Introductory remarks

I am very pleased and honoured to be writing for Richard's celebratory collection. I first studied philosophy of religion in 1969 with Richard's predecessor, Basil Mitchell, for a BPhil that I never completed, but, though I have taught it at all levels since then, I have published very little in it and nothing before 1994, so I cannot claim to be a serious figure in the subject. The importance of Richard's contribution to the development of philosophy of religion as mainstream within analytical philosophy, especially in the US, is, of course, beyond doubt and needs no elaboration from me. So, I hope, is the importance of the topic on which I am now to talk, and in the debate on which Richard has been very prominent.

Statements of the two views

Richard Swinburne states his own position thus.

Brian Davies contrasts this view and Aquinas's.

A positive account of the traditional picture is given by Vallicella.

This article is an attempt to understand what the difference is between these accounts and in what ways the differences may be important – or not. There is a sense in which the two parties are not in a simple disagreement. The properties that Swinburne and other ‘perfect person’ theists attribute to God are also attributed to Him by classical theists, and essentially so. All theists agree that God is omnipotent, omniscient, and perfectly good. The disagreement arises, I think, because the classicists think that these properties can belong to a divine being – can have, as one might say, divine proportions – only if the divine being has a very special real essence, whereas the ‘perfect person’ theorists think that attributing such properties as omniscience, omnipotence, and absolute goodness to God is enough to tell you what His nature is. No more mysterious ‘inner rationale’ or ‘real essence’ is needed to rationalize the absolute, unlimited, or infinite nature of these properties.

Why is it thought necessary, by classical theists, to go beyond the agreed properties, and, if one does go beyond them, does it make any difference to how one understands these properties as they manifest themselves in God?

Dolezal explains the need for the classical conception thus.

Dolezal explains the rationale of the modern rejection of the traditional picture in the following way.

And he says the following about the unfortunate consequences of the modern developments:

Sorting out the issues

I apologize for starting with such long quotations from the traditionalists, but I hope they help to make that position and its rationale clearer; Swinburne's view – or Plantinga's – is much more straightforward and needs fewer words. This gives them a head-start in the debate. And not all the reasons given for the traditional position are entirely clear. For example, denying ‘ontological univocity’ – often expressed as the slogan that God is not a ‘being among beings’ – may not seem to be entirely clear. One possible sense is that, if you list the things that you take to exist, then God will not be on that list. Some apophaticists – and some Philipsians – may hold this, but certainly not anyone in a normal orthodox tradition. Aquinas thinks that we can prove deus est even though we cannot know His essence. Another interpretation is that God is very deeply different from creatures, but this is both innocuous and uninformative, unless spelled out. Swinburne's God might be intelligible in the same terms as human persons are, but His radical perfections make Him very different indeed from the rest of us and even from angels. So on its own ‘ontological univocity’ does not seem to say anything specific.

There seem to me to be at least the three following genuine issues.

1. (1) Is God eternal or sempiternal? The traditional view says the former; the modern, the latter.

2. (2) Can we know the Divine essence? The traditional view is that we cannot, but, if the modern view is correct, there seems to be no trouble in the idea that we can. There is the associated issue of what the traditionalists mean by ‘knowing the essence of God’ – what would it involve?

3. (3) Is God simple and, if so, what is meant by this expression? The traditionalist says that God is simple, and this in a sense that implies both that His properties are not distinct from Him and that they are not distinct from each other. Furthermore, His essence is identical to His existence. Moderns may accept His simplicity, in some sense (Swinburne does), but not in the ways that I have just glossed it.

The issue over eternity is primary, in the sense that, if God is in time, then He will not be in all respects immutable, as the classical theory claims He is, and this puts constraints on in what sense He can be simple. Sempiternity fits much more easily with the ‘perfect person’ conception.

The problem of eternity: (i) the coherence of timelessness

Swinburne is together with many moderns in denying God's timelessness. Unfortunately, in The Christian God (1994) he seems to rely, for his account of the traditional view, on the account by Stump and Kretzmann (Reference Stump and Kretzmann1981), the inadequacy of which had already been demonstrated by Paul Helm (Reference Helm1988), who proceeded to show what the eternalist should say. Swinburne does not reply to Helm in Swinburne (Reference Swinburne1994).

The normal accounts of timelessness talk of God being simultaneous with or present to all times at once or in an instant or moment. Stump and Kretzmann do not avoid this kind of talk and it is easy for Swinburne to show that this leads to contradictions. Helm's response is to avoid all use of temporal talk when describing God and His relation to the created world. He is not ‘simultaneous with’ or temporally ‘present to’ anything, nor in any ‘instant’, or ‘moment’.

Swinburne might believe that this is impossible because of his views on the relationship between causation and time.

He cites his principle called the ‘causal theory of time’ as entailing this:

Swinburne then defends this by arguing that it is not possible to influence the past.

If one takes the quoted words as the complete truth, then, if God can causally affect what happens at time t, then t must be future for him, so he must be in time. But, as far as I can tell, all that Swinburne has actually shown is that, if cause and effect are both in time, the effect must be future to the cause. This is not particularly controversial, but, more importantly, it is not relevant to the issue of divine timelessness. It does not prove that a cause cannot run from timelessness to time – nor that a timeless being cannot cognize events in time. It may well show that nothing that could be called a mechanism could realize either of these things, but God's causing and knowing do not require this on any account.

In the latest edition of The Coherence of Theism (2016), Swinburne responds to arguments of Helm's type as follows.

Swinburne clearly asserts that God's knowledge of human action involves the action causing the knowledge in God. I think that a classical theist would deny that God is passive in this relationship; it is by His action that He knows, not by something's having an effect on Him. If it were a matter of the action affecting God, then it is difficult to see how an event in time could affect a timeless being, but God's directly cognizing the action does not seem to have this consequence. It is no doubt true that, in some obvious sense, God knows about the action because the action occurred, and that, in a sense, therefore, God's knowledge depends on the fact that there was the action in question, but it does not follow from this that this is a causal ‘because’ or a causal dependence.

Helm's approach to the problem seems to me, therefore, not to have been refuted.

The problem of eternity: (ii) if God were in time, would he fail to be absolutely powerful?

The proponents of divine timelessness argue that if God were in time He would be its captive and not sovereign over it. Is this true?

There seem to me to be at least the following ways of understanding God's presence in time, if one is not an eternalist.

1. (1) Time is a necessary existent, alongside God, like Platonic forms are or would be.

2. (2) God is in time because He changes certain accidental properties – for example, His beliefs – as He comes to know what is currently happening in time.

3. (3) Time exists when there are objects that change and God is then ‘caught up’ in that time. But either (a) there is no time when there are no changing objects, or (b) there is undifferentiated time when there are no such objects.

The first might be the Newtonian absolutist position. It does not entail that God is actually in this time, but it would give time a status independent of God. I shall not be discussing it.

The second relates to time through the doctrine of immutability, but I think that if we can make sense of divine timelessness, then there is no great problem in allowing Him to know timelessly of the changing world. One model would be that of the specious present, which involves no differentiation in the temporal structure of the knower, only in the object of knowledge. This would give support to Aquinas's claim (if I remember rightly) that God's knowledge of temporal changes involves only change in that to which He is related, not in Himself.

The third, with (b), is Swinburne's position. I agree with the common clause in 3 – differentiated time needs change, but think that (a) is correct. Swinburne's reason for accepting (b) is that, if metric time begins with the appearance of changing substances, then there is a ‘before’ this occurs, so there is time before, though undifferentiated.

I think there are two things to say about this. First, that even if this were true, it would not follow that God was subject to the time. It would, however, as with the ‘Newtonian’ option, seem to make this time independent of God. Second, I'm not convinced that undifferentiated time is needed to give intuitive, if not literal, sense to ‘before time started’. Would it not do that there is always the possibility that God might create a temporal system, irrespective of whether there is actually one, and He did in fact create a series of temporally ordered events running from e¹ to eⁿ?

So would God's sempiternity make God not absolute? Not if it only applied whilst there was a world which He had created and need not have created, but if time were necessary, like Forms, even in undifferentiated form, then His absoluteness would be undermined.

Knowing God's essence

It is important to Aquinas that we cannot give a real definition of God, only a nominal definition. This importance stems from the fact that Aquinas believes that God's essence is identical to His existence, and if we could give a real definition it would be notum quoad nos that existence was part of it and that, therefore, the ontological argument would work, and he both thinks that it does not and that it is important to his general strategy that it does not.Footnote ¹

Not having a real definition is equivalent to not knowing the real essence. But there is something of a paradox here. Usually, in a modern context, not knowing the essence of something is because it is some hidden, complex structure, and advanced science is needed to unearth it. But God is simple, so what is it that we do not know? Paradoxically, there seems to be no scope for ignorance.

Furthermore, there is a contrast between real essences of creatures as conceived of by the Aristotelian tradition and as they might be understood nowadays. For Aristotle, man was a rational animal, and this was his real definition. Basic matter consisted of air, fire, earth, and water, and these could be really defined in terms of the hot, the dry, the wet, and the cold. So real essence of kinds was available by some straightforward philosophical analysis.Footnote ² By contrast, modern real essences depend on arcane science. On Putnam's analysis, water is H₂O, but what is to guarantee that all hydrogen is the same, etc. The assumption that kinds go all the way to the bottom – if there is a bottom – is built into Putnam's theory, but the problematic nature of this assumption is ignored. It is even problematic what ‘the real essence’ of the quantum field, which, in some sense, constitutes the whole of matter, could conceivably be. Could it be that there is a sense in which God's real essence – as expressed in Swinburne's conception – is more transparent than the real essence of matter?

In order to answer this question, it helps to see what Aquinas thinks is involved in knowing, or ‘comprehending’ an essence.

Aquinas argues that, by the light of divine grace, the human intellect can see God and His essence, but it cannot comprehend him.

The first striking thing about this passage is the very high standard required for knowledge. It is a clear and well understood scientific demonstration based on incontrovertible first principles. The fact that we can see God, but only imperfectly, as a result of ‘the light of glory’ suggests that the mode of such ‘knowledge’ as we have is experiential and more like perception than like a demonstration. Either way, one can say that knowledge proper requires that one have a ‘clear and distinct idea’ of the truth or phenomenon in question, and this we cannot have of God. However much ‘light of glory’ we are granted, we can see His nature only in a ‘blurred’ fashion, even if not ‘through a glass darkly’.

It seems that, for Aquinas, knowing an essence has two components. One is knowing that it exists actually – that is, is realized; the other is having as total an account of its nature as the nature of the thing allows. The beatific vision and the cosmological arguments both, in their different ways, give one knowledge of the existence, but neither gives one a thorough grasp on the Divine nature. It is still not wholly clear to me why the complete philosophical account does not provide this, on an intellectual level. What it cannot give us is a definitive experiential understanding of that nature. Maybe a modern expression of this limitation is that it is totally impossible for us to comprehend what it could be like to be God.

One way of coming to appreciate why this experiential comprehension of God is impossible for us is to reflect on the special way that God is related to His properties. This latter issue is Dolezal's second challenge. It was that the absoluteness of God is undermined if He is allowed to possess any positive properties in a straightforward way. To approach this argument one begins from the idea that God cannot have parts, in any sense. It is not controversial that he cannot have parts in the straightforward sense, because these parts might ‘come apart’ and He would be contingent upon these parts. But it is far from obvious that this rationale can be extended to what one might call ‘conceptual parts’, such as properties. The properties of an object cannot ‘fly apart’ in the same ways that ordinary physical parts can be dispersed. The traditional view, however, is that conceptual complexity means that the complex object cannot be the foundation of the objects or features that constitute its complexity. This is what leads Dolezal – and Plantinga – to say that the modern conception allows a parallel eternity to universals. I think an argument for this position can be reconstituted as follows.

1. (1) Assume that one is a realist, in some sense, about universals/properties.

2. (2) In so far as properties/universals are real, if God is the source of everything, He must be the source of them.

   Therefore

3. (3) If He instantiates properties (including essentially), He must be the source of those properties.

4. (4) If it were a primitive fact that He instantiated them essentially, He could not be their source, for without them He could not be what He is.

This line of argument about properties could be generalized, as follows.

1. (5) For any conceptual distinction held to be real in God the same form of argument would apply. For example, form and matter, potentiality and actuality or existence and essence would have to be independent of God.

Therefore

1. (6) If God is the absolute source of everything, then in so far as these things are real, God does not instantiate them.

Form and matter might be dismissed on the grounds that God has no matter and the other two might be thought to be eccentricities of the Aristotelian and scholastic tradition. Plantinga, however, is happy to accept that the Platonic realm exists alongside God.

The controversial step in the argument is (4). It rests on

1. (7) An entity cannot be the source of something if its own existence logically depends on it.

Someone might challenge this by asserting an in re form of realism.

1. (8) Properties/universals only exist in their instances.

2. (9) Instances of properties depend on the thing that instantiates them.

   Therefore

3. (10) God's properties depend on God.

This is the line that Swinburne takes: if one can avoid Platonism, there is no problem. Whether this is a reply to the original argument depends on whether the kind of dependence that a property-instance has on its subject is the right kind of dependence. Dolezal is not satisfied with this answer. He has two objections.

He quotes with approval Barry Miller:

The force of the ‘absolute self-sufficient’ is that He must be the source of properties also.

This is made clear in the second objection:

Vallicella cites these arguments as sufficient to dismiss the harmony theory, but, as they are expressed, they may not seem clear. Talk of the attributes being the reason for God, for example, seems strange and I'm sure Swinburne would not accept such a description of the situation, but I think the worry can be stated less oddly.

It might seem that although an object's properties depend on it, because, on the in re theory, a property cannot exist without its owner, neither can the owner exist without its essential properties. So it is the case that if there were not properties as such, in re or ante rem, there could not be God. It might seem that for this reason in re realism will not do the trick.

It is still possible to challenge the argument, along Plantinga-esque lines, by disputing whether such a radical degree of dependence of everything on God is really required by orthodox theism. Is the fact that God could not exist without his properties really a limitation?

One way of indirectly tackling this last question is to see whether adequate sense can be made of the attempt to make God transcendent of even His nature. If it can, then there is something to be said for the project. In fact, most of the arguments against the traditional position are directed at the thought that no sense can be made of it. That is the burden of Plantinga's arguments in ‘Does God have a Nature? (Reference Plantinga1980). He makes the obvious points that two properties – like knowledge and virtue – cannot be the same property, as such, and that it makes no sense to say that a non-abstract thing is the same as a property. In Robinson (Reference Robinson and Stone2008) I said that this was an overly literal interpretation of the traditional doctrine and suggested instead what I called the neo-platonic model of God, as follows.

I now have two reflections on this. First, I'm not sure that the contrast between the ways that predicates apply to God and to us is so absolute.

In the case of, for example, a red ball, the properties of redness and sphericality are quite distinguishable in the object itself, and this may be true of all normal physical properties. Even more arcane properties, such as mass and force or spin and charm, show up in quite distinct measurements. But this is not so clearly so for certain psychological properties. If I say that someone is intelligent and imaginative, I might be picking out two separate features. I might mean that they are both very good at grinding quickly through complicated equations and very good at inventing entertaining stories for children on the spot, thinking of these as quite different kinds of talent. But I might also be focusing on the same feature – the highly inventive way they find original solutions to difficult problems. If I call someone ‘insightful’, I am probably ascribing imagination and intelligence as one unified feature. To a certain degree, one's overall cognitive capacity is somewhat artificially divided into kinds. Perhaps God is the limiting case of this characteristic. The words we use to describe Him are literally true, in a sense, but not by chunking out discrete properties, which could be thought of as the most literal account of literalness. The reality is such that these predicates are satisfied, but not in a standard, discrete ‘patchwork’ fashion, but by some mode of being that in its richness satisfies them all.

The second point concerns the ineffability of the essence of the human mind, according to Aquinas. Aquinas thinks that only material things can disclose their essences to us, because only perceptible things allow us to abstract their form. So we cannot know the essence of anything immaterial, which, he rightly holds, includes the human mind. Aquinas also holds that the intellectual soul is the core of the total human soul and so, in a sense, is the self, though this is not the language that he uses. So if we put together the unknowability of the essence of the human self with the (in a sense) non-standard way that psychological predicates apply to it, we can see that the human soul stands, as it were, half way between the more directly intelligible world and God. This fits in with the neo-platonic understanding of the self. In Ennead VI.9 Plotinus tries to characterize the nature of the One. There is much that is mysterious and problematic in what he says, but he is clear that we can find a model for the unity and simplicity of the One in the unity and simplicity of the self. The ineffability of the One has much in common with the traditionalist conception of God, and it emphasizes the way in which the human self or person is an image of the divine.

My second thought concerns the relation between what I called above absolute, pure, or unlimited being, or pure act, and its relation to Swinburne's conception of God as ‘pure limitless intentional power’. Are these notions so different? My first thought on this concerns the relationship of these conceptions of God to the more standard properties that are attributed to Him. Perhaps Swinburne would want to say that his divine entity literally possesses the divine properties, whereas the traditional one does not, but I am not convinced of this difference. In connection with the issue of whether God simply possesses properties or is prior to them, one might ask whether pure limitless intentional power is qualified by omniscience, omnipotence, and perfect goodness as separate features, or whether the property predicates are all ways of characterizing the nature of such limitless power and not distinct items within it. In other words, the relation of the properties to the essence is no different from that specified in the traditional conception, nor from the model that I suggested for certain cases of psychological predicates in the human case.

It might still be argued that PLIP is still very different from ‘pure act’ or ‘subsistent being’. There are two obvious differences. The first is that the traditional terms depend on Aristotelian and scholastic categories which are not generally in play in modern philosophy and not in Swinburne's writing. Second, Swinburne's claim that the power is intentional emphasizes that God is a mind or, at least, mind-like, and so fundamentally a person, but the categories of act and potency, and essence and existence are much more general than that. It is not immediately obvious why pure act, or subsistent being should satisfy the psychological predicates of omniscience, omnipotence, and goodness, rather than such physical predicates as perfect gravitational force, perfect positive charge, or perfect spin. For the traditional categories to work, there must be a sense in which pure act and pure being are necessarily immaterial and spiritual forms. This should at least give pause to those, like Davies, who deny that the traditional conception regards God (as opposed to the ‘persons’ of the Trinity) as a person. Maybe this is easier to do with pure act than with pure being. If pure act lacks any remaining potentiality and matter is potentiality, then pure act will be immaterial, and as an agency it cannot be immaterial in the mere way that an abstract object is, for abstract objects are not agents. If one thinks of God as subject to no limitations and of form or essence as imposing limitations – ruling positive alternatives out – then God's nature will be His unrestricted existence. This does not look too different to me from pure unlimited intentional power, which in its plenitude, necessarily satisfies the standard predicates of omniscience etc., without their marking strictly separate features within Him.

I argued as follows in the Festschrift for Paul Helm.

Swinburne, I think, agrees that it is metaphysically necessary that everything should depend on God, so pure limitless intentional power gives us the key to the nature of existence, in the thick sense, and so this power is the thing – esse – itself.