1.1 The sovereignty–aseity intuition as the foundation of simplicity

The characteristic approach of PBT is notorious in Alvin Plantinga’s famous presentation of DDS. He takes divine greatness, a widely accepted divine attribute, as a starting point for his argument:Footnote ⁶

1. (3) Divine greatness or perfection. God is a being of incomparable greatness, that being than whom it is not possible that there be a greater.

Plantinga then describes divine greatness as multifaceted and points out two attributes among the various aspects it comprises:

1. (4) Aseity. God is uncreated, self-sufficient, and not dependent on anything else.

2. (5) Sovereignty. God has control over all things, and everything else depends on His creative and sustaining activity.

Thus, Plantinga links DDS to what he calls sovereignty–aseity intuition,Footnote ⁷ i.e., he takes our natural intuition to be that (4) and (5) are included in (3):

1. (6) Sovereignty–aseity intuition. A being greater than which there can be no other must not depend on anything else, and anything else that exists must depend on Him.

However, ‘the whole realm of abstract objects—the whole Platonic pantheon of universals, properties, kinds, propositions, numbers, sets, states of affairs and possible worlds’ supposes a severe reduction to (6), since ‘it is natural to think of these things as everlasting (…). Abstract objects are also naturally thought of as necessary features of reality, as objects whose non-existence is impossible.’Footnote ⁸ Thus, if such abstract entities were to exist, they would be completely independent from God, which undermines (5). Besides, if we believe that God has properties (such as goodness, omniscience, or omnipotence), these would be properties that, at least in God, exist eternally, without Him having done anything to have them or being able to do anything in order to get rid of them.

This undermines not only (5), but also (4). So, we have that:

1. (7) If God has a nature, there is no sovereignty–aseity. If God has a ‘nature’ (an essential property such that God cannot exist without that property), and that property is not identical to God himself, then God is not absolutely independent from everything else, nor is it true that everything else depends on him.

There are three ways to avoid such impasse: (i) positing DDS and holding that all of God’s properties are identical to God himselfFootnote ¹⁰; (ii) embracing nominalism and denying that there are such things as ‘natures’ and ‘properties’Footnote ¹¹; or (iii) denying that God has a nature and asserting that absolutely anything is possible for him.Footnote ¹² After extensive examination, Plantinga concludes that none of these options is fully convincing. In short, there is an insurmountable collision between our intuitions, which forces us to abandon those that seem most dispensable in favor of the more crucial ones. Confronted with this dilemma, Plantinga chooses to preserve our intuition that God has a nature, and therefore abandons (6).Footnote ¹³

The way in which Plantinga presented the motivations for DDS reappeared practically unchanged in other critics of the doctrine. For example, Tom Morris explicitly starts his argument from the assumption that ‘The most powerful and profound philosophical conception of God is given to us by perfect being theology, that method for conceiving of deity which has as its primary and controlling consideration the idea of a greatest possible or maximally perfect being.’Footnote ¹⁴ He points out as a peacefully accepted point in PBT that aseity and sovereignty are ingredients of perfection or great-making properties and focuses on arguing that the denial of DDS does not entail a type of dependency that undermines these properties in a relevant sense.

Other critical discussions of the foundations of the doctrine take up the basic argument that associates DDS with sovereignty and aseity, but fail to trace these other two attributes back to the more basic claim that God is maximally perfect.Footnote ¹⁵ These authors seem to take aseity and sovereignty as a starting point for their discussion, and I do not know if they do so with the open intention of avoiding a commitment with PBT. However, the lack of any alternative development of the reasons to posit such basic attributes in the first place leads to a somewhat vacuous a priorism that does not adequately represent the approach followed by Aquinas. This gap is actually crucial in Baddorf’s argument, since along his paper he drops the principle according to which what is composite must be ontologically dependent or caused.Footnote ¹⁶ He does not consider the fact that this principle he has so easily dismissed is but one of the variants of the principle of causality on which cosmological arguments depend (and, as we will see in next section, such arguments are essential for the Thomistic project). Thus, although nothing is said about the God as a maximally perfect being, PBT continues to loom as an implicit background for the argument.

1.2 The PBT approach in defenses of divine simplicity

This presentation of the motivations for DDS within the framework of PBT was peacefully accepted and appropriated by many defenders of the traditional doctrine.

For example, William Mann said:

The same happens in the presentation of Vallicella, according to whom a central motivation for DDS

The second motivation presented in this article has a somewhat different structure than Plantinga’s argument, but it also follows the matrix of PBT:

Another example is Brian Leftow, who has generally defended the usefulness of the framework provided by PBT.Footnote ¹⁹ ‘The medievals found divine simplicity important because they explained the concept of God by explicating the concept of a perfect being, and they considered simplicity a keynote of perfection.’Footnote ²⁰ However, Leftow also associates the intuition of sovereignty–aseity with what he calls ‘ultimacy assumption’: it is part of the ordinary theist’s concept of God that no true explanatory return can go beyond God.Footnote ²¹ In fact, in other publications Leftow seems to present perfection as an aspect of ‘ultimacy’, which would be the most distinctive feature of God.Footnote ²² Leftow is therefore somewhat ambiguous regarding whether the real starting point is God as that greater than which nothing can be thought, or God as the ultimate explanatory foundation of all reality. Both notions seem to be presented as equivalent, though they are actually completely different starting points, which lead respectively to PBT or ANT.

Stump and Kretzmann avoid going into detail about the motivations for DDS, but they succinctly state that it ‘derives from the metaphysical considerations that have led philosophers and theologians to maintain that God is a being whose existence is self-explanatory, an absolutely perfect being or pure actuality’.Footnote ²³ As can be seen, this presentation has an ambiguity similar to that of Leftow, insofar as both starting points, that of PBT and that of ANT, are presented as equivalent. In any case, in a footnote of this paper it is said that ‘no conception of God that omits either perfect-being theology or biblical accounts of God as a person can count as fully Christian, on a historical understanding of orthodox Christianity’ (p. 378).

James Dolezal’s approach is harder to classify.Footnote ²⁴ It seems that in his book the fundamental attribute of God is his ‘absoluteness’, but it is not clear what exactly is meant by this word and what are the reasons that justify God having that condition. On the one hand, the shift of focus from perfection to absoluteness intensifies the commitment to aseity, bringing us closer to ANT. But if we consider God’s absoluteness as given a priori, or as something that must be deduced from the eminent divine perfection, Dolezal will still remain within the framework of PBT.

Barry Miller, however, openly blamed PBT as responsible for the current misunderstanding of DDS. This author argues contemporary Anselmians make the mistake of conceiving God’s perfection in terms of God ‘having the maximally consistent set of great-making properties or perfections’.Footnote ²⁵ By doing so, they understand the difference between the perfection of God and that of creatures as one of degree only, and ‘therefore, God turns out to be simply the greatest thing around, some kind of super-being that would be quite capable of evoking admiration and wonder, but who could scarcely be described as being absolutely transcendent, or as being worthy of worship’.Footnote ²⁶ Instead of placing God at the summit of a scale of perfections, Milles conceives divinity ‘as that to which the items on the scale merely point or that towards which they merely tend to converge without ever actually doing so’.Footnote ²⁷ For Miller, God is not the maximum or the limit simpliciter within a series, but it is a limit case that falls outside the series. This critique is clever, but it fails to clearly articulate an alternative model to PBT for the development of divine attributes and therefore fails to provide an adequate explanation of the motivations that lead Aquinas to posit DDS. Although he criticizes PBT, Miller generally approves it as a ‘time-honoured procedure for discussing God’s nature (…) to settle upon some controlling notion to use as the yardstick against which to judge what can or cannot be said of God’.Footnote ²⁸ In that sense, he suggests that Aquinas differs from Anselmians in that he takes ‘God as Subsistent Existence’Footnote ²⁹ as his control notion. But this assumption has the problem of being a description of God to which Aquinas only arrives through DDS.

1.3 Should Aquinas’ claims about DDS be interpreted within the framework of PBT?

Three paradoxes arise when interpreting Thomistic arguments for DDS along the basic lines of (3)–(7).

The first and most notorious is that in both SCG and ST (and also in the shorter Compendium Theologiae), simplicity is developed just after the proofs of God’s existence and before discussing divine perfection. Thomas could hardly have introduced DDS as a way of accommodating (7) and (6), since (3) – which according to Plantinga grounds (6) – is not introduced until after DDS has been established. That is, Thomas does not posit simplicity to avoid collisions between attributes derived from the eminent perfection of God, since in all the works in which he makes a systematic development of the divine attributes, divine perfection is always defended once simplicity has already been posited.

A second paradox is the frequent merge of sovereignty and aseity that usually takes place in these discussions. Through Plantinga’s discussion, the concern for divine sovereignty acquires progressive preeminence, until finally claiming that the really important point here is God’s absolute control over Creation: God’s nondependence on anything else and the dependence of other things on God must be understood in the sense that the existence of creatures depends ontologically on an action that God performs freely, while God’s existence does not depend analogously on anything that creatures do.Footnote ³⁰ As Immink observed very early, this is one of the most objectionable points of Plantinga’s discussion.Footnote ³¹ It does not reflect at all the thought of Aquinas. By redirecting the problem of ontological dependence to that of control, Plantinga suggests that DDS ultimately pretended to safeguard God’s creative omnipotence. But again, this has little or no textual support. In fact, in ST the problem of God’s power is addressed much later, almost at the end of the treatise on Deo Uno (q. 25), and the thesis that creatures depend in their being on an efficacious action of God is only developed in detail from q. 44 onwards. Again, then, DDS is presented as grounded in claims that, within the architecture of the Thomistic treatises, only appear at very advanced stages of the argument, long after DDS has been established.

Finally, it should be noted that within the theological tradition, DDS appears closely linked to the theme of divine transcendence. As Immink observes, among the theological motivations of the traditional doctrine, the otherness of God has at least as much preeminence as aseity.Footnote ³² This motif is clearly dominant in pseudo-Dionysius, who is perhaps the most important source for Aquinas in this matter. DDS must be placed in the context of the negative moment of Thomistic theology. The very introduction of q. 3 warns us that here we will rather develop the way in which God is not, the way in which he radically differs from creatures. It is not clear how this aspect of the doctrine can be rescued when it is conceived and elaborated from the matrix of PBT.

2.1 ANT in the context of the meta-theological debate

The dominance of PBT as a philosophical method for elaborating the divine attributes was quite noticeable when most of the articles discussed in the previous section were written. But more recently, this approach has become the target of frequent criticisms.Footnote ³³

Kvanvig has recently proposed a systematic approach to the meta-theological problem.Footnote ³⁴ In his recent book, this author contrasts PBT with two other fundamental theological projects: Creator Theology and Worship Worthiness Theology. Of these three approaches, the one closest to the Thomistic perspective seems to be Creator Theology, which places at the starting point the relationship of ontological dependence between the creature and the creator, with a consequent strong commitment to aseity. However, the way in which Kvanvig himself conceives it makes it still inadequate for the interpretation of Aquinas. Creator Theology as presented by Kvanvig

Daniel De Haan gives a better outline of the project that governs Thomistic reflection on divine attributes. For this author, Aquinas joined Albert the Great, Bonaventure, and many other scholastics in drawing on pseudo-Dionysius triplex via to have ‘a principled heuristic for understanding God by way of causality, negation, and super-eminence that keeps philosophical theology from being either wholly negative and agnostic or too optimistic and falling into ontotheological idolatry’.Footnote ³⁶ The strategy of the triplex via consists in getting to know God from creatures through the paths of causality, removal, and eminence. These three routes constitute intertwined steps that interact in the articulation of our natural knowledge of God, allowing us to develop a natural theology that fulfills a semantic and hermeneutical function toward revealed theology.Footnote ³⁷

I have been tempted to join De Haan and instead of talking about ANT refer to Aquinas’ project as TVT (Triplex Via Theology). However, De Haan himself admits that TVT entails many metaphysical assumptions that today would be considered highly contentious.Footnote ³⁸ I agree with this author that relying on the background of an entire metaphysics is not a disadvantage of the Thomistic approach but precisely one of its strengths. Nevertheless, it seems to me that for present purposes it is not necessary to commit ourselves to the entire Thomistic metaphysical background. For this reason, when presenting ANT I preferred to describe the project in broader terms, so that not only Aquinas but many other philosophers can be interpreted within this trend. Anyone who begins with a natural theology where God is known indirectly through creatures, and ultimately grounds the divine attributes on a reflection about how a complete explanation of creatures requires the postulation of a transcendent cause for them, will fall into ANT.

2.2 The semantic function of the cosmological argument in ANT

While PBT begins with the definition of God as that greater than which nothing can be conceived, ANT grounds our conception of God in prior considerations about the natural world. This means that for ANT the basic description of the divine essence is not a given datum that can be taken for granted and then used as a touchstone to evaluate the merits of certain theological doctrines. We should rather start from our total inability to define God, whose essence is not an object that is offered to our experience nor be directly apprehended by the intellect (among the numerous texts where Aquinas develops this idea, see ST I q. 12 a. 12). This gives rise to what can be called the ‘semantic function’ of cosmological arguments.Footnote ³⁹

1. (8) Semantic function of cosmological arguments. Cosmological arguments allow us to fix the referent for the word ‘God’, which designates that whose existence is concluded by such arguments.

Assigning a semantic function to cosmological arguments means approaching the proofs of the existence of God from a nonprescriptive point of view. That is, instead of positing in advance a definition of God to which the result of the argument must conform in order for it to be considered successful, we must let the argument itself provide us with the relevant definition from which we will elaborate the notion of God. This is so because, as Aquinas himself maintains following Aristotle, ‘when the cause is demonstrated by its effect, it is necessary for the effect to take the place of the cause’s definition, in order to prove that cause’s existence’ (ST I q. 2 a. 2 ad 2).

In this sense, there is a close connection between the arguments for the existence of God and what Aquinas usually presents as the first step of the triplex via: the via causalitatis. Footnote ⁴⁰ While PBT postulates the infinite perfection of God as its starting point, ANT starts from the imperfection and finitude of creatures, upon which metaphysical reflection reveals its ontological dependence on a higher principle. The famous five ways to prove that God exists should not be read as a piece of polemic apologetics aimed at settling the debate between theism and atheism, but rather as a starting point to open a conversation about God and thus make theism intelligible.Footnote ⁴¹

There are several versions of the cosmological argument. This is relevant, since not all such arguments in question are equally close to DDS. Since our effort here is to help a more adequate exegesis of Aquinas , it is natural that we take as a point of reference the famous five ways of ST I q. 2 a. 3. Without pretending to make a complete reconstruction of the arguments, I am interested for these purposes in calling attention to at least three elements present in each one: (a) an empirical fact that serves as a starting point; (b) some variant of the principle of causality, which can be applied without exception to all those entities for which any of the conditions specified in (a) are fulfilled; and (c) the postulation of God as ultimate cause of the entities that meet the conditions (a), and which must be described in such terms as to prevent the principles (b) from being applied to him (since that would prevent God from fulfilling the required explanatory function). Thus, among other things, Aquinas’ arguments state:

1. (9) Argument from the unmoved mover. (a) The things of this world move; (b) everything that moves is moved by something else; and (c) God is that being who moves all things without being himself susceptible of being moved by anything.

2. (10) Argument from the uncaused cause. (a) There are chains of efficient causes in which we observe that none of the members is the cause of its own causing; (b) if there were no first cause, there could be no subordinate causes; and (c) God is that which causes all other things, and nothing can be His cause.

3. (11) Argument from the necessary being. (a) Natural things are contingent, as is seen in their generation and corruption; (b) every contingent entity has a cause; and (c) God is that being necessary in itself on which depends everything contingent or everything that has the cause of its necessity in another.

4. (12) Argument from the most perfect being. (a) In natural things, we observe certain transcendental properties which all of them have according to different degrees; (b) what is maximum in one genus is the cause of the others that participate in that genus; and (c) God is the cause of all the limited perfections of creatures, which He himself contains in an eminent manner.

5. (13) Argument from the universal orderer. (a) We observe that in natural things some things which lack knowledge act for the sake of an end, which is proven by the fact that they always behave in the same way in order to achieve what is most convenient for them; (b) things that lack knowledge are not ordered to an end unless directed by some intelligent agent; and (c) God is the intelligence which governs all natural things, ordering them toward their end.

This is not the context for asking whether these arguments are successful or not or whether we can come up with better arguments than Thomas’. The relevant point here is that, within Thomistic theology, the basic descriptions of God that serve as a starting point for the development of the divine attributes are the various variants of (c), and that those initial descriptions can only be reached once we have already accepted the various variants of (a) and (b). So, the challenge for someone who opposes DDS from the approach of ANT is either to deny them and develop another successful variant of the cosmological argument that can serve as a good starting point for developing a robust notion of God, or to accept them and avoid DDS from following from what has already been granted. So far, I have not found in the literature any critics of DDS who manage to successfully meet this challenge.

2.3 From the cosmological argument toward divine simplicity

So, we can concede to Anselmians that the motivations for DDS must be associated with aseity and sovereignty – (4) and (5). But these attributes are not deduced from (3) – divine greatness or perfection – through an argument as weak as (6) – intuition of sovereignty–aseity. Rather, we have that (4) and (5) follow from the description of God that is made in cosmological arguments, that is, in premise (c) of arguments (9)–(13).

Now, the deduction that goes from cosmological arguments to DDS, passing through aseity and sovereignty, is also complex. Both in SCG and in ST Thomas conducts an exhausting (and sometimes even repetitive) exclusion, one by one, of the different types of composition that there could be in God. This introduces us to the second moment of the triplex via: the via negationis or via remotionis. This stage of the theological discourse delves deeper into the absolute transcendence of the God who is concluded in the cosmological arguments. We must specify the ways in which this God is radically different from any creature, showing that if certain distinctions that are inherent to finite beings were applied to him, he would not be fit to play the role that we assigned him in our metaphysics. It is thus very important to keep in mind that simplicity is not a positive attribute that Aquinas directly postulates but rather the result of a set of successive negations.Footnote ⁴² This results in a notion of God that seems utterly obscure and prevents us from using univocal language to refer to divinity. This mysterious darkness is indeed part of what Aquinas seeks to safeguard with DDS. Simplicity introduces a dose of moderate agnosticism into our image of God, without which our discourse on Him would be deceptively clear. We are at the end of the day talking about a Being whose essence escapes us, a God of whom we have no direct experience but who only reveals himself to us indirectly through creatures.

Among the different forms of composition that Thomas Aquinas excludes from God, the culminating point seems to be reached when the identity between being and essence is established. If such identity is granted, all the others must also be accepted. At this point, the incisive criticisms that Christopher Hughes made of Aquinas’ DDS will be helpful. At the beginning of his book, Hughes argues that at the core of Thomas’ defense of divine simplicity is his identification of essence and existence in God, since this characterization of God as ipse esse subsistens would by itself exclude the possibility of any other kind of composition in God.Footnote ⁴³

Now, Hughes states he could find in the texts of Aquinas five types of arguments in favor of this peculiar identification between being and essence:Footnote ⁴⁴

1. (14) Argument from the nature of the composite beings. (i) The following properties are common to every composite being: (a) having a potential element; (b) being subsequent-dependent on its components; (c) being corruptible; (d) needing an efficient cause to make the components, which are in themselves diverse, enter into composition; and (e) having parts less perfect than itself. But (ii) none of these features can be reasonably attributed to God. Therefore, (iii) God cannot be a composite being.

2. (15) Argument of act and potency. If God is not identical to his being, he is related to it as potency relates to its respective act.

3. (16) Argument of participation. If God is not identical to his being, he can only exist by participating on being.

4. (17) Argument from causation. If God is not identical to his being, God must be caused.

5. (18) Argument from the adequate cause. (a) The existence of an effect common to several causes requires the existence of a higher order cause whose proper effect is that one; (b) but God is the proper cause of being; furthermore, (e) every proper cause is similar to its effect; therefore, (f) God is subsistent being himself.

Hughes’ treatment of this topic is clearly much more careful than Plantinga’s, and as far as the strict textual basis is concerned there is no doubt that this is a good synthesis of the arguments Thomas uses to establish DDS.

So, in order to see whether we can dispense with DDS within Aquinas’ ANT, the reader need only consider arguments (14)–(18) and check whether it is possible to make a convincing refutation of all of them without severely spoiling the argumentative complex (9)–(3). Hughes at least fails to do so. Hughes’ refutation of (14)(a) – presence of potential elements in composites – (14)(b) – subsequent-dependent character of the composite upon its parts – and (15) – argument from act and potency – supposes the denial of premises (b) and (c) of (13) – argument from the unmoved mover – since it implies denying that there are any difficulties with God having potentialities.Footnote ⁴⁵ Refutation of (14)(c) – corruptible nature of composites – supposes a denial of premise (a) of (11) – argument of the necessary being. Likewise, his refutation of (15)(d) – necessity of a cause that makes the elements in themselves diverse of the composite enter into composition – is resolved in a denial of premise (b) of (11) – argument of the necessary being.Footnote ⁴⁶ The refutation of (16) – argument of participation – supposes abandoning premises (b) and (c) of (12) – argument of the most perfect being.Footnote ⁴⁷ I think it is idle to continue with this. I leave the exercise to the reader.

Perhaps the only of the Thomistic ways that seem to survive the critique of divine simplicity is (13) – the argument from the universal orderer. But I suspect that this argument is by itself insufficient to ground a strong enough theology. Separated from the other arguments, it allows at most to postulate a demiurge who infuses order and purpose into the universe, but not a God who makes everything out of nothing. In any case, the nonprescriptive approach that arguments for the existence of God take in ANT means that this proof cannot be declared insufficient just because its conclusion is a God who fails to fulfill all the standard attributes of classical theism. How far one can go in developing an image of God based on teleological arguments is an interesting challenge still open to those who feel inclined to a project along the lines of ANT but have serious reservations about DDS.