counter-example to the premise as stated, but the argument can be reformulated to avoid the objection.
Here is a reformulation:
(6*) If it is the case that x can be such that (i) x is identical to y, or (ii) x is not identical to y, and if it is the case that y is not merely an expression we use to refer to x, then x has (intrinsic) potencies.
Notice that the example with Aquinas is no longer a counter-example to (6*).
The question, on (6*), is now: is God’s being identical to his creative act a mere fact about us and our linguistic expressions? It seems not. For God’s act is out there (as it were). God’s creative act — his exercise of causal power — seems to be a genuine, objective feature of extramental, extralinguistic reality.
Here is another counterexample: Venus is such that it could have not been identical to the Evening Star. But obviously, this does not posit any intrinsic potencies in the planet Venus. It merely posits an intrinsic potency in us to have given Venus a different name or no name at all.
Well, as Tomaszewski is well aware, whether Venus is such that it could have not been identical to the Evening Star depends on whether we treat ‘the Evening Star’ as a rigid designator (a proper name) or a description. If we treat it as a description, such as ‘the heavenly body of such-and-such color, shape, etc. that is visible in the evening”, then it’s true that Venus is not necessarily identical to this description. But if we treat it as a rigid designator, then Venus couldn’t have failed to be identical to the Evening Star.
 Here is another example to illustrate Tomaszewski’s point. The president of the United states (a description) is identical to Donald Trump. But is could have been identical to Hillary Clinton had she won the election.
Now, (6*) essentially states that when X satisfies a description but could have failed to satisfy that description, but where the description in question is not a mere artifact of our linguistic expressions (but instead has its basis (as it were) ‘out there’, in extramental and extralinguistic reality), then X has potency.
And this seems to be supported by examples that involve descriptions which are not merely linguistic. For instance, Neil Armstrong (a proper name) is identical to (the description) the first mammal on the moon. But Armstrong could have failed to be identical to such a description (say, if he became a philosopher instead of an Astronaut). And such a description is not merely linguistic — whether a mammal satisfies that descriptions is a fact ‘out there’, it seems, unlike whether someone is deemed an Angelic Doctor by the church. And it seems that Armstrong could have failed to satisfy the description precisely in virtue of possessing variegated potencies, some of which he simply didn’t actualize (like the potency to become a philosopher, or a milkman, or a meme expert).
 Here is a technical complication. Armstrong also could have failed to satisfy the description due to something wholly extrinsic to him as he is in himself and hence not in virtue of intrinsic potencies he possesses. For instance, the Soviet dog that was the first mammal in space could have satisfied the description had the dog landed on the moon as well.
But this doesn’t seem to be an option for the classical theist. For whether there is anything extrinsic to God already presupposes God’s creative act, in which case God couldn’t satisfy the description ‘having a creative act’ in virtue of anything extrinsic to him (lest we admit vicious circularity). In other words, whether or not God is identical to his creative act cannot be in virtue of extrinsic relations (like the technical case with the commie dog, or whatever), since whether there are any extrinsic relations to begin with already presupposes whether or not God has a creative act and hence presupposes whether or not God is identical to that creative act.
The fundamental problem here is a confusion about identity. According to Aquinas and other scholastics, numerical identity is a purely logical relation (ST I.28.1 ad 2). That is, it is a relation arising between a thing and itself in our mental consideration of the thing alongside itself. In reality, there is no relation of numerical identity at all, and so the possibility of the relation obtaining or not obtaining between two (logical) things requires no real potencies in the thing itself.
It seems, though, that I’ve already made provision for this consideration in my comments about the extralinguistic nature of the descriptions in question.
I’ll still say a couple things, though. First, identity is a reflexive relation, but it’s not at all clear why this reflexive relation is purely one of reason. It seems that, in reality itself, for any x, x is identical to itself. This is not a conventional truth but is a metaphysically necessary truth that is grounded in the very nature of identity, which in turn pre-exists any of our linguistic representations, expressions, and so on. God could really (and not merely logically or mentally) be identical to his creative act and could really (and not merely logically or mentally) fail to be identical to his creative act. Similarly, Armstrong could really be identical to the first mammal on the moon and could really fail to be so identical. Or so it seems to me.
Overall, from the considerations adduced above, I conclude that the ninth argument retains its force.
 The epistemic weight I attach to all of these problems is, of course, not identical in each case. Some are stronger than others. This argument is one that I don’t view as extremely forceful, but it does seem to have at least some teeth. And, so it seems to me, Tomaszewski’s response does not remove those teeth. I thus stand by my conclusion that the ninth argument retains its force as a prima facie problem.
Problem 10: Freedom and Potency (response by Nemes)
To understand the argument, see my original blog post.
The actual world exists contingently. It exists but it does not have to exist. Another world could have existed. One might say that the contingency of the actual world is grounded in the freedom of God to create it. But does this therefore mean that God possesses an unactualized potentiality to create another world (or no world at all)? Such an inference would be mistaken. In the first place, it assumes the difference principle, i.e. it assumes that a possible difference in the created world as effect must be grounded in a possible difference in God as cause.
Even if it does presuppose EDP, I wouldn’t view that as particularly problematic (as we’ve seen). Though, it’s unclear to me that it does presuppose it. It seems to arrive at the conclusion by a means that is tangential to the difference principle. More particularly, it seems to be based on an analysis of freedom as requiring the possibility of doing (and hence being) otherwise (i.e. other than what is in act). And since potency and act are a complete division of being, it seems the possibility of doing otherwise (and hence being otherwise) entails the possession of potency. Or so the argument goes. This doesn’t seem to explicitly or implicitly presuppose EDP.
 I guess it might have it as a consequence. But it doesn’t seem to presuppose its truth to get there.
Nemes (with whom it is super fun to discuss on Facebook, I might add, and who is super funny) continues:
But God is not subject to the difference principle. Rather than grounding the contingency of the world in some supposed divine freedom, understood by analogy with human freedom, it would be better to think of the “freedom” of God as cause in terms of the contingency of the world. For God to create the world “freely” means that another created order could have come into existence through God, even though God remains totally unchanged across all possible worlds. In other words, it means that God, such as He is in Himself, is not intrinsically or essentially ordered to the creation of this world or any world at all.
As I hope to have shown, though, the argument doesn’t presuppose EDP, and so objecting on the basis of God’s not being subject to the principle seems to miss the mark.
What’s more, it seems that God’s freedom cannot be had in virtue of the contingency of creation, since it seems that the contingency of creation already presupposes God’s freedom. For if God wasn’t free ontologically prior to his creative act, then he couldn’t have done otherwise — in which case, there couldn’t be any contingency to begin with.
Or so this all seems to me.
I conclude, then, that the tenth problems retains its force.
(For those who want a short break to listen to music, this is truly beautiful)
Problem 11: Begotten Not Made (response by Chutikorn)
Check the original post for a brief exposition of the argument.
The term ‘begotten’ simply means something generated or even “fathered.” When the Nicene Creed utters the words “begotten, not made” it makes a clear distinction between something generated eternally as opposed to something generated in a created way. As I explained in response to Problem 2, the procession of the Son from the Father is one of interior operation (opera ad intra). The mode of generation of the Son is thus different from the mode of generation of things in the created order. That being said, while the Trinitarian processions are understood by faith, in order to understand this procession through the means of human reason, we must first think from what we can know a posteriori.
Fair enough — I don’t see anything particularly objectionable here.
St. Thomas Aquinas often uses the example of intellectual generation in the human soul since this most closely approaches in likeness to God. Just as the soul by its act of thinking begets a word (as interiorly conceived), so too is there in God an interior expression of what he knows – an emanatio intelligibilis. But again, it is because of the absolute simplicity of God that his word is of his very essence, for in us, a word (even considered as an interior concept) is accidental. That is to say, the concept is not identical to our substance. In God, there can be no principle of potency or additional actuality received since God is pure act. To say that God does not have an eternal “Word” would be to say that God is entirely without knowledge, but just as God is eternal, his word must also be eternal.
The ‘problem’, though, is not about eternal begetting, but begetting simpliciter. It seems to be some sort of priority/posteriority relation (like a dependence relation, say). But that seems disallowed under absolute divine simplicity — there cannot be two distinct things, one of which is prior to the other, intrinsic to God. Or so it seems.
 For lack of a better word. I just mean something along the lines of ‘the referent of an expression’.  Indeed, if the Father is really identical to God and the Son is really identical to God, it seems God is both prior to and posterior to himself, for if the Father begets or generates the son, it seems there is some sort of ontological priority and posteriority relation here. But this requires more thought, of course.
It was said that:
(i) J and F are distinct, (ii) at least one of J or F is not identical to God simpliciter, and (iii) J and F are intrinsic to God.
Reply to (i): It is true that the Son and the Father are distinct, but relatively distinct as I explained in response to Problem 2.
As I have already addressed (and found to be inadequate) Chutikorn’s response to Problem 2 earlier, this doesn’t seem to jeopardize the argument.
Reply to (ii): This is not true. The Son and the Father are in fact identical to God simpliciter but not identical with respect to origin since the towardness of each person of the Trinity is derived from the very nature of intellect and will. Yet, there is nothing in the Son that is not in the Father and there is nothing in the Holy Spirit that is not in the Father or the Son. In the procession of the Word from the Father, for instance, all the divine attributes are communicated concomitantly as the divine essence is fully given to this subsistent relation by way of intellect. Likewise, the divine essence is communicated by the bringing forth of the Holy Spirit proceeding by way of the will, as from the mutual love between the Father and the Son.
It’s not clear, however, how to make sense of the claim that the “Son and the Father are in fact identical to God simpliciter but not identical with respect to origin since the towardness of each person of the Trinity is derived from the very nature of intellect and will.” The reason has to do with LL: if x and y are identical, then whatever is true of one is true of another. If that’s the case, then if the Son and the Father are both identical with the Father, then it cannot be the case they have different ‘origins’, for then something would be true of one that is not true of another. Or so it seems to me, that is.
Appealing to relative identity will seem to land in the problems I discussed in Problem 2.
Reply to (iii): The Son and the Father (and the Holy Spirit) are indeed intrinsic to God insofar as they are God, but only distinct inasmuch as they are personal modes of subsistence in God who proceed by way of intellect and will in the intra-Trinitarian life of God.
The relation of ontological posteriority of the Son to the Father cannot lead to an absolute distinction since relations do not constitute natures. What the persons are is the same, but who they are is different since each person either proceeds or is the principle of the processions but share the same essence and mutually indwell through their perichoresis.
While my understanding of the latter paragraph here is admittedly a bit fuzzy, it seems to be making the same (or similar) point about constituting natures that was discussed and addressed in Problem 2. I therefore direct readers to my response in Problem 2.
I conclude, then, that the eleventh argument retains its force.
 Though, it’s entirely possible that my fuzzy understanding is causing me to miss something. Nevertheless, I at least temporarily stand by the (tentative) conclusion.
Here is a final break from philosophy.
Problem 12: Distinction Between Knowledge and Creative Activity (response by Stanislaw)
Here is the original argument:
(1) Knowledge involves a non-causal relation to propositions.
(2) Creating does not involve a non-causal relation.
(3) Necessarily, if x and y are identical, then whatever is true of x is true of y.
(4) So, knowledge is not identical to creating.
(5) If knowledge is not identical to creating, then God’s knowledge is not identical to God’s creative act.
(6) If God’s knowledge is not identical to God’s creative act, then ADS is false.
(7) So, ADS is false.
Stanislaw (who is a legend) writes:
The Thomist rightly rejects 1, as God’s knowledge is the cause or principle from which all things come to exist.
Correct, but the question at issue is whether the Thomist is justified in such a rejection, especially given the intuitive claim that knowledge is essentially a relation to abstacta whereas causal activity is not essentially a relation to abstracta.
However, even accepting premises 1-4, premise (5) is problematic because the term, “God’s creative act,” may signify in three separate ways: (i) the agent from which the act proceeds; (ii) the effect of the act itself; or (iii) the event as a whole without separating (i) and (ii).
The way that I used ‘creative act’ throughout the post was specified in one of the problems of mine: “God having a creative act (else: an act of creation) just means God exercises his causal power to create (i.e. bring something into being).” So, I understand it as the act of bringing something into being. In other words, I mean neither the effect nor the cause, but rather the cause’s causing of the effect.
Premise (5) states that if knowledge is not identical to creating, then God’s knowledge is not identical to God’s creative act. Depending on whether (i), (ii), or (iii) are in mind, the premise is false, for while God’s knowledge is not identical to the effect he produces or the entire event as a whole, his knowledge is indeed identical to himself qua agent or principle from which the act proceeds (just as heat is the agent or principle of heating), namely the divine essence itself.
But God, under classical theism, is such that he is identical to his act of bringing things into existence. There is no distinction between ‘being’ and ‘action’ in God, or between substance and the substance’s exercising its causal power(s). God is, in other words, identical to God’s causing things to be. So, it seems that the trichotomy laid out in (i)-(iii) doesn’t fully capture what I was getting at in my problem.
And — at least this is how things seem to me — nothing said thus far has shown (or undermined the negation of) the claim that [knowledge is not essentially a relation to abstracta]. In other words, premise (1) still seems in tact.
Here, premise (6) If God’s knowledge is not identical to God’s creative act, then ADS is false, only follows from (ii) or (iii), but not (i), and ADS is only committed to (i). Thus, the argument does not go through and ADS is safe.
But classical theism, from my understanding, also maintains that God’s knowledge is identical not only to the agent but also to the agent qua acting. And that is precisely the sense of creative act I have meant from the start. And from this, the argument still arises, since that means God’s knowledge is identical to God’s causing. But ‘causing’ is not essentially a relation to abstracta whereas it seems knowledge is essentially a relation to abstracta — in which case, God’s causing cannot be identical to God’s knowledge (as something is true of one that isn’t true of another).
The problem with the first defense of premise (1) has to do with accepting that a change in the world involves a change in God’s essence, something the Thomist rejects. God, in knowing his essence, knows it not as static but, rather, in all its dynamic fullness of being and as imitable. God’s knowledge of the variability of creatures, both possible and actual, does not entail a variability in himself nor an accrual of some new perfection via received knowledge.
The ‘first defense’ to which he refers ran as follows (this is me from the original post):
The classical theist will likely object to (1) and affirm that God’s knowledge is non-propositional. Instead, God knows things by knowing his own essence.
But this cannot be correct. For God’s essence is utterly unchanging from world to world. But there are contingently true things that can be known, i.e. there are bits of knowledge in some worlds that are not bits of knowledge in other worlds. Because bits of knowledge change from world to world, but God’s essence is utterly unchanging from world to world, God’s knowing God’s essence is categorically insufficient for God’s knowing all bits of knowledge. So, God’s knowing his essence cannot be the whole story to God’s knowledge.
But notice that the inference was not that variability in things known entails variability in creatures. The inference, rather, is that since God’s essence is utterly unchanging from world to world, it follows that any knowledge of God’s essence is unchanging from world to world. But the contingency of the world means that the world is not unchanging from world to world. And from this, it follows that any knowledge of the (contingent) world is not unchanging from world to world. And now the following straightforward and valid deduction follows:
- Any knowledge of God’s essence is unchanging from world to world.
- Any knowledge of the (contingent) world is not unchanging from world to world.
- So, knowledge of God’s essence is not identical to knowledge of the (contingent) world.
Finally, where necessary truths are concerned, these truths known by God are no different than any other truths, and this problem once again surfaces by holding the view that God knows as human knowers do, again something the Thomist rejects. Whether necessary or contingent, God knows by knowing his own essence as imitable. It isn’t clear why one ought to accept that a multiplicity of truths known by God entails a multiplicity of distinct features within him as if his thoughts were received and constitutive.
Note, first, that I already addressed the central criticism in here (God knows by knowing his essence) above.
Secondly, there is a subtle shift at play here, it seems. God might know his essence as imitable, but that neither means nor entails knowledge of the various ways God’s essence is actually imitated. Indeed, this is precisely where the inference above comes into play: while God can know his essence unchangingly from world to world as possibly imitated (since whatever is possible is necessarily possible), it is precisely because the various ways that God’s essence is actually imitated change from world to world that God’s knowledge must change from world to world.
 This is the S5 axiom of modal logic.  Whether or not this is a real (intrinsic) change is beside the point, as this simply isn’t at issue here. The question at issue here is the allegation that my Problem 12 fails to take into account that God knows the contingent world by knowing his essence.
I conclude, then, that the twelfth argument retains its force.
 Once again, the epistemic weight I attach to all of these problems is not identical in each case. Some are stronger than others. This argument is one that I don’t view as extremely forceful, but it does seem to have teeth. And, so it seems to me, Stanislaw’s response does not remove those teeth. I thus stand by my conclusion that the twelfth argument retains its force.
Whelp, that concludes my survey of the responses. Whew! That was exhilarating, don’t you think? Regardless, I have some final remarks that you should attend to.
For those interested, only a couple days after the post was up, I modified it to include a thirteenth prima facie problem. It is labeled as ‘Problem 12′ in the original post. The original post’s problem 12 was bumped to ‘Problem 13’. I’m obviously not blaming any of my classical theist interlocutors for not addressing it, since it was added after a couple days had elapsed since the original post.
From the foregoing analysis, here are all of my (thus-far-held, that is) verdicts on the plethora of prima facie problems:
- One: retains its force
- Two: retains its force
- Three: retains its force
- Four: retains its force
- Five: retains its force
- Six: retains its force
- Seven: retains its force
- Eight: very unclear whether the argument retains its force — undecided/more thought needed
- Nine: retains its force
- Ten: retains its force
- Eleven: retains its force
- Twelve: retains its force
Honestly y’all, we should all write a book together on classical theism. It would be lit.
Now, I’ve been keen that the arguments I presented were not knock-down, irrefutable ‘demonstrations’. Those are — for the most part at least — myths. However, here is an irrefutable, decisive proof that classical theism is false.
Before leaving, I encourage you to check out my book on critical thinking in philosophy! It’s currently the #1 New Release in philosophy methodology on Amazon. And get this: the kindle version is only $1.99 — that’s less than a cup of coffee!
Email: NaturalisticallyInclined@gmail.com (no, I’m not a naturalist)