Hostname: page-component-74656f9884-qr5l4 Total loading time: 0 Render date: 2026-11-03T22:02:34.104Z Has data issue: false hasContentIssue false
  • English
  • Français

Divine Simplicity and Analogy: Clarifying Some Conundrums

Published online by Cambridge University Press:  03 August 2026

Maikki Aakko Open the ORCID record for Maikki Aakko [Opens in a new window]
Maikki Aakko*
Affiliation:
Theology and Religion, University of Helsinki, Finland Campion Hall, University of Oxford, Oxford, UK
*
Email: maikki.aakko@campion.ox.ac.uk; maikki.aakko@helsinki.fi
Rights & Permissions [Opens in a new window]

Abstract

This paper defends a broadly Thomistic understanding of simplicity against a range of contemporary criticisms. Critics of divine simplicity argue that simplicity leads us to identify God with an abstract property and that it renders theological language semantically vacuous because divine simplicity cannot ground a plurality of true intrinsic predicates. I defuse these criticisms by employing some central Fregean distinctions – specifically, the distinctions between object and concept, predication and identification, and sense and reference. While these Fregean distinctions help to dissolve initial logical difficulties, they do not by themselves explain how theological predicates can have determinate content or truth-conditions. I therefore situate the relevant Fregean diagnostics within the Thomistic metaphysical framework of analogy. This framework supplies principled constraints on theological predication without reintroducing composition into God. I conclude by arguing that the Thomistic doctrine of analogy does not eliminate semantic openness but renders theological language sufficiently determinate for meaning and practice, grounding meaningful God-talk in both metaphysics and the formative practices of religious life.

Keywords

analogysemanticssimplicitytheological languageThomism

Information

Type
Article
Information
New Blackfriars , Volume 107 , Issue 4 , July 2026 , pp. 407 - 425
Check for updates
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2026. Published by Cambridge University Press on behalf of Provincial Council of the English Province of the Order of Preachers.

1. Introduction

The aim of this paper is to defend a broadly Thomistic account of divine simplicity against a cluster of contemporary metaphysical and semantic objections. Critics of divine simplicity argue that simplicity renders God identical with an abstract property and/or renders our language about God semantically vacuous since a simple God seems unable to ground a plurality of true intrinsic predicates. In this paper, I employ the resources of modern analytic philosophy – specifically, selected Fregean distinctions – to demonstrate how these critiques can be defused.

The two central Fregean distinctions that are employed in this paper are those between object and concept and between sense and reference. Several authors have employed the Fregean sense/reference distinction in their defences of divine simplicity, arguing that a plurality of attributes can be predicated of a simple referent.Footnote 1 Thus, names with multiple senses can be truly said of a simple divine being and the charge of semantic vacuity is avoided.

The object/concept distinction, on the other hand, is far less discussed in contemporary literature on divine simplicity.Footnote 2 Unlike the sense/reference distinction that largely corresponds to Aquinas’s ratio/res significata distinction, the object/concept distinction is also not found in Aquinas as such. This paper argues that bringing this distinction to bear on Thomistic defences of simplicity can be illuminating and make a genuine contribution to these defences.

However, the clarifying procedure involving the employment of these distinctions – Fregean diagnostics, as I’ve termed it – does not yet explain how theological predicates can have truth-conditions and sufficiently determinate content.Footnote 3 I argue that in order to address these underlying concerns of the critics of simplicity, Fregean diagnostics must be located in the broader semantic-metaphysical framework provided by the doctrine of analogy. It is this framework that introduces a crucial further distinction that Fregean diagnostics do not give us – namely, the distinction between res significata and modus significandi – and thereby provides the resources needed to ground the meaning of theological predicates. Once this overall framework is in place, the coherence of central Thomistic commitments becomes evident.

Thus, this paper contributes to existing debates surrounding simplicity by arguing that the more often-used sense/reference distinction remains an incomplete defensive strategy if it is isolated from the metaphysical framework of analogical semantics. The same goes for the more rarely used object/concept distinction: it, too, remains insufficient on its own to ground the meaning of theological predicates. This paper, then, will not simply introduce a previously neglected distinction to clarify simplicity-related conundrums: it also clarifies the role of broader metaphysical commitments in debates over simplicity.Footnote 4

More specifically, I will proceed in the following order. Firstly, in Section 2, I offer a brief exposition of a broadly Thomistic understanding of simplicity. Then, in Section 3, I introduce Fregean diagnostics to solve some of the apparent conundrums that this understanding of simplicity produces according to its critics. In the following section, Section 4, I demonstrate how a Thomistic doctrine of analogy supplements Fregean diagnostics by providing the needed semantic-metaphysical grounding to render theological predicates sufficiently determinate. In the last section, Section 5, I offer some remarks on the significance of practice for how we understand the emergence of analogical meanings for finite beings like us.

2. Divine simplicity: a brief exposition

Classically, theologians have tended to affirm that the divine being is simple, i.e., utterly one: there are no parts – spatial, temporal, of any kind – in God.Footnote 5 God is necessarily and utterly and only Godself, identical with God’s own nature. According to classical Thomistic understanding, the doctrine of simplicity follows necessarily from God’s being the first cause: we designate as God that which answers the question ‘why is there something rather than nothing?’, and whatever answers this question must be ontologically prior to everything else. Since ‘every composite is posterior to its component parts, and is dependent on them’, God cannot have components – otherwise these components would be more fundamental than God.Footnote 6 God must also be pure act without any potentiality, because otherwise potentiality would be prior to God.Footnote 7 There is also no distinction between essence and existence in God, because as the first cause God must be uncaused: God has the reason of God’s existence in God’s essence because God’s essence simply is existence – or, as David Burrell puts it, ‘for God to be, is for God to-be’.Footnote 8 God is One, because God is First.Footnote 9

More specifically, the doctrine of divine simplicity entails that the following kinds of distinctions and compositions are denied of God due to God’s ontological priority: matter–form composition, the essence–individual distinction, the essence–existence distinction, and the substance–property distinction.Footnote 10 Note that these are apophatic denials: they do not posit anything positive of God. Rather, they only entail knowledge of what God is not. Footnote 11

3. The semantic logic of divine simplicity: a Fregean diagnosis of two conundrums

This paper focuses on the last of the aforementioned denials, i.e. the substance–property distinction. According to the doctrine of divine simplicity, God does not have properties: everything that God is, God is all at once, in one pure act. Why? Because a distinction between what God is and God’s properties would entail composition in God. Composition, in turn, implies dependence and potentiality: the whole depends on its parts and the parts could in principle be organised differently.Footnote 12 If this were the case, we would have once again identified a metaphysical reality that was ontologically prior to God.

This is where we get to critiques of divine simplicity that are the subject of this paper.Footnote 13 For two things follow from this. Firstly, it follows that God is identical with God’s essence: God does not have ‘goodness’ as a property, but God simply is essentially good. Secondly, it follows that divine attributes all have the same referent: we are not picking out some distinct ‘part’ in God when we speak of God’s goodness, nor yet another when we discuss God’s wisdom.Footnote 14

The critics of simplicity like to point out that this leads us to two conundrums, one metaphysical and one semantic. Here’s the metaphysical (à la Plantinga):

Conundrum A

  1. 1. God is good.

  2. 2. As simple God is identical with God’s goodness.

  3. 3. Therefore, goodness is God.

That is, for critics of simplicity, Conundrum A leads us to identify God with some abstract property such as goodness. This, in turn, would indicate that God ‘isn’t a person but a mere abstract object’.Footnote 15 We can appreciate the force of this objection even if we are not committed to divine personhood. For even a Thomist would want to speak of God in agential terms: God acts, God knows, God wills.Footnote 16 God does all these, because God is Life itself. God is not a thing, a nature, or an abstract property.Footnote 17 Yet, Conundrum A indicates that this is exactly what divine simplicity will force us to conclude. Furthermore, it would seem to follow from all this that God can only be one kind of thing or one kind of property: either God is identical with goodness or God is identical with wisdom, but certainly God cannot be at the same time fully and exhaustively identical with both goodness and wisdom.

So much for the metaphysical conundrum, let’s look at the semantic one next.Footnote 18 It goes like this:

Conundrum B

  1. 4. God is good.

  2. 5. God is wise.

  3. 6. As simple God lacks any and all metaphysical distinctions.

  4. 7. Therefore, goodness and wisdom are the same thing in God.

The implication of this conundrum is that if God is simple, we have very little grasp of what different divine attributes signify in their multiplicity. What is goodness that is identical with wisdom? The doctrine of divine simplicity, therefore, threatens the coherence and meaningfulness of our language about God because it makes it impossible to meaningfully predicate multiple attributes of God. It, in other words, seems to render theological predicates semantically vacuous.

Now, Aquinas is aware of counterarguments against the doctrine of divine simplicity, and he already possesses a sophisticated semantic-metaphysical apparatus for divine naming, some of which reflects Fregean distinctions (e.g., ratio/res significata distinction that largely corresponds to the Fregean sense/reference distinction). These allow Aquinas to explain how many names can be truly predicated of a simple God. My argument is not that this Thomistic apparatus should be replaced with Frege. Rather, my argument is that with the help of Frege’s classic distinction between objects and concepts, we can diagnose these conundrums as semantic misreadings. This move is primarily clarificatory and reconstructive: the aim is to articulate a Thomistic defence of simplicity in the idiom of contemporary analytic philosophy, thus also demonstrating the continued relevance of Thomistic understanding of simplicity.

In Section 4, I will then argue that Fregean distinctions are not sufficient to ground the meaning of theological predicates. Hence, I propose that divine simplicity motivates non-standard but sufficiently determinate semantics, namely analogical semantics. When understood in its metaphysical context, analogical semantics also supplies principled constraints on admissible divine predications that are grounded in participation and eminence (i.e., it explains why ‘God is good’ is eligible for intrinsic predication while ‘God is evil’ or ‘God is a chair’ is not).

Firstly, however, my approach involves clarifying the relationship between objects and concepts and distinguishing different forms of predication.Footnote 19 This will help us see how our conundrums are based on a semantic misreading. According to Frege’s famous distinction, objects, to put it simply, pick out things in the world, they are logically complete by themselves.Footnote 20 Examples of objects include: the White House, Venus, cat, and pencil. The term ‘God’ can also function as an ‘object’ in Frege’s technical sense, i.e., as the kind of term of which predicates are affirmed.Footnote 21 Objects can relate to concepts, but they do not need to. Concepts, to put it simply again, fall under objects and to mean something, they need an object. Concepts require something to which they apply; they are parasitic on objects, if you like. (Hence, I say ‘wide’ and you ask ‘wide what?’)Footnote 22 Concepts include things like: wide, thick, red, and Trinitarian.Footnote 23 However, some words can be used both as objects and as concepts, for example, words such as planet, triangle, or colour.Footnote 24

Following this, note also Frege’s crucial insight with regard to the word ‘is’, namely, that there are two kinds of ‘is’: the ‘is’ of predication and the ‘is’ of identification.Footnote 25 Consider:

  1. 8. The barn is red.

  2. 9. The evening star is planet Venus.

Sentence (8) is a matter of predication: a concept (red) is predicated of an object (barn). It follows that these cannot exchange places: (8) does not entail that ‘red is barn’. Sentence (9), on the other hand, is a matter of identification: it’s an object–object predication. We can therefore say that ‘Venus is the evening star’. Consider another example: Bruce Wayne [object] is Batman [object]. This can be turned around: Batman is Bruce Wayne. Both sentences make sense and are true. Try the following, though: Bruce Wayne [object] is tall [concept]. This cannot be turned around: tall is not Bruce Wayne.

Now, for Aquinas, terms such as ‘goodness’ are abstract perfection terms that can be properly used of God.Footnote 26 Abstract names denote divine perfections but, given simplicity, they do not denote properties in God or any other distinct metaphysical entities. Rather, they signify the simple divine being itself, under different formal aspects. Nevertheless, these abstract names can seem to be functioning logically much like Fregean objects and in so doing produce something like Conundrum A. Designating abstract perfection terms as Fregean concepts, however, clarifies the semantic-logical situation: abstract perfection terms such as goodness are concepts. Indeed, I’m arguing that this is what divine simplicity requires us to say: if God is simple, then divine being is not composed out of a plurality of distinct perfections. Abstract perfection terms, therefore, cannot refer to distinct objects in the Fregean sense but rather they denote the one, simple divine being only in the concept sense. Hence, in ‘God is good’, ‘good’ is not ‘a thing’ but rather ‘good’ is a concept predicated of ‘a thing’ we call God. In light of this, we can diagnose Conundrum A as the following kind of confusion: saying that [God is good = goodness is God] is like saying that [the barn is red = red is a barn]. We are confusing object–object predication with object–concept predication.Footnote 27

This being said, we can also use nominalised perfection terms as singular terms to refer to God. We can, for example, speak of ‘the Goodness from which all creaturely perfections flow’ or we can say that ‘God is goodness itself’. Speaking in these ways does not necessarily mean that God has goodness as an inhering property; rather, it is simply a way of expressing in a derivative and nominalised idiom that God is good essentially. Here, goodness is a nominalised perfection term whose ordinary grammar might tempt us into reading it as a name for an abstract object or a property. However, in divine discourse, it functions to signal essential predication without positing a property-item in God. Hence, words like goodness, divinity, simplicity, or Trinity can be used both in object and concept sense, but this does not indicate a shift into the kinds of metaphysics of God, where things such as goodness are properties in God.

But how about Conundrum B? According to the critics of simplicity, ‘God is goodness’ and ‘God is wisdom’ are akin to (9): they are a case of object–object predication.Footnote 28 Hence, the logical sequence in (4)–(7) is thought to be the same as in the following:

  1. 10. The evening star is the planet Venus.

  2. 11. The morning star is the planet Venus.

  3. 12. Therefore, the evening star is the morning star.

However, this is in fact not the case. In actuality, (4)–(7) represent a logical fallacy of the sort that was just introduced:

  1. 13. The barn is red.

  2. 14. The barn is wide.

  3. 15. Therefore, red is wide.

This logical fallacy stems from not distinguishing between the ‘is’ of predication and ‘is’ of identification. Logically, the words ‘good’ and ‘wise’ are not objects in (4)–(7) but concepts. This, as I argued, is one of the central semantic results of the doctrine of simplicity: divine attributes cannot be predicated of the divine essence as separate things or objects because they are not ‘parts’ or ‘things’ in God. Rather, they are concepts that signify the undivided, utterly simple divine essence. Hence:

  1. 4. God [object] is good [concept].

  2. 5. God [object] is wise [concept].

  3. 16. Therefore, God is both good and wise.

Now the critic might jump in, though, pointing out that we have predicated a plurality of attributes of a simple God (God is both good and wise). For, the critic might say, in the barn-case, ‘wide’ and ‘red’ are indeed distinct attributes of the barn, because the barn is not an undivided, simple entity. Hence, there is nothing objectionable in the conclusion that we predicate a plurality of attributes of the barn (‘wide’ and ‘red’), but it is incoherent to predicate such a plurality of God, who is supposed to be simple.

We can respond to this objection by availing ourselves of another central Fregean distinction, that between sense and reference.Footnote 29 ‘Good’ and ‘wise’ both have the same referent: the simple, undivided divine being. In that sense, the reference works as in ‘morning star/evening star’. However, the doctrine of simplicity does not force us to say that we mean the same thing when we say ‘God is wise’ and when we say ‘God is good’. Rather, these are just terms that pick out the same object (God) under different descriptions.Footnote 30 Hence, divine simplicity is compatible with the multiplicity of divine attributes that finite human minds employ to talk about God, because all those attributes have the same, simple referent while having a different sense.

Now, as noted earlier, the Fregean distinction between sense and reference largely corresponds to Aquinas’s distinction between concept (ratio) and the thing that is signified by the concept (res significata).Footnote 31 As critics have pointed out, however, the Fregean distinction does not yet explain how theological predicates can have sufficiently determinate content and truth-conditions. That is, it does not yet provide us with a theological theory of the meaning of divine attributes.Footnote 32 Thus, the critic’s fundamental concern about semantic vacuity remains. Answering these worries about meaning requires turning from semantics to metaphysics: Fregean diagnostics must be embedded in a broader metaphysical framework, e.g. analogy. This framework will provide a basis for a theological account of meaning and also supply theological predication with some salient constraints.

4. Semantic logic within metaphysics: how divine metaphysics grounds analogical semantics

Our aim in this section is not to rehearse a generic version of the Thomistic doctrine of analogy but to demonstrate how this doctrine can answer the residual worries about meaning from Section 3 by providing principled constraints and sufficiently determinate content for divine predicates. This provides further theological basis for the use of Fregean diagnostics to clear the conundrums introduced earlier. It will also show why the Fregean distinctions introduced in Section 3 are not an external imposition, but a helpful way of making explicit what the Thomistic account already presumes in some sense.

This being said, the discussion in the previous section also demonstrated that Frege does have something genuinely interesting to contribute to a Thomistic defence of simplicity, namely, his distinction between concepts and objects. Unlike the sense/reference distinction, which is largely already in Aquinas, the concept/object distinction as such is not found in Aquinas, and it is also rarely discussed in the context of debates concerning divine simplicity. Hence, where the ratio is quite close to Fregean sense because both concern conceptual content or mode presentation, the Fregean notion of ‘concept’ is distinct from these, because it is not to do with meaning. Rather, the Fregean concept is a logical and semantic entity signifying a predicative expression that is logically incomplete (or ‘unsaturated’, as Frege sometimes puts it). Aquinas does, of course, make a distinction between that of which something is predicated and what is being predicated, but he does not quite have a Frege-like sharp logical distinction between complete referential expressions and unsaturated predicative expressions.

However, this strategy, like the sense/reference strategy, remains incomplete because it cannot fully address residual worries about the content of theological predicates. For that, we must bring in a further distinction from Aquinas that Frege does not have, namely the distinction between what is signified (res significata) and the mode of signification (modus significandi).Footnote 33 When it comes to applying such terms as ‘good’ to God, the mode of signifying is creaturely (i.e., speaking of goodness as a property that inheres in an entity), but what is signified (res significata) is truly found in God, except in a metaphysical mode that we cannot properly conceive. We can only say negatively that this mode is not such that it entails any kind of composition.

Here is also where analogy, in its semantic dimension, finds its originating principle: for creaturely predication is made for a world such as ours – a world of multiplicity and difference – and can, therefore, only speak of God in a creaturely mode. ‘Good’ and ‘wise’ are distinct concepts because our intellect abstracts different aspects from creaturely perfections; in God, there are not distinct aspects but one act.Footnote 34 The plurality, in other words, is in the mode of signifying, not in the signified reality. As Eleonore Stump puts it: ‘finite human minds contemplating the deity must break the one divine nature into separable attributes; limited human minds cannot take in as one whole the greatness of the divine nature’ and so ‘human minds, like a prism, break the white light of God’s nature into a rainbow of divine attributes’.Footnote 35 Or, indeed, as Aquinas himself says it:

We can speak of simple things only as though they were like the composite things from which we derive our knowledge. Therefore in speaking of God, we use concrete nouns to signify His subsistence, because with us only those things subsist which are composite; and we use abstract nouns to signify His simplicity. In saying therefore that Godhead, or life, or the like are in God, we indicate the composite way in which our intellect understands, but not that there is any composition in God.Footnote 36

According to the doctrine of analogy, then, ‘good’ has a different meaning when applied to God but this difference is not random. Rather, ‘good’ when applied to God has a different meaning in just this way: that God does not have goodness as a property. God is all that God is in one single, undivided actus purus. However, while we might be capable of negatively characterising God’s distinct mode of being in abstract terms, we can hardly imagine a simple mode of existence or express it linguistically.Footnote 37 Our language and thinking, made as they are for multiplicity, lack the conceptual resources for that. This is why the meaning of goodness is not exactly the same when applied to creaturely beings and when applied to God; this is why there is a similarity in the midst of even greater dissimilarity (as Lateran IV would have it).

It is important to make this point, because sometimes it is assumed by the critics of analogy that the difference involved in analogy is just any kind of difference and we have no way of specifying what this difference is. And so it is argued that analogy obscures the meaning of terms when applied to God: human goodness might, for example, mean that one does not needlessly cause the innocent to suffer, but we have no way of knowing if divine goodness entails the same.Footnote 38 Following this, such critics as Thomas Williams think that we must be able to spell out the difference between ‘wise’ in ‘God is wise’ and ‘wise’ in ‘Socrates is wise’ in univocal terms, or we’ll fall into an infinite regress of meaning that renders our language about God unintelligible.Footnote 39 Critics such as Rory Misiewicz locate the issue more explicitly in divine simplicity because divine simplicity entails that God is not a member of any genus and, therefore, we do not know how terms such as ‘good’ apply to God:

If we talk of God’s love, power, or goodness, and that involves operating with some particular sense or senses associated with those terms (say, a parent’s love, a monarch’s power, or a hero’s goodness), we must immediately rescind whatever sense is cognized, for otherwise we would place God in a species of a genus (say, God’s love is that of a parent).Footnote 40

According to Misiewicz, the defenders of analogy would respond to this by arguing that attributes such as ‘love’ are not entirely senseless when applied to God; we just do not know ‘the particular sense in which God’s love, power, or goodness obtain’ with respect to God.Footnote 41 Nevertheless, ‘we can trust that there is something beside the unlikeness of analogy that renders our language intelligible’.Footnote 42 This is true, but the thing that leads Misiewicz astray is thinking that the defenders of analogy have no way of specifying what this something is.Footnote 43

I reckon that the reason Misiewicz, Williams and others think of the difference in analogy as unspecified – a semantic can of worms that opens up the possibility of anything goes predication – is that they treat the semantic dimension of analogy in isolation from its metaphysical background. They miss the fact that analogy is a semantic thesis that is grounded in a primary metaphysical thesis: creatures participate in perfections pre-contained in God.Footnote 44 This is the something that grounds the intelligibility of theological language: because God is the creator of all things, all things – to the degree that they have being – reflect God. This also puts a stop to the kind of semantic regress Williams worries about: there is a fixed metaphysical relation between God and creatures that grounds the meaning of analogical predicates. This metaphysical relation is not itself an intrinsic perfection predicate, but an extrinsic denomination. Its meaning is, therefore, not analogical.

It is, therefore, not just that certain potent creaturely realities – parent’s love, monarch’s power, hero’s goodness – can be, as it were, images of God’s goodness. In fact, strictly speaking, it is not the case that individual entities are analogies of God.Footnote 45 Rather, being itself is analogical. All things – spiders, apples, trees, galaxies, human persons – reflect the goodness of God because all things come from God. Since all things receive their being from God, the perfections of all created things exist in God pre-eminently. Everything together – in its manifold richness – reflects the white light of one, simple divine essence. So yes, as simple God is beyond all genera in the sense that God does not belong in any genus, but this is not to say that God is not, as it were, reflected in all genera. It is precisely this – that God is not in any genus but the source of all genera – that requires a distinct mode of predication when it comes to God, i.e., an analogical mode. As Aquinas says:

Everything is therefore called good from the divine goodness, as from the first exemplary, effective, and final principle of all goodness. Nevertheless, everything is called good by reason of the similitude of the divine goodness belonging to it, which is formally its own goodness, whereby it is denominated good. And so of all things there is one goodness, and yet many goodnesses.Footnote 46

Hence, analogy – for Aquinas, anyways – is not simply a matter of causality.Footnote 47 If by ‘God is good’ we meant only that ‘God is the cause of goodness’, then goodness would, in fact, always be first spoken of creatures and spoken of God only in a secondary sense. However, God is not simply the cause of goodness, but rather God is the one who is essentially Good. And so, since goodness is contained primordially in God, it is always first spoken of God. Or, to be more precise, in the order of being, goodness is always first spoken of God because God is the primordial goodness from whom flow all creaturely goodnesses. In the order of knowledge, we know that God is good by knowing goodness first from creatures and by then ascending from creatures to divine goodness via analogy. We can therefore know, through analogical reasoning, that God’s goodness means that God would never needlessly cause the innocent to suffer. We know this by paying attention to the world God has created (including the kinds of human moral goodness that it contains).

This metaphysical background explains why divine simplicity does not require ‘immediately rescinding whatever sense is cognized’. It requires, rather, recognising that creaturely goodnesses reflect God in their limited mode because creatures have their goodness from God participatorily while God is good essentially. That is, God just is essentially good while creatures are good in a composite way that involves, among other things, movement from potentiality to actuality (i.e., creaturely things attain to their goodness). Thus, the semantic difference in ‘God is good’ and ‘Socrates is good’ stems from the different metaphysical modes of God’s being good and Socrates’s being good. This is not to say that metaphysics replace semantics. Rather, it explains why non-univocal semantics is required.

By situating analogy within the appropriate metaphysical framework, we can also come to see how, within the Thomistic system, there are certain constraints that apply to theological predication. To see how, consider the following sentences that have God as their subject: (i) God is good, (ii) God is a chair, and (iii) God is capable of doing evil. The critics of analogy would like to suggest that analogy results in a semantic indeterminacy that does not allow us to explain why (i) is true while (ii) and (iii) are not. However, according to the Thomistic framework proposed here, only predicates that signify pure perfections – such as goodness, wisdom, knowledge, being, and life – can be predicated of God analogically since such perfections exist in God eminently and without limitation. By contrast, ‘God is a chair’ involves a category mistake because being a chair entails a material mode of being that God does not have. ‘God is capable of doing evil’, on the other hand, involves the incoherence of assuming that God can fail to be what God is or lack a perfection.Footnote 48 These constraints on theological predication are not ad hoc – they follow coherently from the metaphysical framework in which the Thomistic doctrine of analogy is embedded.Footnote 49

Lastly, for this section, let us consider two objections. Firstly, there is the objection that all predication requires a subject who exemplifies a property.Footnote 50 If God is simple and, therefore, does not exemplify a property, then we have no way of predicating anything of God and no way of knowing what is the truth-value of any proposition p in the form of ‘God is X’ because we do not know what in God renders p true. Following Bergmann and Brower, we can solve this issue by adopting a truth-maker view of predication when it comes to divine predicates: it is God in Godself – the simple divine being – that functions as the truth-maker for p.Footnote 51 For God to be a truth-maker for p does not require that God exemplifies X because ‘to characterise an entity as a truth-maker is to characterise it in terms of a certain function or role – that of necessitating the truth of the predications it makes true’ rather than in terms property exemplification.Footnote 52 As Bergmann and Brower write:

From the perspective of the truthmaker theory of predication, there is nothing problematic about saying that God is identical with his nature, goodness, power, wisdom, or any other such things. To put the point more carefully, there is nothing problematic about saying that God is identical with the referents of abstract singular terms corresponding to each of the true intrinsic predications that can be made about him.Footnote 53

Another objection that can be raised against all this is that we have, in articulating this metaphysical framework that grounds analogical semantics, tacitly placed God in a genus by, for example, characterising God as ‘a truth-maker’ or ‘a source’. Are we not thereby placing God in the genus of truth-makers or sources? This, I reckon, is ultimately the concern of both Williams and Misiewicz: analogy is inherently incoherent because we do not seem to be able to characterise the doctrine itself without falling back on some apparently univocal predicates that place God in a genus (such as ‘truth-maker’, ‘source’, and ‘being participated in’).

The key mistake in this objection is the assumption that every true predication must be classificatory – i.e., it must locate its subject within a genus by attributing it an intrinsic feature. We can answer this concern by distinguishing between intrinsic perfection predication, extrinsic denominations, and metatheoretical role descriptions. Analogical predication applies to the first of these because intrinsic perfection predication (God is ‘good’, ‘wise’, ‘simple’) involves statements about what God is in Godself – God’s simple essence. To characterise God as the source of creaturely perfections, however, is not to add an additional perfection to God and thereby characterised God’s simple essence as such – it is to simply characterise a metaphysical dependence relation between God and creatures. It is, in other words, to assign a certain metaphysical role to God and to thereby extrinsically denominate God based on God’s effects. Similarly, on a truth-maker approach: to say that ‘God is good’ is true is to say that God’s very being is sufficient for (and thus metaphysically grounds) the truth of the proposition. This is a claim about what makes the sentence true, not an additional intrinsic predicate that places God in a further category. This is why Bergmann and Brower characterised truth-makers ‘in terms of certain function or role’.Footnote 54 Thus, to speak of God as ‘the truth-maker of p’ is to speak of God in terms of metatheoretical role descriptions.

5. Conclusion: meaning grounded in practice

In the previous section, I employed a generic version of the Thomistic doctrine of analogy to answer worries about semantic meanings that the Fregean diagnostics alone could not address. Critics such as Misiewicz and Williams might still be unsatisfied, however. They might accept that the Fregean object–concept distinction that was introduced in Section 3 can clear some initial logical inconsistencies that simplicity implies – such as that in God ‘goodness’ is the same thing or property as ‘wisdom’ – but this still leaves theological predication too much in the dark: if divine simplicity implies that perfection terms apply to God analogically there is still too much indeterminacy on the level of meaning. What, they might ask, does it actually mean to say that God is good?

My main issue with this objection is that it involves a false dichotomy: a theological theory of meaning must either be able to provide the semantic tools for determinately fixing the meaning of theological predicates or accept that theological predicates are semantically vacuous. I would like to suggest, however, that it is unlikely that any theological theory of meaning can determinately fix the meaning of theological predicates because our understanding of the meaning of such predicates as ‘good’ when applied to God is always incomplete – what finite mind, after all, could comprehend infinite goodness? Thus, there is always – regardless of the theoretical framework adopted – a semantic openness to theological predication.

Therefore, in my view, a better standard of success for a theological theory of meaning is that our theory should provide theoretical grounding for thinking of theological meanings as sufficiently determinate. By sufficiently determinate, I mean that our theory of meaning should give us more than ad hoc grounds for excluding certain statements about God as false and for accepting other statements as true but incomplete or semantically open. I think that the Thomistic doctrine of analogy can meet this standard.

Ultimately, our understanding of such divine predicates as ‘good’ or ‘wise’ emerges in practice – that is, through lived engagement with resources of the Christian form of life. The doctrine of analogy, too, is therefore more than a set of logical rules the application of which automatically produces correct meanings. Rather, analogy is a metaphysical doctrine that grounds certain religious practices of sense-making and language-use that cannot be reduced to logical principles. As Burrell writes:

Whether we use perfection expressions properly when we use them of God depends on acquired skill: knowing how to respect the grammatical difference which logic demands for discourse in divinis. Genuinely learning how to respect these difference [sic], however, is more than just a matter for logic. That task requires the disciplines traditionally associated with religious living and practice. For attending to differences, we must call upon the metaphorical resources of analogous terms. And the formation of apt metaphors is a matter for sensitivity, not for science.Footnote 55

Analogy, then, cannot be used as a mechanistic logical device to determinately fix the meaning of theological terms in such a way that no semantic overplus remains. Rather, analogy invites us to approach theological thinking and language as a continuous journey – the mind’s pilgrimage that cannot be completed on this side of death. Given that God’s simple mode of being – and, therefore, the ultimate truth-maker of theological predicates – is inconceivable, there remains an important element of apophasis to theological language. Together, analogy and apophasis mark the contours of divine mystery – that which ultimately eludes our conceptual grasp and linguistic expression. And so, incompleteness remains a feature of theological language. The need to engage in formative practices that can give rise to the kind of practical sensitivity analogy requires also remains. As Joshua Lee Harris puts it:

Perfections such as goodness and life are predicated literally of God, but it is important to note that these perfections are only understood to a limited degree. That is, because our modus significandi is irreducibly composite and therefore inadequate to the task of representing God as he is in himself, it is also the case that it is inadequate to the task of grasping divine perfections as they are in themselves. After all, on Aquinas’s doctrine of divine simplicity, we know that whatever is predicated literally of God is simply identical with God. Goodness, life and the like are ultimately expressions of the divine essence itself. To the extent that we fail to comprehend the divine essence, then, we also fail to comprehend these divine perfections.Footnote 56

I’d therefore also like to suggest that the doctrine of analogy is salutary because, rather than inviting the surrender of all meaning, it invites us to consider the meaning of theological language as a continuous process of ever-increasing understanding.Footnote 57 This understanding cannot be separated from praxis: we cannot simply fix the meaning of theological terms in the abstract without the corresponding formation of our thinking selves. Nor can one skirt the difficulties of talking rightly about an object like God by simply adopting a different theory of theological predication as if talking rightly about God was simply a matter of logical rigour rather than holistic formation.

So then – divine simplicity entails that divine attributes all reliably pick out the same, simple object under plural senses. This is the insight recourse to Frege provided. The Fregean distinctions introduced in Section 3 also helped us to clear some apparent logical incoherencies that seemed to follow from simplicity – such as that we would be forced to say things such as ‘wisdom is power’ or ‘omnipotence is loving-kindness’. At the same time, I argued that Fregean diagnostics alone cannot provide sufficient grounding for theological meanings – for that analogy is required. When it comes to theological predication, then, it’s not that the meaning of ‘good’ and ‘wise’ is the same, but that their referent is the same while their meaning is analogical.

The central claim of this paper, then, has been that divine simplicity does not render theological language vacuous, but instead requires a disciplined, analogical mode of predication grounded in participatory metaphysics. Now, this being said, analogy does not give us a complete conceptual grasp or control over theological meanings. However, nor does univocity (for one can also be mistaken about univocal meanings or have a severely incomplete understanding of a concept). This lack of control is salutary: in addition to being able to elicit the development of certain kinds of virtues (such as humility), it can pull us towards an encounter with the semantic excess of God’s superabundant Life. Encountering this Life, we might come to know ourselves not just as those who address God with the help of various predicates but as those who are, first and foremost, transformatively addressed by God.Footnote 58

AI statement

This article has been written by the author, and AI tools have not been used to generate content for it. However, during the editing process, the author used ChatGPT-5 to improve some aspects of readability and language. The author takes full responsibility for the contents of this article.

Declaration of interest

There are no conflicts of interest or funding to declare.

References

1 As discussed in, for example, William Mann, ‘Divine Simplicity’, Religious Studies, 18 (1982), 451–71; Eleonore Stump and Norman Kretzmann, ‘Absolute Simplicity’, Faith and Philosophy, 2 (1985), 353–82; Gregory P. Rocca, Speaking the Incomprehensible God: Thomas Aquinas on the Interplay of Positive and Negative Theology (Washington, DC: Catholic University of America Press, 2004), p. 293; Barry Miller, A Most Unlikely God: A Philosophical Enquiry into the Nature of God (Notre Dame: University of Notre Dame Press, 1996); Jordan P. Barrett, Divine Simplicity: A Biblical and Trinitarian Account (Minneapolis: Fortress Press, 2018), p. 90; Brian Leftow, ‘Divine Simplicity’, Faith and Philosophy, 23 (2006), 365–80; Jeffrey E. Brower, ‘Making Sense of Divine Simplicity’, Faith and Philosophy, 25 (2008), 3–30.

2 The only author I’m aware of who has discussed the Fregean concept–object distinction in relation to arguments against divine simplicity is Miller (A Most Unlikely God, pp. 47–59). I largely agree with Miller; this paper adds to the discussion by addressing some more recent criticisms of divine simplicity and by connecting analogy more directly with Aquinas’s own metaphysics. It also grounds analogical use in practice in Section 5, something Miller does not explicitly do. I thank the anonymous reviewers for pointing me to Miller.

3 I use the term ‘determinate’ in a semantic sense: if some semantic item (predicate, sentence, etc.) is semantically determinate, its meaning or content (I use these terms interchangeably) is fixed or complete. There is a debate as to whether complete semantic determinacy is possible (cf. Delia Belleri, Semantic Under-Determinacy and Communication [London: Palgrave Macmillan, 2014]). I will specify later what I mean by sufficiently determinate in a theological context, but for now it suffices to say that I use the term broadly to refer to semantic items whose meaning or content is not entirely vacuous (but not necessarily complete or fixed either). Thus, when I say that the meaning of theological predicates is sufficiently determinate, I do not mean to imply that there are property relations in God between determinates and determinables, see ft. 28 later.

4 Classic detailed studies of Aquinas’s doctrine of simplicity and its relation to language (Peter Weigel, Aquinas on Simplicity: An Investigation into the Foundations of His Philosophical Theology [Oxford: Peter Lang, 2008]; W. J. Hankey, God in Himself: Aquinas’ Doctrine of God as Expounded in the Summa Theologiae [Oxford: Oxford University Press, 2004]) do not assume a strongly defensive task with respect to contemporary analytic criticisms and they usually do not make explicit use of Frege, although Weigel (Aquinas on Simplicity, p. 231) does discuss Frege’s notion that existence is a second-level predicate and its implications for the doctrine of simplicity. Those classic defences of divine simplicity that discuss contemporary criticism more explicitly (Brian Davies, ‘A Modern Defence of Divine Simplicity’, in Philosophy of Religion: A Guide and Anthology [Oxford: Oxford University Press, 2000]; Nicholas Wolterstorff, ‘Divine Simplicity’, Philosophical Perspectives, 5 (1991), 531–52; Stump and Kretzmannn, ‘Absolute Simplicity’) do not discuss Frege’s object–concept distinction in relation to something such as the Conundrums A and B in this paper, although Davies does discuss existence as a second-level predicate and Stump and Kretzmann mention the sense/reference distinction. In addition (as discussed by Katherin Rogers, ‘The Traditional Doctrine of Divine Simplicity’, Religious Studies, 32 (1996), 165–86), Wolterstorff does not quite get Aquinas’s Aristotelian metaphysics right; pace Wolterstorff, Aquinas did not think of natures as concrete existing things (although, in so far as a particular existing individual has a nature that nature can be thought of as existing as a concrete thing). Some contemporary defenders of simplicity have focused especially on defending divine simplicity on a dogmatic and biblical basis (Stephen Duby, Divine Simplicity: A Dogmatic Account [New York: T&T Clark Bloomsbury]; D. Stephen Long, The Perfectly Simple Triune God: Aquinas and His Legacy [Minneapolis: Fortress Press, 2016]; Barrett, Divine Simplicity) none of them – as far as I’m aware – make use of Frege’s object/concept distinction.

5 Part is here understood as anything or principle that combines with another thing or principle to make a whole. Since my interest is in defending a broadly Thomistic understanding of simplicity, I’ll leave aside Scotus and formal distinctions and the question of whether formal distinctions also constitute parts in the relevant sense.

6 St. Thomas Aquinas, Summa Theologiae (trans. The English Dominican Fathers, 1920), I-I, q. 3, a. 7.

7 Aquinas, ST I-I, q. 3, a. 1.

8 David B. Burrell, Aquinas: God and Action (Scranton and London: University of Scranton Press, 2008), p. 66.

9 Note, pace Rory Misiewicz (‘Divine Simplicity’, International Journal of Philosophy and Theology, 82 (2021), 51–63, p. 53), simplicity does not therefore merely result from ‘God being conceived as a perfect being’. This is important because sometimes critics of classical theism assume that classical divine attributes stem from a specific (outdated) understanding of perfection and that the relevant conception of perfection can, therefore, be exchanged for another conception of perfection based on one’s intuitions about what makes for perfection in beings. Another thing to note at this stage is that, pace Yann Schmitt (‘The Deadlock of Absolute Divine Simplicity’, International Journal for Philosophy of Religion, 74 (2013), 117–30), simplicity does not entail that God has one ‘super-property’. Rather, it means that God does not exemplify any properties because property exemplification presupposes a distinction between the property that is exemplified and the subject that exemplifies it.

10 For more details, see Thomas Joseph White, ‘Divine Simplicity and the Holy Trinity’, International Journal of Systematic Theology, 18 (2016), 72–82.

11 This being said, it should be noted that negative attributes are not privative: they signify the plenitude of divine life that exceeds our capacity to express it.

12 This would be true even if God was somehow eternally composed of such properties as omniscience, omnibenevolence, omnipresence, and omnipotence: even then, these properties would be more fundamental than God because God would only exist if these properties came together.

13 There also other conundrums – such as modal collapse (Ryan T. Mullins and Shannon Eugene Byrd, ‘Divine Simplicity and Modal Collapse: A Persistent Problem’, European Journal for Philosophy of Religion, 14 (2022), 21–52) – that simplicity seems to produce, but they are not the topic of this paper. For an overview of various arguments against simplicity, see William F. Vallicella, ‘Divine Simplicity’, in The Stanford Encyclopedia of Philosophy, ed. by Edward N. Zalta and Uri Nodelman (2023). <https://plato.stanford.edu/archives/win2023/entries/divine-simplicity/> and James E. Dolezal, God without Parts: Divine Simplicity and the Metaphysics of God’s Absoluteness (Eugene: Pickwick Publications, 2011).

14 As Aquinas puts it (ST I-I q. 13, a. 5): ‘[F]or instance, by the term “wise” applied to man, we signify some perfection distinct from a man’s essence, and distinct from his power and existence, and from all similar things; whereas when we apply to it God, we do not mean to signify anything distinct from His essence, or power, or existence. Thus also this term “wise” applied to man in some degree circumscribes and comprehends the thing signified; whereas this is not the case when it is applied to God; but it leaves the thing signified as incomprehended, and as exceeding the signification of the name’.

15 Alvin Plantinga, Does God Have a Nature? (Milwaukee: Marquette University Press, 1980), p. 47.

16 In addition, the Thomists are, of course, also committed to language about Trinitarian divine persons – in that context, however, the term ‘person’ is not employed in the same sense as theistic personalists employ it (e.g., to indicate that God is, for example, more or less like a human person except that God has limitless mental capacities and is disembodied (cf. Richard Swinburne, The Coherence of Theism [Oxford: Clarendon Press, 1993], p. 1). Rather, the term ‘person’ takes on a distinctly theological meaning.

17 In fact, the Thomist would deny that God is a thing, a nature or an abstract property for the same reason the Thomist would deny that God is a person: all these would entail that God belongs in some genus.

18 Versions of this conundrum are articulated and discussed by various authors such as Charles Hodge, Systematic Theology, 3 vols (Grand Rapids: Eerdmans, 1965), pp. 369–72; Ronald H. Nash, The Concept of God (Grand Rapids: Zondervan, 1983), pp. 88–89; Christopher Hughes, On a Complex Theory of a Simple God: An Investigation in Aquinas’ Philosophical Theology (Ithaca: Cornell University Press, 1989); Wolterstorff, ‘Divine Simplicity’; Miller, A Most Unlikely God; Thomas V. Morris, Our Idea of God: An Introduction to Philosophical Theology (Downers Grove, IL: InterVarsity Press, 1991), p. 115; Anders Kraal, ‘Logic and Divine Simplicity’, Philosophy Compass, 6 (2011), 282–94.

19 This solution is inspired by a paper given by Nicholas E. Lombardo OP at Campion Seminar at Oxford titled The Grammar of Divinity: A Genealogy of Confused Thinking About the Divine Essence in Western Trinitarian Theology. I employ elements of his argument concerning Trinitarian grammar to solve the conundrums just introduced.

20 Gottlob Frege, ‘On Concept and Object’, in Translations from the Writings of Gottlob Frege (Oxford: Blackwell, 1960), pp. 42–55.

21 Note that this is a statement about language; this is not to say that God is an object in the sense of being a discrete, spatio-temporally located entity.

22 Frege’s terms for the distinction between objects and concepts is to also say that concepts are ‘unsaturated’ while objects are complete (Gottlob Frege, ‘Function and Concept’, in Translations from the Writings of Gottlob Frege [Oxford: Blackwell, 1960], pp. 21–41, p. 24).

23 Notice that here ‘concept’ is used in the logical sense only and not in the sense of, say, a mental representation or an idea in the mind.

24 For example, consider: ‘Mars [object] is a planet [concept]’ versus ‘the planet [object] is beautiful’. Or: ‘red is a colour [concept]’ versus ‘this colour [object] is bright’.

25 Frege, ‘On Concept and Object’, pp. 43–44.

26 Aquinas, ST I-I, q. 3, a. 3; I, q. 13, a. 3. Cf. also Duby, Divine Simplicity, pp. 226–27.

27 With respect to Plantinga’s argument that is the basis of Conundrum A, there have, of course, also been other kinds of responses. Some (William Vallicella, ‘Divine Simplicity: A New Defense’, Faith and Philosophy, 9 (1992), 508–25; Mann, ‘Divine Simplicity’; Brian Leftow, ‘Is God an Abstract Object?’, Noûs, 24 (1990), 581–98) have suggested that we should, indeed, interpret the doctrine of simplicity as implying that God is identical with a property; they then proceed to construe divine properties in such a way that identifying God with them is compatible with maintaining that God is a person. Brower (‘Making Sense of Divine Simplicity’) and Rogers (‘The Traditional Doctrine of Divine Simplicity’) have critiqued these approaches on the basis that the traditional doctrine of simplicity does not identify God with God’s properties – rather, it denies that God has any properties at all. My response differs from all these in that it employs the Fregean object–concept distinction, thereby diagnosing Conundrum A as a semantic conundrum rather than a metaphysical puzzle. It also explicitly situates its defence of divine simplicity within the framework of analogical metaphysics.

28 Schmitt (‘The Deadlock of Absolute Divine Simplicity’) presents a variation of Conundrum B. He assumes that the defenders of absolute (i.e., Thomistic) divine simplicity are committed to God having ‘the super-property of being the maximal determinate of all the determinates of being perfect’ (ibid., 120) so that in both ‘God is good’ and ‘God is wise’, ‘good’ and ‘wise’ refer to the same super-property. Thus, ‘God is good’ and ‘God is wise’ are a form of object–object predication where the object is identified as one super-property. This, then, undermines divine simplicity because determinable–determinate relations generate property plurality (e.g., if ‘God is perfectly wise’ then God has the properties of ‘wisdom’ and ‘being perfectly wise’). On the Thomistic account defended in Section 3, however, divine predicates are analogical and do not function as determinates of a shared determinable in God.

29 Gottlob Frege, ‘Sense and Reference’, in Translations from the Writings of Gottlob Frege (Oxford: Blackwell, 1960), pp. 56–78

30 Beyond Frege’s distinction, I haven’t committed myself to any particular theory of reference (and I need not since Frege’s distinction is not dependent on any particular theory of reference for its validity). For an overview of various theories of reference in a religious context, see Michael Scott, Religious Language (Basingstoke: Palgrave Macmillan, 2013).

31 Cf. Rocca, Speaking the Incomprehensible God, p. 293. For Aquinas’s distinction between ratio and res significata, see Aquinas, ST I-I, q. 13, a. 4; Aquinas, Summa Contra Gentiles, ed. by Joseph Kenny OP, trans. by Anton C. Pegis (New York: Hanover House, 1955–1957), I.35–36. See also, Aquinas, Questiones Disputatae de Veritate, trans. by Robert W. Mulligan SJ (Chicago: Henry Regnery Company, 1952), q. 2, a. 1: ‘These attributes [that are used of God] do not differ as regards the reality which is signified, but only in our manner of understanding them. For God’s essence, life, knowledge, and whatever else of this sort that may be predicated of Him are all the same; but in understanding essence, life, and so forth in His regard, our intellect has different concepts for each.’

32 Schmitt (‘The Deadlock of Absolute Divine Simplicity’) notes this incompleteness of Fregean approaches to divine simplicity and then proceeds to argue that all attempts to ground content without sacrificing simplicity (such as minimal truth-makers, trope theory, property exemplification, divine attributes as determinables of a super-property) fail. Schmitt does not, however, discuss Thomistic analogy in any depth – in what follows, I’ll demonstrate how the doctrine of analogy can ground the meaningfulness of theological language without implying composition in God.

33 For this distinction in Aquinas, see ST I-I, q. 13, a. 3; SCG I.30; De Pot. q. 7, a. 2, a. 4. In some sense, the res significata/modus significandi distinction presupposes the ratio/res significata distinction: whereas the latter is utilised to explain how concepts with plural meanings can successfully refer to a simple referent, the former encapsulates the creaturely – and, therefore, composite – character of our language and thinking more generally. That is, ratio/res significata distinguishes a conceptual account of X from X that is being signified. The res significata/modus significandi distinguishes the X that is signified from the linguistic/intellectual mode under which it is signified. Indeed, one could suggest that in Aquinas earlier work, such as Summa Contra Gentiles, these distinctions are not yet fully distinguished from each other. Cf. SCG I.36: ‘For, although our intellect arrives at the knowledge of God by various conceptions, as stated above, it understands that what corresponds to them all is absolutely one, because our intellect does not ascribe its mode of understanding to the things which it understands, even as neither does it ascribe immateriality to a stone, although it knows it immaterially.’ For a discussion and overview of the development of Aquinas’s thought on this, see Weigel, Aquinas on Simplicity, ch. 6.

34 This is not to say that Thomists deny all real distinctions in God, as they do affirm real distinctions between the divine persons. Within the limits of this article, however, we cannot address in detail the relationship between the doctrine of simplicity and the doctrine of Trinity. For discussion, see, for example, Réginald Garrigou-Lagrange, The Trinity and God the Creator: A Commentary on St. Thomas’ Theological Summa, Ia, q. 27–119 (St. Louis/London: B. Herder, 1952); Robert M. Burns, ‘The Divine Simplicity in St Thomas’, Religious Studies, 25, no. 3 (1989), 271–93; Hankey, God in Himself; D. Stephen Long, ‘Thomas Aquinas’ Divine Simplicity as Biblical Hermeneutic’, Modern Theology, 5 (2019), 496–507; Jonathan M. Platter, Divine Simplicity and the Triune Identity (Berlin/Boston: De Gruyter, 2021).

35 Eleonore Stump, Atonement (Oxford: Oxford University Press, 2018), p. 409.

36 Aquinas, ST I-I, q. 3, a. 3.

37 Interestingly, Plantinga (Does God Have a Nature?, p. 53) comes surprisingly close to realising something like this as he suggests in an off-hand remark that perhaps Aquinas, in arguing that God is simple, ‘doesn’t mean to identify God with a property or state of affairs at all, but with something quite different. If so, it is not easy to see what sort of thing it might be’. That it is not ‘easy’ – or that it is, in fact, impossible – to say ‘what kind of thing’ God is precisely Aquinas’s point. In some sense, Stephen R. Holmes (‘“Something Much Too Plain to Say”: Towards a Defence of the Doctrine of Divine Simplicity’, Neue Zeitschrift fur Systematische Theologie und Religionsphilosophie, 43 (2000), 137–54) is right, then, in arguing that much of the confusion here lies in not appreciating the apophatic character of the classical doctrine of simplicity.

38 Sometimes this assumption is made also by those who are sympathetic to the doctrine of analogy. Cf. Susannah Ticciatti (A New Apophaticism: Augustine and the Redemption of Signs [Leiden: Brill, 2013], p. 38) who writes of Thomistic doctrine of analogy: ‘The result can only be that we lose our hold on the meaning of words as we use them of God.’

39 Thomas Williams, ‘The Doctrine of Univocity Is True and Salutary’, Modern Theology, 21 (2005), 575–85.

40 Misiewicz, ‘Divine Simplicity’, p. 57.

41 Ibid., p. 57.

42 Ibid., p. 57. Italics mine.

43 More specifically, Misiewicz argues that Cajetan’s version of analogy as a proportional unity fails if God is not in any genus (as simplicity would require us to say). In summary, Cajetan argued in De Nominum Analogia (1498) that there is an analogical relation R₃ between two kinds of relation, creaturely relation R1 (A: B) and divine relation R2 (C: D). For example: analogical concept R3 (goodness) is drawn from the proportional unity between creaturely concept of goodness R1 (moral perfection: human goodness) and the concept of divine goodness R2 (metaphysical perfection: essential goodness), because ‘moral perfection’ stands in a similar relation to ‘human goodness’ as ‘metaphysical perfection’ does to ‘essential goodness’. Misiewicz says (ibid., pp. 56–60) that making this relational comparison requires putting God in a genus because proportionality presupposes measure and comparability (and an entity can only have these if it belongs in a genus). In addition, Cajetan’s ‘abstraction by confusion’ (e.g., the strategy of taking the relations involved in R1 and R2 and blurring the differences between them so as to arrive at a higher-order formal resemblance R3) presupposes compositionality (in terms of distinguishing the relata), which contradicts simplicity. Thus, there can be no analogy between some creaturely entity X and a simple God. I think we can respond to these counterarguments by paying attention to the metaphysical character of analogy. Firstly, in relation to measure and proportionality: the proportional unity is not an arithmetic one but rather a metaphysical one, stemming from God’s causal relationship to creation. It is therefore a sui generis relation – indeed, it is a sui generis relation precisely because God does not belong in any genus. Secondly, as to analogy seemingly presupposing composition: here we must bear in mind that analogical predication is a way our creaturely concepts can be predicated of God. The concepts – and, therefore, analogical predication – presupposes composition, but this does not mean that composition must be read into God in Godself. In fact, it is precisely because there is no composition in God that the predication in question is analogical. When it comes to abstraction by confusion, we are not, therefore, abstracting from God but from creaturely concepts (which are composite). For more on Cajetan and analogy, see Joshua P. Hochschild, The Semantics of Analogy: Rereading Cajetan’s De Nominum Analogia (Notre Dame, IN: University of Notre Dame Press, 2010). (It should be noted that although I offer this brief defence of Cajetan, in this paper, I’m not committing myself more specifically to Cajetan’s version of analogy.)

44 Cf. Philip A. Rolnick, Analogical Possibilities: How Words Refer to God (Atlanta: Scholars Press, 1993), pp. 41–42; W. Norris Clarke SJ, The Philosophical Approach to God: A Contemporary Neo-Thomist Perspective (Winston-Salem: Wake Forest University Press, 1979), pp. 55–56.

45 We can, of course, say that ‘God’s love is like a parent’s love’ but this is, strictly speaking again, a metaphor – an illuminating but non-literal way of talking about divine love.

46 Aquinas, ST I-I, q. 6, a. 4.

47 Ibid., q. 13, a. 6.

48 What is presupposed here is an understanding of evil as a metaphysical privation.

49 In addition, negative terms – such as simplicity or immutability – are also strictly speaking not analogous because they do not entail predicating anything positive about God. This is not to say that sentences such as ‘God is immutable’ and ‘this stone is immutable’ are univocal in meaning. Rather, in the theological context, negative terms point to the excess of divine life that exceeds our capacity to conceptualise or linguistically express it – an excess that transcends simple negation that applies to creaturely things (e.g. either X is A or not-A; God is neither A nor not-A).

50 For an overview of these theories, see Michael Bergmann and Jeffrey E. Brower, ‘A Theistic Argument against Platonism (and in Support of Truthmakers and Divine Simplicity)’, in Oxford Studies in Metaphysics, Vol. 2 (2006), pp. 357–86. For a classic view, see David M. Armstrong, A World of States of Affairs (Cambridge: Cambridge University Press, 1997).

51 We can adopt this view with respect to God without having to adopt Bergmann’s and Brower’s view that all predication is based on truth-makers rather than property instantiation.

52 Bergmann and Brower, ‘Theistic Argument’, p. 380. Note also, pace Schmitt (‘The Deadlock of Absolute Divine Simplicity’), that nothing in this account requires the identification of distinct minimal truth-makers for distinct divine predicates; the truthmaking role is fulfilled by the simple divine being itself. Dean Da Vee (‘Why Truthmaker Theory Cannot Save Divine Simplicity’, International Journal for Philosophy of Religion, 90 (2021), 43–60) has argued that Brower is pushed towards either accepting the implausible view that ‘true essential intrinsic predications about an object are made true by that object alone, not by its instantiating essential properties’ (ibid., 49) or towards making an ad hoc exception in the case of God so that – unlike for all other objects – in God’s case essential intrinsic predications are made true by God alone and not by God’s having some essential properties. However, if we begin from taking divine simplicity seriously, then adopting the view that God provides an exception case for how truth-makers usually function is not ad hoc – it is just what we would expect to follow from God’s ontological uniqueness.

53 Bergmann and Brower, ‘A Theistic Argument’, p. 383.

54 Ibid., p. 380.

55 Burrell, Aquinas, p. 77.

56 Joshua Lee Harris, ‘Analogy in Aquinas: The Alston-Wolterstorff Debate Revisited’, Faith and Philosophy, 34 (2017), 33–56, 38.

57 We could therefore also say that analogy grounds the value-laden intelligibility of the cosmos. In addition, the doctrine of analogy might invite us to view the world in a contemplative manner as a reflection of God’s own Life; through slow, formative spiritual practices, we might acquire an epistemic-perceptual stance that allows us to see the world as richly endowed with theological meanings. These meanings can be found in anything that is part of the world insofar as it has being and in the world as a whole: for according to the doctrine of analogy, it is not simply the virtuous human being that shows us what God’s goodness is like – the vast constellations of spinning galaxies and blades of grass do too.

58 I would like to thank the two anonymous reviewers for their generous engagement with the article and for the detailed suggestions they offered.