An introduction to Theistic Conceptual Realism

One way to motivate the view that divine concepts can ground the character of particular things is to demonstrate that the conjunction of the doctrines of divine omniscience and creation ex nihilo naturally lends itself to a relationship between God, divine concepts, and particular created things that is sufficient to ground the phenomena typically associated with the problem of universals. The doctrine of creation ex nihilo says that God creates out of nothing everything distinct from Godself.Footnote ³ The doctrine of omniscience – which claims, roughly, that God knows everything that can be knownFootnote ⁴ – provides prima facie evidence for the belief that, logically prior to the act of creation, God both knows and intends to create whatever God in fact creates.Footnote ⁵ It would be surprising if it turned out that God knows what God creates only because God creates it, rather than creating it because that is what God intends (logically prior to creation) (see Ross (Reference Ross, Wainwright and Audi1986), for a defense of this surprising view). Even without a set of necessary and sufficient conditions for divine knowledge, it is not a stretch to think that if some knowledge can properly be said to be God's knowledge, then something or other in God must bear an appropriate relation to whatever the act of creation brings into existence. I will take some anthropomorphic leeway and call that something in God a concept. We need not say that this concept is chronologically prior to God's creative act, but the concept must be both logically and causally prior to God's act of creation if it is to be true that God knows what God creates rather than creating ignorantly.Footnote ⁶

From the doctrines of creation and omniscience we can infer something of the relationship that holds between the divine being and the particulars that God creates. On the one hand, God's creation entails that particulars derive their being from God. This is a causal dependence. On the other hand, God's knowledge and intentions entail some sort of correspondence between the divine concept and the particular that God creates. The particular must resemble, be like, or imitate the divine concept in accordance with which God intends to create. We expect God's will to successfully bring into being exactly that thing which God intends to create, and that this thing is like the concept in accordance with which God creates it. We might think of these divine concepts as the extrinsic, formal causes of particulars in the world. Particulars come into being as a result of the act of creation, but they are the kind of thing that they are because of the relation in which they stand to divine concepts. That is, particulars have their properties in virtue of the relation in which they stand to the divine concepts. Therefore, the relation of a particular to its extrinsic formal cause is a sort of ontological dependence. It seems reasonable to think that, if it is part of the essence of some thing, x, that it is what it is because of its relationship to y, and not vice versa, then x ontologically depends on y. For example, E. J. Lowe defines two kinds of ontological dependence, one weaker and one stronger as follows (Lowe (2003), 194):

1. (ND1) x is rigidly existentially necessarily dependent on y ↔ _df. Nec (x exists only if y exists).

2. (ED1) x is rigidly existentially essentially dependent on y ↔ _df. E_x (x exists only if y exists).Footnote ⁷

If we take a particular person, x, the doctrine of creation entails that, necessarily, x only exists if God does. Furthermore, the doctrine of omniscience entails that it is part of the essence of a particular human that they exist only if God's concept of humanity exists. The same could be said about any particular created thing and its properties. Within Lowe's framework, then, we can say that particulars rigidly, existentially, necessarily depend on God (ND1)Footnote ⁸ and that they rigidly existentially essentially depend on divine concepts (ED1). Because essential dependence entails necessary dependence, it follows that particulars ND1-depend on the divine concepts as well. If ED1 is transitive,Footnote ⁹ and divine concepts are dependent on God in the same way, as plausibly they are,Footnote ¹⁰ then it follows that particulars are also ED1-dependent on God. It is important to note here that this is not intended to endorse Lowe's particular way of cashing out ontological dependence. Rather, the point is that for any notion of ontological dependence that one formulates, it should turn out that created particulars ontologically depend on their respective divine concepts, because, according to TCR, the creation relation entails both causal and ontological dependence.

I indicated in the introduction that I follow D. M. Armstrong, Michael Loux, and others in taking the problem of universals to be a problem of explaining the character of particulars – of explaining what grounds them being the kind of thing that they are and the properties that they have.Footnote ¹¹ As such, it requires a theory of properties as its solution. With a theory of properties in place, one is then situated to explain all of the phenomena typically associated with the problem of universals: the truth of subject–predicate discourse, abstract reference, and property agreement. I have argued elsewhere (Panchuk (Reference Panchuk2016) ), and briefly summarized here, how divine concepts can do that metaphysical work for which traditional Platonists posit abstract property universals. If I want to know why the dog is brown, one sort of explanation that can be given is because the dog exemplifies the property being brown. We can call this the ontological ground of character.Footnote ¹² But what it means for the dog to exemplify the property being brown, according to my formulation of TCR, is that it bears a resemblance relation to the divine concept, being brown.

The simplicity of divine character

If one accepts that divine concepts do the work of explaining the character of particulars and that they ground the truth of subject–predicate discourse, then if multiple predicates can be applied to a single particular, it seems that this particular must bear resemblance relations to multiple divine concepts and, therefore, exemplify multiple properties. What then are we to say about God? The fact that we speak of God using a variety of distinct predicates provides at least a prima facie reason to think that God has some sort of positive, complex character. According to classical theism, God is omniscient, omnipotent, omnibenevolent, etc. If this sort of predication is appropriate, then we have some (but not conclusive) reason to think that God exemplifies various properties. Indeed, this is the conclusion that Alvin Plantinga (Reference Plantinga1980) draws. He argues that God must exemplify properties because, according to the law of the excluded middle, for any predicable, we must predicate either it, or its negation, of God. Either God exemplifies the property being holy, or God exemplifies the property not being holy. Footnote ¹³ God exemplifies the property being cruel, or God exemplifies the property not being cruel. Thus, claims Plantinga, property exemplification is unavoidable. But if we follow Plantinga in accepting that the principle of character grounding applies to God, serious difficulties loom for the DDS. For, as Plantinga points out, if both the principle of character grounding and the DDS are true, then God is identical to a property. The DDS entails that God is identical to the divine essence and that God has no properties apart from that essence. Therefore, God is identical to divine love, and to divine power, and to divine knowledge. But, if God is identical to God's love, and God's love just is the property being loving, then God is identical to a property. Because Plantinga thinks that properties are abstract objects, he concludes that the DDS entails that God is an abstract object: ‘But no abstract object could have created the world’, because abstract objects are causally inert (ibid., 47). So it seems that we must reject the DDS altogether.

One possible solution to this worry is to accept that God is identical to a property, but deny that this claim is problematic. William Mann has taken this approach, arguing that all people are identical to a special kind of particularized property-instance he calls a rich property, consisting of all of the time-indexed property-instances that they ever exemplify (Mann (Reference Mann1982), 466ff.; see Morris (Reference Morris1985) for a response). If this is right, God can be both identical to a property and the personal creator of the universe. It is doubtful, however, that such a response is open to the Theistic Conceptual Realist, because it is likely to result in a vicious form of bootstrapping, as others have argued in detail elsewhere (see Bergmann & Brower (Reference Bergmann, Brower and Zimmerman2006); Gould (Reference Gould2011b); Panchuk (Reference Panchuk2016); Gould & Davis (Reference Gould and Davis2017) ). The brief explanation of that argument is that if God exemplifies properties within the TCR framework, God's nature ends up being both causally prior and causally posterior to God's own nature, because properties are identical to divine concepts.

Michael Bergmann and Jeffery Brower endorse an approach that is more promising for proponents of TCR (although they themselves reject platonic theories such as TCR and Theistic Activism). They deny that one should accounted for divine predication in terms of exemplifiables, and instead adopt a combination of substance ontology and a truthmaker theory of predication. In so doing, they deny that the truth of our discourse about God's character requires that God exemplify properties. According to their account of truthmaker theory, ‘If an entity E is a truthmaker for a predication P, then “E exists” entails the truth expressed by P’ (Bergmann & Brower (Reference Bergmann, Brower and Zimmerman2006), 376).

There are theoretical reasons, which they acknowledge, for taking this only as a partial analysis, but the general suggestion should be clear. For every essential predication (predications of properties essential to a thing's nature) of God, P, God's existence necessitates the truth of P, because P is true in every possible world where God exists. The simple divine substance is the obvious candidate for being the truthmaker for all divine predications. If that is right, the divine substance is the truthmaker for the proposition ‘God is love’, and for the proposition ‘God is omniscient’. We need not appeal to properties at all.

There is something right about this approach. First, it seems that the DDS fits more naturally within the sort of substance ontology that was standard in the mediaeval world, but I will not argue for that view here. Second, Bergmann and Brower are right to reject exemplifiables as the only candidates for explaining the truth of our discourse about God. Other theories of the truth of subject-predicate discourse are available, including truthmaker theory. However, it is critical to remember that the apparent tension between the DDS and TCR arose not primarily from issues related to truth, but out of considerations of ontological structure (Leftow (Reference Leftow2015) ). As I have formulated it, TCR is not primarily a thesis about predication (though it has important implications for predication); it is a theory about what metaphysically grounds a specific phenomenon. The phenomenon under present consideration is that particulars apparently possess distinct and complex characters. We might wonder what makes the proposition ‘Socrates is human’ true, but we can also ignore this question and wonder what grounds the fact that Socrates is human. In what does the state of affairs of Socrates being human consist? Depending on one's ontological commitments, the answers will vary: Socrates being an instance of the kind-universal humanity; Socrates being constituted by a bundle of tropes that necessarily has his humanity _s trope as a constituent; or Socrates’ participating in the divine concept of humanity. The same is true in the case of God. One might grant that the divine substance is the truthmaker for all essential divine predications while still wondering how it can be that a simple being like God has an, apparently, rich character. What explains the fact that a single, ontologically simple substance can be the truthmaker for diverse predications? This question is not necessarily answered by any particular theory of predication. Thus, if we wish to embrace the spirit of Bergmann and Brower's intuitions, and deny that God exemplifies a multiplicity of properties, we must still say something about what explains the fact that the divine character apparently has characteristics which created particulars could only possess by being ontologically complex. I will suggest that the distinction between the created and the uncreated, the finite and the infinite, can explain why the character of finite beings must be grounded in the properties they exemplify while the character of an infinite being need not be.

I propose that two things are true of property exemplification according to TCR: (1) property exemplification arises from the creation relation, and (2) property exemplification entails finitude. Each gives us reason to deny that God exemplifies properties. The first claim follows straightforwardly from our discussion in the first section. I have argued that particulars ontologically depend on divine concepts because they are the result of God's intentions to create in accordance with those concepts. But, God does not depend in any way on the intentions of another.

To understand the justification for (2), it is helpful to think about what follows from the fact that any particular exemplifies a property. To exemplify a property is to exist in a certain way (Armstrong (Reference Armstrong1989), 96–97). It is to exist in this way, rather than in that way. In other words, to exemplify a property is to be delimited and finite.Footnote ¹⁴ For this wine to be red, it must not be (in the same way and at the same time) any of the other colours of the rainbow. To be a dog is to fail to be a cat and all of the other animals at the zoo. Furthermore, being an instance of a kind-universal or exemplifying one property may limit the range of other properties that a particular can exemplify. Being an instance of the universal-kind dog means that the particular cannot exemplify the property being prime.Footnote ¹⁵ Properties impose a specific set of limits on particulars. If we are right to think that property exemplification follows from the act of creation, then being finite (on at least some axis or another) would follow necessarily from being created.Footnote ¹⁶ Only created particulars depend for their existence on the intentions of another, and only created particulars are finite expressions of being.

If properties explain the limits on the being of particulars, and God is absolutely infinite, then it follows that God does not exemplify any properties. However, we do not want to be forced to say that because God is property-less God is also character-less. I suggest that if properties are delimited ways of being, an infinite being, far from having a more impoverished character than finite things, has a richer one – an infinitely richer one. The infinite encompasses all that finite beings are, because, not having any limitations on being, it is more than them, without being (exactly) any particular thing that they are. It is hard to conceptualize this along all axes simultaneously, but if we use a comparison with properties along one axis it may be helpful. A line that is infinitely long lacks the properties ‘being (exactly) one inch long’ and ‘being (exactly) two inches long’ not because it lacks these inches of length but because its infinite length encompasses and surpasses them both, without possessing either particular length. I claim that something similar is true when comparing God to finite things, only it holds along all possible axes. The divine being is simple, but infinite, and thus ‘rich’.

There is precedent for this claim in the history of philosophy, stretching back through the neo-platonists at least back to Plotinus and perhaps even to Parmenides. But given our present concern with the relationship between God and universals, Nicolas of Cusa's discussion of the nature of God as infinite universal is particularly suggestive. Cusanus is committed to the via negativa. He argues that we cannot properly or truthfully speak of what or who God is; we may only affirm what God is not. Foremost among those properties that God does not exemplify is finitude. Cusanus thinks that from the fact that God is not finite we can infer that God is the ‘absolute universal’. On this view, universals exist as distinct universals only in particular things, but they exist in God as identical to God. To explain how it is possible for ‘all possible things’ to be in God actually, while still be identical to God, Cusanus turns to the mathematics of infinity for a metaphor that he claims transcends our power of imagination, but is ‘graspable by the intellect’ (Hopkins (Reference Hopkins1985), 22f.). He reasons that an infinite circle is a circle, the diameter of which is infinitely long. Because any given arc of a circle is more or less curved in proportion to the circle's diameter, an infinite diameter requires a minimally curved line that is infinitely long – i.e. an infinite, straight line (Hopkins (Reference Hopkins1985), 21). Thus, an infinite line and an infinite circle are one and the same. Cusanus expresses this relationship by saying that the infinite circle is ‘enfolded’ in the infinite line. The example is meant to prove that, at the infinite limit, things lose their differentiation (ibid., 24–25). Thus, something infinite simpliciter (rather than only in length, or only in size, or only in goodness) would lack any particular qualities while encompassing all of them. Even if Cusanus's understanding of geometry and infinity leaves something to be desired by contemporary standards, I believe he has metaphorically communicated something about the nature of God that may not be expressible via literal discourse. Perhaps we grasp something in his metaphors that, while not literally true about lines and circles, gets us closer to understanding the relationship between the infinite divine nature and finite created particulars. What we are supposed to grasp, along with the rest of the Neoplatonic tradition, is that the infinite surpasses the finite by encompassing it. God is wholly other, while being more, not less, than any created particular.

Given the foregoing discussion of the ontological character of the divine substance, we are now in a position to ask questions about discourse concerning God. While giving a thorough account of predication is not within the scope of this article, I think at least two accounts are compatible with the picture I have presented. If one is inclined to truthmaker theory, then the infinite substance can be the truthmaker for all essential predications of God as Bergmann and Brower suggest, only now we have some explanation of how and why it is the truthmaker for diverse predications, despite being simple. If we accept this view, some questions will still remain. For one, if God encompasses everything, why are some predications true of God while others are false? I will return to this question in a moment. But on my account, unlike Bergmann and Brower's, if one rejects truthmaking one may still be able to give a satisfactory explanation of divine predication. If properties always indicate limits on being, one might draw on the venerable tradition of the via negativa and argue that divine predications refer not to the properties that God exemplifies – since God exemplifies none – but to the absence in God of various limits that are present in created particulars. That is, when we say that God is omniscient, we claim that God is not limited with respect to knowledge. When we say that God is omnipotent, we claim that God is not limited with respect to power. There are two ways that one might interpret these claims.Footnote ¹⁷ On the first, God is not limited with respect to power because God does not possess power at all. To possess power would be to possess a limit. Few theists would be satisfied with such radically revisionary entailments. On the second, God is not limited with respect to power, not because God lacks power, because God surpasses the limits that the property ‘having power’ indicates in finite creatures, having infinitely more. When I say God is not limited with respect to power or to knowledge, I mean ‘God does not lack (any amount of) power’ and ‘God does not lack (any amount of) knowledge’. In other words, God does not lack what a finite being possesses when they exemplify the property ‘being powerful’ (when they instantiate limits with respect to this mode of being), in the same way that the infinite line does not lack what the inch-long line segment has when it exemplifies the property ‘being (exactly) one inch long’. It is just that at the infinite limit, power is not distinct from knowledge, is not distinct from God. This acknowledges the intuitions behind negative theology and perhaps Thomistic analogical truth (a theory beyond the scope of this article) while affirming as true the positive claims that many theists believe to be essential to theology.

However, one might think that the ‘with respect to’ locutions used above reveal a subtle (or perhaps not so subtle) acknowledgement of divine complexity in my theory. The absence of limits with respect to this, rather than that, might suggest that one can distinguish various aspects of the divine, which might be limited, but are not. The divine has knowledge, which might be limited, but isn't, and power that might be limited, but isn't. This would be the wrong interpretation of these locutions. In the context of the picture I have proposed, the phrase ‘with respect to’ refers not to a distinction between aspects of the divine nature, but to the complexity that a limit would constitute in the divine nature if it were there.Footnote ¹⁸ That is, once we conceptualize a property, with it we grasp the specific kind of limit that property would constitute. The limit would be a limit with respect to something only because the limit itself would introduce complexity. Being familiar with all of the ways in which finite creatures are limited, we can (relatively) easily conceptualize such limits that are absent within the divine nature. Our subject-predicate discourse, then, is the denial of these limits – and the properties that entail them – in God.

One might object to this proposal on the grounds that while the ‘omnis’ fit neatly within it, other properties do not. For example, it isn't clear how ‘being the creator of Abraham’, or ‘being loving’ could be explained along these lines. We cannot, after all, interpret ‘being the creator of Abraham’ as the absence of a limit on God's creation, for there certainly are such limits, even if they are self-imposed. God chose to create Abraham, kittens, and church mice rather than Frodo, centaurs, and house elves. God does not create everything that God could possibly create, which seems to suggest the very limits that I deny above.Footnote ¹⁹ This does not pose a particularly serious problem for someone who, like me, is partial to a sparse theory of properties, since such a person already assumes that there is no one-to-one correspondence between the predicates in natural language and the properties that things exemplify. Predicates like ‘being the creator of Abraham’ refer to relational properties, which are extrinsic, rather than intrinsic. It is the property being a creator or being capable of creating or, even better, being omnipotent that aptly characterizes God's essence. Having the capacity to create follows from the absence of a limit on divine power. Similarly, ‘God is loving’ or ‘God is love’ means something like ‘In God there is no lack or limit to love.’ That is, there is an absence in God of the finite limits that make human love distorted and deficient.Footnote ²⁰

Finally, if God encompasses all universals, we must explain why some predications express truths about God, while others do not. Traditional theism says that God is loving, but that God is neither evil, nor ugly, nor four-legged. With respect to the first two, I suggest holding to the Augustinian intuition that negative properties are privations. If a property is a particular, limited way of being, a negative property refers to the absence of that particular way of being in that entity – one that is appropriate to its substance-kind. So while the via negativa suggests that God lacks limits by surpassing them, these privations are cases of lacking by deficiency or absence. Something that is evil lacks goodness. It fails to resemble the divine being in the delimited way that it should and is less like it. With respect to the second, I mentioned above that exemplifying one property, or being an instance of a particular kind, may circumscribe the other properties that the particular can simultaneously exemplify. Being a dog precludes being a prime number. Thus, a great number of properties will not make sense to predicate of God because those properties are dependent on the entity already existing in a particular, delimited way.Footnote ²¹

In this section, I have explained a prima facie reason one might have for thinking that my Theistic Conceptual Realist account of properties conflicts with the DDS. I then argued that some of the defences of the DDS conflate accounts of truth with accounts of ontological character, and demonstrated that the latter is fundamental. I suggested that a simple but infinite being has the rich character we attribute to God because the infinite is more, not less, than particular finite things. And finally, I argue that our discourse about God can either be accounted for via truthmaker theory or by a theory of reference to absences of limitations in the divine being.

The simplicity of divine thought

The second task in reconciling TCR and the DDS is to demonstrate that the latter is compatible with the claim that divine concepts ground all of the numerous properties that particulars do, or could, exemplify. No doubt, the explanation of the relationship between the finite and the infinite in the previous sections suggests a promising starting point for such a demonstration, but as it stands it is insufficient. As I pointed out in the introduction, on a straightforward interpretation TCR suggests that multiple, discrete concepts exist in God. If my argument in the previous section is successful, we can think of God being simple while still having a ‘rich’ character without accepting that God literately exemplifies properties, but it is difficult to see how God could entertain distinct concepts, which serve as extrinsic formal causes, without introducing complexity into the divine mind. The doctrine of simplicity requires not only that there be no distinction between God and God's properties, but that there be no distinction between God and God's thoughts. A promising account of a related issue – how God's knowledge of contingent things, which differs across possible worlds, is compatible with simplicity – suggests that God's knowledge of contingent things is extrinsic to God's nature, such that there is no intrinsic difference in God across possible worlds (Pruss (Reference Pruss and Kvanvig2008); Grant (Reference Grant2012) ). But the multiplicity of concepts for which I have argued is both necessary and intrinsic. They are necessary because God not only has concepts for those properties that God chooses to instantiate in created particulars, but also all those that God could instantiate in created particulars, whether or not God does so. Thus, they are constant across possible worlds. They are intrinsic because God's knowledge of creation, qua exemplar cause, is causally prior to the act of creating and metaphysically prior to the instantiations of those properties; therefore, they are not the result of something external to God.Footnote ²²

Thomas Aquinas addresses this issue in both the Summa Contra Gentiles (I, c. 51–2) and the Summa Theologiae (I.15.2). In Summa Theologiae he argues that the divine ideas must be many, because God does not create one thing, which is causally responsible for the rest of the universe. Rather God creates the order of the entire universe. To have an idea of a whole requires that one have the ideas of all of the parts. Therefore, God must have multiple ideas if God creates all of the parts of the universe. Aquinas explains, however, that this is not at odds with the doctrine of simplicity because God does not have multiple images in God's mind, which God forms on the basis of external things. That is, the ideas in the mind of God are not intelligible species by which God knows particular things (as an intelligible species actualizing God's intellect), but are that which is understood. The ideas that serve as exemplar causes for finite particulars are identical to God's divine self-knowledge. God can know Godself in two ways – as God is in Godself and as God can be participated in by finite creatures. Each species participates in the divine essence in its own unique way and measure. Thus, Thomas claims that the multiplicity is only from the perspective of the creation, while the principle or mode of knowledge is the simple divine nature, so knowledge of multiple things need not compromise God's simplicity.

Placed within the context of our earlier discussion, the ways in which finite particulars imitate God arise from the presence in them of boundaries or limits on being that are absent in the divine nature. Any finite imitation of the infinite would have to be in some respect or another. In knowing Godself, God knows not only the divine nature, but also the many ways in which limited things can be like that nature. So, while the object of divine knowledge is multiple, the mode by which God knows it – i.e. through the simple divine nature – is simple. Or at least so Thomas suggests (Doolan (Reference Doolan2008), 83–122). One might wonder if the inference is valid. Even if the mode of knowledge is simple, what God knows in virtue of knowing that simple nature is decidedly complex. In virtue of knowing Godself, God must also know that P¹ – a is like God in x way – and P² – b is like God in y way – and so forth for every particular, every species, every genera, every property, and every relation. If these propositions are distinct, then we have a prima facie reason to think that God's concepts and thoughts are truly distinct in God. The second problem, then, for reconciling TCR and the DDS arises from conceiving of God's knowledge as propositional. Propositional divine knowledge is problematic because God knows a multiplicity of propositions and, if propositions turn out to be complex, knowing each proposition itself involves further complexity. The pressing question, then, is whether or not it is reasonable to deny that God's knowledge is propositional.

It should be obvious that non-propositional knowledge, in principle, exists. Humans possess non-propositional knowledge in spades. We know how to do things. We know what things are like. Consider Frank Jackson's famous thought experiment of Mary in the black and white room (Jackson (Reference Jackson1986) ). Prior to leaving her black and white existence, Mary is quite knowledgeable about the science of colour vision. We can even imagine that she believes every true proposition relative to colour. Nonetheless, when she leaves her room and experiences colour for the first time, it is reasonable to think that she gains knowledge. She gains knowledge of what seeing colour is like, which was not captured by any of the propositions she previously knew to be true. One might think (although it certainly isn't uncontroversial) that knowledge of what something is like is a distinct category of knowledge that is not reducible to its propositional cousins. Perhaps God's knowledge is more like our knowledge of what things are like than our knowledge that P.

Unfortunately, the mere existence of non-propositional knowledge does not obviously help our case. In order to defend the DDS, we need to demonstrate that it is possible that God non-propositionally knows the content of what we know when we know that P. But one might think that knowledge of what something is like is fundamentally different from the kind of propositional knowledge that concerns us now. Our use of the term ‘knowledge’ in both contexts may be merely polysemous.Footnote ²³ We have propositional knowledge, and we have phenomenal knowledge, and we have how-to knowledge, but we cannot know via one mode what we know via the other. After all, I can know that This rose is red and I can know what experiencing this rose's redness is like, but I can't know that ‘This rose is red’ in a non-propositional way. At least, so goes one popular view of perception and experience. Many philosophers have argued that experience is fundamentally conceptual and propositional. At the very least, it is controversial to think that we regularly experience anything without experiencing it as something.

However, there is another strain of thought that places a brute givenness of experience at the foundation of empirical knowledge. Consider again the experience of a red rose. We might think the content of the experience fundamentally includes the concepts ‘rose’ and ‘redness’ and a certain phenomenal content – the what-it-is-like-ness of the experience – as distinct constituents (e.g., Brewer (Reference Brewer, Steup and Sosa2005); Byrne (Reference Byrne, Steup and Sosa2005) ). Both aspects, the conceptual and the phenomenal, are equally fundamental on this picture. Now consider a second picture. When I experience the rose, I experience certain perceptual sensations that constitute the what-it-is-like-ness of the experience. Then, on the basis of that phenomenal content, I formulate concepts, and then entertain propositions, as a way of giving an account of my experience. On this picture, the non-propositional experience is the raw, unified ‘data’ from which concepts are extracted. Linda Zagzebski suggests a similar picture, claiming that:

If she is right, every experience is non-propositional in a fundamental way. But we don't need anything quite this strong to make our point. Even if it is doubtful that adult humans who already possess the relevant concepts can consciously experience the world in this brute, non-conceptual way, we still have a coherent possible account of experience that places the non-propositional at the foundation. This makes it more plausible to think of God as having non-propositional knowledge that is more fundamental and more complete than ours. Indeed, several philosophers have suggested a picture similar to Zagzebski's for understanding divine knowledge. Aquinas claims that God knows by a simple intuition all that can be known about a thing, rather than by ‘composing and dividing’ as humans do (Summa Contra Gentiles, I. 38; Summa Theologiae, I.14.16). Thomas Sullivan similarly claims that God ‘attains intimate epistemic acquaintance’ with all particulars by a single intellectual act that does not rely on ‘impoverished abstraction’ (Sullivan (Reference Sullivan1991), 30). And, William Alston suggests that F. H. Bradley's account of experience that places at its base a state of ‘pure immediacy’ where there is no distinction between subject and object provides a helpful framework for thinking about divine cognition (Alston (Reference Alston1986), 290). Human beings cannot grasp a concrete whole in its fullness. Instead ‘we recognize certain abstract features . . . which we proceed to formulate in distinct propositions’ (ibid., 291). Furthermore, we need propositions in order to extend our knowledge by inference (ibid.). God suffers from neither of these cognitive limitations, and therefore has no need to ‘divide’ up divine knowledge into concept-sized bits.

Although it may be open to someone who endorses this mode of non-propositional, non-conceptual knowing to follow the Thomistic view that God knows particulars through the divine essence (I will not here evaluate the success of this possibility), I think the proponent of TCR need not do so. For the purposes of defending the compatibility of TCR with the DDS, it need only be the case that God knows properties (finite modes of being like the infinite being) by knowing the divine essence, because this knowledge must be causally and metaphysically prior to creation. If God directly, non-propositionally, non-conceptually, and eternally intuits all that is, there doesn't seem to be any, in principle, reason to think that God cannot do so with the external world and the divine essence.

However, the above perspective is not obviously compatible with Theistic Conceptual Realism as I have described it. In the first section I referred to God's concepts of created particulars. The relation in which particulars stand to God's concepts is the foundation of TCR. Here I claim that the sort of ‘composing and dividing’ that necessitates the existence of concepts is below the divine mode of knowing. At this point it may be helpful to recall the purpose for which proponents of TCR appeal to divine concepts. On TCR, divine concepts are supposed to do the work that abstract universals do in theories of platonic realism about property universals. That is, they metaphysically ground the properties that particulars exemplify, and in so doing explain the truth of subject-predicate discourse, abstract reference, and property agreement. Can the divine mind fill these ontological and theoretical roles without containing distinct concepts?

One obstacle to answering this question is that we, the metaphysicians, cannot entertain or answer the question without entertaining and appealing to distinct concepts. That is, we must reason about a mode of knowing to which we have no access. From our perspective, the divine mind will necessarily appear to ‘contain’ distinct concepts, but this may say more about our mode of knowing that it does about God's. As such, any claim that I make will be necessarily tentative and suggestive, perhaps even more metaphorical than metaphysicians would prefer. But if the infinitely simple divine nature is necessarily richer than the finite particulars that participate in it, and if God can know in one simple act of direct intuition both the divine essence and all to which the power of that divine essence extends, and if finite particulars can exist by being like God in finite delimited ways, such that they have their properties in virtue of that imitation, then it seems like the divine essence could perform the functions for which abstract universals are usually posited. Particulars exemplify properties in virtue of the delimited resemblance relations they bear to the infinite and simple divine essence. Multiple particulars exemplify the same property (attribute agreement) by bearing the same resemblance relations to the divine being. Those very same ‘aspects’ of the divine being ground the truth of subject-predicate discourse. And cases of abstract reference (at least in cases of fundamental abstractions) refer to those ‘aspects’ of the divine mind. Returning to Aquinas's and Cusanus's arguments, particulars are like the aspects of the simple infinite nature that we can distinguish by imagining the limits that fail to exist in God. God, on the other hand, need not distinguish these limits using discrete concepts as we do to know how the divine nature is imitable by finite particulars.

If this speculation is close to the truth, we can say the following things about non-propositional divine knowledge: (1) it is distinct from, but continuous with, the propositional knowledge that humans have; (2) it is richer than propositional knowledge; (3) it is more direct than propositional knowledge. As odd as the idea of non-propositional knowledge of what humans can only know propositionally may initially sound, all three of these features – difference, richness, and directness – are what a traditional theist should expect of divine knowledge. Even if we cannot prove by any empirical mechanism that God in fact has such knowledge, we have at least shown that a theory that proposes such knowledge is not self-referentially incoherent, incomprehensible, or even completely discontinuous with some theories of human knowledge.