Introduction
At the core of the Chalcedonian Statement of 451, which set the standard of orthodoxy in the Christian understanding of the doctrine of the Incarnation, are two theses, one concerning identity, the other concerning the dual nature of Christ. In a single person we have ‘one and the same Christ, Son, Lord, Only-Begotten’: Christ, in other words, is numerically identical to God the Son, the second person of the Trinity. Moreover, Christ is ‘to be acknowledged in two natures, without confusion, without change, without division, without separation’. As a result of this dual nature, Christ is both ‘truly God and truly man’.Footnote 1
But can one person contain two completely opposed natures? Nowhere is this idea more vividly and shockingly portrayed than in Robert Louis Stevenson's The Strange Case of Dr Jekyll and Mr Hyde. Here is Henry Jekyll's eloquent description:
With every day, and from both sides of my intelligence, the moral and the intellectual, I thus drew steadily nearer to that truth, by whose partial discovery I have been doomed to such a dreadful shipwreck: that man is not truly one, but truly two. I say two, because the state of my own knowledge does not pass beyond that point. Others will follow, others will outstrip me on the same lines; and I hazard the guess that man will ultimately be known for a mere polity of multifarious, incongruous and independent denizens…. I saw that, of the two natures that contended in the field of my consciousness, even if I could be said to be either, it was only because I was radically both; and from an early date, even before the course of my scientific discoveries had begun to suggest the most naked possibility of such a miracle, I had learned to dwell with pleasure, as a beloved daydream, on the thought of the separation of those elements. If each, I told myself, could but be housed in separate identities, life would be relieved of all that was unbearable; the unjust might go his way, delivered from the aspirations and remorse of his more upright twin; and the just could walk steadfastly and securely on his upward path, doing the good things in which he found his pleasure, and no longer exposed to disgrace and penitence by the hands of this extraneous evil. It was the curse of mankind that these incongruous faggots were thus bound together – that in the agonized womb of consciousness, these polar twins should be continuously struggling. (Stevenson (Reference Stevenson and Robert2002), 55–56.)
Reading Stevenson's lucid narrative, the notion of two natures in one person appears not only intelligible but compelling. Why, then, should it be thought so problematic in the context of the Incarnation, which is regularly charged with inconsistency by philosophers and theologians alike?
Perhaps the reason the Jekyll-and-Hyde phenomenon is not thought paradoxical is that we can represent what is going on in such a case in terms of two sets of competing dispositions, each manifesting themselves at different times. Such conflict is not confined to the psychological. Take a rubber ball: it is disposed to absorb shocks, to compress when a force is exerted upon it. But it is also disposed to resist such compression, and to exert a force against other objects. Or consider a crystal of salt: it is disposed to dissolve when placed in water, but not to dissolve when placed in benzene. Different behaviour is exhibited in different circumstances, so we do not suppose that the disposition to exhibit a certain pattern of behaviour in some circumstances rules out the disposition, in one and the same object, to exhibit a precisely opposite pattern of behaviour in other circumstances. What is ruled out is the manifestation, at one and the same time, and by one and the same object, of those incompatible patterns. So it is in human nature: there is the disposition to good, and the disposition to evil; the disposition to be active, and the disposition to be indolent. They sit side by side, and we may feel their competing pull, but when it comes to action, there is a choice to be made.
In contrast, the two natures ascribed to Christ are not readily interpretable in this way. The divine nature is not merely disposed to acquire all knowledge, it already possesses it. The human nature does not possess it, nor, in its limited bodily frame, could it do so. The divine nature is not merely disposed to be immortal, it is immortal. The human nature, in contrast, is mortal. And even where human and divine properties can be defined dispositionally, it seems that one nature has dispositions the other lacks. The human nature is disposed to evil (not invariably, and not without there being counteracting dispositions, but still somewhat disposed to evil). The divine nature wholly lacks this disposition. Here, it seems, we cannot escape incompatible properties.
That does not mean, however, that we can learn nothing from the Jekyll-and-Hyde case that would carry over into the theological context. For what that case presents us with are not just two patterns of behaviour, but a reification of the two natures, a realization of each as, in effect, two individuals. This is not quite right, as although Hyde is presented as the undiluted evil nature, Jekyll remains a composite. But we might imagine Jekyll to have achieved his dream of entirely separating the two natures, so that, as he puts it, he would have had the fortune on occasion to ‘come forth an angel instead of a fiend’. It is tempting to describe such a separation as resulting in the dual personification of the natures, so that to two natures there correspond two persons. But that characterization can be resisted, as we shall see later. Perhaps, then, by a reification of the two natures we might try to resolve the contradiction in the idea of the Man-God. What emerges is the idea of the composite Christ, as containing different parts, each with its own set of properties. It is that idea, or one particular version of it, which I shall explore in this paper. I shall argue that although the notion of the composite Christ may appear to provide a satisfactory understanding of one of the Chalcedonian theses, that of the dual nature of Christ, it leads the other thesis, that of the identity of Christ with God the Son, to paradox.
Before going any further, however, we should consider the grounds for saying that divine properties are indeed incompatible with human ones. Perhaps it is a priori that divinity implies omniscience, omnipotence, and moral perfection. But is it similarly a priori that humanity should entail the complement of these properties? Does being human necessitate being limited in knowledge, power and moral character? Brian Hebblethwaite has pointed out that only if we have a clear grasp of the essence of divinity and humanity can we be justified in asserting that the doctrine of Incarnation involves contradiction on these grounds, and that in fact we have no such grasp (Hebblethwaite (Reference Hebblethwaite1987), 3). Following this line of thought, Thomas V. Morris has proposed a distinction between being fully human and being merely human. To be wholly human is to belong to a certain kind. To be merely human is to have nothing more than the typical member of that kind. It can easily be established on a posteriori grounds that the typical member of human kind is limited in power, knowledge, and moral character. It does not follow that this belongs essentially to the kind. Why could there not be an omniscient etc. being who nevertheless satisfied the criteria (whatever they are) for membership of the human kind? Christ could then be fully human without being merely human, just as we are fully animal but not merely animal (Morris (1986), 65–66).
We can concede this point, but still insist that it does not remove the peculiar problem posed by the Incarnation. Scripture does not present Christ as some sort of superman, remarkable though he is. He is not a giant; he does not possess superhuman strength; he does not, apparently, contain all knowledge from the dawn of time to the moment when the Seventh Seal is broken (or if he does, he keeps pretty quiet about it). Moreover it is, surely, physically impossible that a being with a normal human body should be omnipotent. And insofar as that being's mental states are realized by states of a normal human brain, it is surely impossible for such a being to be omniscient. But, crucially, it is important that Christ should be limited – ‘like us in all things apart from sin’. For it is through his becoming like us that we are saved. So the question is not: ‘Does humanity necessarily exclude divinity?’ The answer to that may well be ‘No’, or ‘Not sure, really’. The question is whether Christ shared our condition. And we know what our condition is without having to engage in essentialist talk about the nature of humanity. So the threat of contradiction comes about when we compare the divine properties with those we suppose Christ to have had, and what we, in a sense, need him to have had.
So there is a prima facie conflict between Christ's human and divine natures, a conflict which the composite picture offers to resolve. Before we consider whether it can in fact do so, however, let us take a brief look at alternative resolutions.
The reduplicative strategy and four interpretations
There is a standard formula in which the Incarnation is expressed which offers to get rid of the surface contradiction, a move known as the ‘reduplicative strategy’. It is said that, qua man, Christ suffered and died, but that qua God, He did not; qua man, Christ was limited in knowledge, but qua God He was unlimited. That is a convenient formula, but it leaves the hard work to be done. What is being picked out by the expression ‘qua man’? If all we are saying is that, by virtue of His divine nature, Christ was omniscient, and by virtue of his human nature, he was limited in knowledge, then we cannot escape the contradiction of describing him as both omniscient and limited.
Perhaps, then, we can exploit an analogy. In his discussion of the reduplicative strategy, Morris considers an analogy proposed by R. T. Herbert (Reference Herbert1979): the ‘duck-rabbit’, made famous in philosophical circles by Wittgenstein. The duck-rabbit is an ambiguous picture. It is so drawn that it sometimes looks like a duck, looking to the left. But then, in a sudden gestalt, the mind turns the beak into two ears, and the slight bump on the duck's head becomes the nose of a rabbit, head upturned, looking to the right. How could we have failed to see it? In another moment, the rabbit is gone, and we see the duck again. Now here is a case where the reduplicative strategy has a convenient application. Qua duck, we can say, the animal has a beak but no floppy ears, qua rabbit, it has floppy ears but no beak.
Not surprisingly, Morris finds the analogy inadequate. It is entirely clear what the ‘qua’ locution picks out here, and it is something we bring to the picture. In itself, the drawing is nothing but lines on paper. It causes us, however, sometimes to see a duck and sometimes a rabbit. It is capable of being seen as representing first one thing, and then another. If we see it as a rabbit, we see the floppy ears. If we see it as a duck, we see the beak. These features are entirely extrinsic to the drawing. But omniscience and its complement are not extrinsic to Christ: these are not merely properties that we bring to our interpretation of Christ, as if, in certain moods, we can choose to think him omniscient, and in other moods, choose to think him limited. He is both omniscient and limited, whether we choose to see it or not.
Let us try a different analogy. In a much discussed paper, Donald Davidson (Reference Davidson1970) wrestles with the following paradox. Sudden desires, resolutions, good intentions, newly-acquired beliefs – in short, mental events – can cause us to behave in certain ways: they have, in other words, a physical effect. Now causation is part of the order in the world: it is not merely haphazard, but subject to certain laws. This piece of iron did not just happen to expand when it was heated: iron as a matter of course expands when heated. Causes and effects, then, fall under general laws that link them. But there are no psycho-physical laws, no laws linking the mental and the physical. We cannot say that a such-and-such mental state invariably leads to a such-and-such physical effect. What a mental event brings about depends a great deal on what other mental events are in the vicinity. Moreover, Davidson insists, the vocabulary of mental talk is quite different from that of physical talk: they simply cannot be mixed. Finally, if mental events did fall under strict laws, would that not pose a problem for our freedom? But we know we are free, and there's an end on't.
So we have a problem: if this mental event caused that physical effect, then there should be a law linking them. But there is no such law. Davidson's solution to the problem is what he calls anomalous monism: there are not two kinds of events, mental and physical, there is just one kind of event. But some events can be described both in physical terms and in mental terms. The expression ‘my becoming aware of Rita’, and ‘my brain state φ’ can pick out one and the same event. But whether or not an event falls under a law depends on how it is described. If we choose to describe it in mental terms, then it is not covered by a law; if we choose to describe it in physical terms, however, it is covered by a law. Once again, we can employ the terms of the reduplicative strategy: qua physical event, it falls under a general law linking it with its effect; qua mental event, it does not fall under any such law. So mental-physical causation, the lawlikeness of causation, and our freedom of will,Footnote 3 are all preserved.
This has been a hugely influential idea in the philosophy of mind. But anomalous monism, for all its intrinsic merits, does not provide the appropriate model for the dual nature of Christ, and that is because it is primarily concerned with types, not with tokens. Laws relate types: this kind of property is necessarily connected with that kind of property. Particular instances of those properties are governed by laws by virtue of the types they fall under. It is this feature of laws that allows us to understand how it can be that an interaction can fall under a law qua physical interaction, but not fall under any law qua psycho-physical interaction. Physicalists in the philosophy of mind typically reject type-type identity: psychological type ϕ coincides precisely with no physical type since ϕ is variably realized. That is, different instances of type ϕ will typically differ in the physical types they exemplify. So when we say that this event falls under a law qua physical event but not qua psychological event, we mean that the physical type it exemplifies is the subject of a law, whereas the psychological type it exemplifies is not. This does not legitimize the suggestion that this token event both is and is not governed by a law. In contrast, the dual nature of Christ is essentially about the token, not merely about the types. To say that Christ is omniscient qua divine being, but limited in knowledge qua human being, is to say more than just that omniscience follows from the type ‘divine’, but not from the type ‘human’. The implication is that it is true to say of Christ both that He is omniscient and that he is not.
Consider a third model: the Cartesian person, which is a composite of soul and body. Although Descartes wants to say that he is essentially a thing that thinks, he is not simply identifying the person with the soul:
Nature … teaches me by these sensations of pain, hunger, thirst and so on that I am not merely present in my body as a sailor is present in a ship, but that I am very closely joined and, as it were, intermingled with it, so that I and the body form a unit. (Meditation VI; Cottingham et al. (Reference Cottingham, Robert and Dugald1984), 56.)
Here again is a case where the reduplicative strategy beckons: qua physical being, I am mortal, qua soul I am immortal. This looks much closer to the case of Christ, in that here we are dealing with the intrinsic properties of a token being. But, if the Cartesian conception of the embodied person is correct, ‘physical being’, as applied to the human person, becomes a phase sortal. I am a physical being only during one phase of my existence. To say, then, that ‘qua physical being, I am mortal’, is to say just that the physical, embodied phase of my existence is temporally limited. Now, if we apply this to the incarnational case, and say that, during the human phase of his existence Christ is limited in knowledge, but during the non-human phase(s) He is omniscient, then we have the kenotic view, according to which the Son gave up (at least some of) the characteristic properties of divinity in order to become human.Footnote 2 But the point of the reduplicative strategy is to make intelligible, to save from contradiction, the orthodox view, on which human and divine properties are simultaneously attributable to Christ.
So once again, we have not found a suitable model. What remains? Perhaps the reason Christ can be both human and divine is that he, like Jekyll, is a composite being: he has different parts, one human, one divine. This provides a clear and intelligible understanding of the reduplicative strategy: when we say ‘qua man, Christ was mortal’, we pick out the human part, and ascribe mortality to it. And since these are only parts of the individual, they are not persons in their own right, but they are, or have, natures. Thus, in the composite Christ we can have two natures in one person.
The human and divine parts are parts of a single substance – that is, an object that persists through time, enjoys a certain independence from other objects, and is a single individual. The properties of the parts can carry over to the whole. Thus we say that Christ is human by virtue of having a human part. Christ suffers by virtue of having a part that does so, even though another part does not suffer. Here is a simple analogy: we say the apple is red, even though it borrows that redness from only part of it, its skin. But it would be pointless pedantry to say, ‘the apple itself isn't red, only the skin is, the flesh within is quite a different colour’.Footnote 4
Objects, then, may truly be ascribed properties by virtue of the properties of their parts. The redness of the skin is reported as the redness of the apple. The wounding of the abdomen is reported as the wounding of the person. But given that different parts have different properties, when we come to report the properties of the whole, do some properties not trump others, as it were? Hermione is intelligent by virtue of her brain, not by virtue of her nose. Her nose, considered in itself, is not intelligent. But we describe Hermione in unqualified terms as intelligent. We do not say: she is intelligent and not intelligent. Or (just to make things absolutely plain) Hermione is intelligent qua enbrained individual, but not intelligent qua benosed individual. We think the intelligence of the brain trumps the lack of it in other organs. Why, then, do we not say, with a similar lack of qualification, Christ is omniscient (by virtue of possessing an omniscient part)? For surely the omniscience of the divine part trumps the epistemic limitations of the human part. Perhaps the answer to this is that we continue to ascribe only limited knowledge to Christ, qua human, because that limitedness is manifested: he does not appear to know what will happen, even in situations where that knowledge is crucial. By contrast, Hermione's nose does not manifest its lack of intelligence: it simply is never called upon to do anything intelligent, so is never embarrassed in public.
Perhaps a deeper worry about the composite account is that, despite the apparent unity, we have two persons here: the divine person, and the human person. For, prior to the Incarnation, what became the divine part of Christ was the Second Person of the Trinity. And Christ's human part contains everything that is possessed by human persons (a human soul conjoined to a human body, for instance, or perhaps a purely physical basis of consciousness). But this worry can be addressed: although, in the absence of the divine part annexed to it, the human part would constitute a person in its own right, where it exists merely as part of the composite Christ, it does not do so. And the personal status of the divine part is similarly affected by whether it exists as a part of an individual, or as an individual in its own right. As we might put it: mere parts cannot be people, however accomplished or even divine those parts may be.
As Brian Leftow has pointed out, there is a convincing argument for the principle that no parts of a person can themselves be persons, which can be derived from Peter Geach's (1980) discussion of the paradox of 1,001 cats, which I present in modified form as follows.
Consider the cat Tibbles. Tibbles is sitting on a rug. He has a tail. But he has a part which contains every part of Tibbles except the tail. This tailless but otherwise very catlike part we will call ‘Tib’. Now, here is the question: Is Tib a cat or not? Were we to encounter Tib on his own, and not merely as a part of another, larger entity, we would not hesitate to call him a cat. But he is not on his own. He is a part of Tibbles. Suppose, then, we say that Tib is a cat. Then we are obliged to say that there are two cats on the rug, not one. We have to say this because Tib lacks something that Tibbles has, namely a tail, and it is impossible for one and the same cat to both have, and lack, a tail at one and the same time. And it does not stop there. For there is a part of Tibbles that includes all of Tibbles except the whiskers, call it ‘Ib’. If Tib is a cat, Ib is a cat, but not the same cat as Tib. So there are three cats on the rug. And there is another part, which contains all of Tibbles except the ears, and yet another which differs from Tibbles only by a single hair. And so on. Granting Tib cat-status opens the floodgates: everyone is applying. And soon we have as many cats on the rug as Tibbles has parts.
Geach's own resolution of the paradox is to invoke the notion of relative identity: the truth of an identity statement a=b is sensitive to the sortals we take a and b to fall under. Relative to one sortal they are identical, relative to another, they are not. We can say, then, that Tibbles and Tib are the same cat, but not the same mass of feline tissue (Geach (1980), 216). We shall examine this move in more detail later. But Leftow's suggestion is that the paradox can be avoided if we simply deny that Tib is a cat, by virtue of being a part of one (Leftow (Reference Leftow2002), 281–282). By parallel reasoning, then, the composite Christ is one person, not two.
There is a purely theological objection to the composite model, and that is that it is ruled out by the words of the Chalcedonian definition that we have already quoted: Christ is ‘to be acknowledged in two natures, without confusion, without change, without division, without separation’. The composite picture, it would seem, introduces both division and separation. It cannot, therefore, present itself as an orthodox account. However, the definition goes on: ‘the characteristic property of each nature being preserved, and concurring in one person and one hypostasis, not as if Christ were parted or divided into two persons, but one and the same Son and Only-Begotten God, Word, Lord, Jesus Christ’. This explains the sense in which there is no division or separation; the natures coexist in one person, not two. Provided, then, we do not think of the parts as constituting persons in their own right, the composite picture need not be taken to be directly at odds with Chalcedon.
I am not suggesting that the four models sketched above are exhaustive. I have not yet mentioned one very influential account of Christ's dual nature, namely the ‘two-minds’ account, defended by Swinburne, Hebblethwaite, and Morris. I shall have more to say about this account later, but for now we have before us the essential elements of an interpretation of the reduplicative strategy in terms of parts and wholes, and my focus in this paper is on that interpretation. It undoubtedly avoids one incarnational contradiction. But, I submit, paradox still lurks in the wings.
A paradox of composition
The case of Tib, the catlike non-cat, provides us with an example of what Peter van Inwagen (Reference van Inwagen1981) calls an arbitrary undetached part: arbitrary, because the decision to draw Tib's boundaries as we did was purely stipulatory, not something forced on us by nature; and undetached, because Tib was not separated from Tibbles and his tail. Those who, unlike van Inwagen, believe in the existence of arbitrary undetached parts, see them everywhere. As van Inwagen points out, they believe that Michaelangelo's David existed for years encased in the lump of rock from which Michaelangelo eventually liberated it. There were, in fact, many Davids, or David-like objects, in this rock before one of them got lucky, and the others were destroyed. Before it was thus liberated, David was merely one of many arbitrary undetached parts of the rock. It would have existed even if Michaelangelo had not chosen this rock, and it had remained untouched by any sculptor (van Inwagen (Reference van Inwagen1981), 75).
There is something slightly unusual, but not at all implausible in all this. But van Inwagen thinks that we should deny the existence of arbitrary undetached parts, and his reasons for thinking so, which exploit a paradox originally presented by the Stoic philosopher Chrysippus, have repercussions for the composite Christ.
Since we have been introduced to Tibbles and Tib, let me present van Inwagen's argument with them as the central characters. Just to remind you: Tib is all of Tibbles except for the tail. Tib is all you would get if Tibbles lost his tail. The first step is one we have already taken, which is to recognize that Tibbles and Tib are two different things. Now, on Monday, that remains the situation. First thing on Tuesday, however, Tibbles loses his tail. (Let us not go into the distressing details.) What has become of Tibbles and Tib? Well, although Tibble's loss was not negligible, it was not fatal: he remains in existence, even if in a somewhat diminished state. As for Tib, he has remained completely unaffected by the episode (except, as we shall see in a moment, in terms of status): since he never had a tail, he cannot have lost it. So both Tibbles and Tib are around on Tuesday. But are they different entities anymore? They have exactly the same intrinsic properties, and they occupy exactly the same space. They are, surely, one and the same. And this, incidentally, is good news for Tib, for he can now be recognized as a cat.
But among these commonplace reflections we have the ingredients for paradox. First, Tibbles on Tuesday is identical to Tibbles on Monday, for mere loss of a tail does not make Tibbles history. Second, Tib on Tuesday is identical to Tib on Monday, for he has remained entirely unaltered. Third, Tibbles on Tuesday is identical to Tib on Tuesday: they have exactly the same matter. Fourth, Tibbles on Monday is not identical to Tib on Monday: one has a tail, the other not. But how can this be? For identity is transitive: if Tibbles on Tuesday is identical to Tib on Tuesday, and if Tibbles on Tuesday is identical to Tibbles on Monday, then, inexorably, Tib on Monday is identical to Tibbles on Tuesday. Having established that, we can derive the further conclusion that Tibbles on Monday is identical to Tib on Monday. But this we have already ruled out.
To represent the paradox diagrammatically:
How should we respond to this paradox? I shall follow Mark Heller's (1984) helpful summary of the options:
(1) Reject the doctrine of arbitrary undetached parts. It is not just that Tib does not count as a cat; he does not count as an object at all. This is van Inwagen's own solution, and it seems a hard conclusion to draw. After all, we seem to be able to think about Tib, to describe him in minute detail, even to point him out. Surely, it is just bizarre to suppose that there is nothing there. Well, van Inwagen does not want to say anything as odd as that. There is something there. Or rather, there are many somethings there, only they are very very small. There are fundamental particles, and they are arranged in such a way as to give us the impression of there being a medium-sized object there. For van Inwagen, there are just two kinds of objects in the world: simples, which have no parts, and living organisms (van Inwagen (1990), 98–107).
(2) Deny the transitivity of identity. This is a desperate move. van Inwagen quotes Geach's remark that those who recommend this line are ‘not to be heard’. There is, however, a less desperate move that can appear to have the effect of denying transitivity, discussed below.
(3) Allow coincident objects. We assumed that Tib on Tuesday was identical to Tibbles on Tuesday, on the grounds that they are wholly spatially coincident. But we might resist this, and hold that, by virtue of their different histories (one of them once possessed a tail and the other did not) they remain different objects. This is the line recommended by Judith Jarvis Thomson (1983).
(4) Adopt mereological essentialism. This holds that objects cannot change their parts and remain in existence. The consequence of this is that Tibbles would, on losing his tail, cease to exist, and therefore cannot be a candidate for identity with Tib, who persists. Roderick Chisholm (Reference Chisholm1973) is an example of such an essentialist.
(5) Reject strict numerical identity over time in favour of something weaker. According to the ‘perdurance’ approach to persistence, an object is extended through time by virtue of having different temporal parts at each of the times that compose its temporal extension. Thus extension is time is analogous to extension in space. No temporal part is numerically identical to any other, but the existence of a sufficiently close connection between the parts – a continuous chain of causal connections running through them, for example – legitimizes ordinary talk of identity through time.Footnote 6 This is the resolution Heller alights on.
(6) Accept the thesis of the relativity of identity. We have already introduced this proposal of Geach's above. Here is how it would impact on van Inwagen's paradox of composition. Two propositions are central to the paradox: that Tibbles on Monday is identical to Tibbles on Tuesday, and that Tib on Monday is identical to Tib on Tuesday. We assumed that we were dealing with exactly the same relation in each case. But when we start relativising identity to particular kinds, a crucial difference emerges. Tibbles on Monday is the same cat as Tibbles on Tuesday, but he is not the same lump of matter, because by Tuesday there is a bit missing. Tib on Monday is the same lump of matter as Tib on Tuesday, for, unlike the case of Tibbles, nothing has befallen Tib. But Tib on Monday is not the same cat as Tib on Tuesday because Tib on Monday is not a cat at all, merely part of one. Now, to be able to appeal to transitivity, we have to make sure that we are talking about the same kind in all our identity assertions. It does follow that if A is the same cat as B, and B is the same cat as C, then A is the same cat as C. It does not follow that, if A is the same cat as B, and B is the same lump of matter as C, then A is the same cat as C. So the Tibbles paradox, arguably, depends on suppressing the kind-sensitivity of identity statements.
Insofar as it undermines an argument that crucially appeals to the transitivity of identity, it might seem that the relative identity thesis actually makes identity a non-transitive relation. In fact, it does not, since the relativized relation ‘is the same F as’ will still turn out to be transitive.
One way or another, it seems that the paradox of composition demands some revision in our ideas either of objects or of identity. What happens when we put the paradox in the context of the Incarnation?
A Christological paradox of composition
Combining the orthodox understanding of the Incarnation with the idea of Christ as composite we get the result that the Second Person of the Trinity becomes the divine part of Christ. The divine part is not intrinsically changed by the Incarnation, it simply becomes a part of the composite Christ. The divine part of Christ, then, existed before the Incarnation, but it was, of course, not then a part of Christ. We need a name to refer to this, and I shall follow Eleonore Stump in calling it ‘the divine nature’.Footnote 5 The divine nature, then, persists. But the divine nature is not identical to Christ, for Christ has parts that the divine nature lacks. Christ is composite, the divine nature not. Yet according to orthodoxy Christ is the Second Person of the Trinity incarnate. Again, we have the ingredients of paradox:
(a) The pre-incarnate divine nature=the incarnate divine nature (since nothing intrinsic has happened to it).
(b) The pre-incarnate divine nature=the Second Person of the Trinity (since the three members of the Trinity exhaust its composition, and nothing else is divine, so only the Second Person of the Trinity could be identical to the divine nature, given that neither the Father nor the Holy Ghost become incarnate).
(c) The Second Person of the Trinity=Christ (since Christ is the Son made man).
From (a), (b), and transitivity:
(d) The Second Person of the Trinity=the incarnate divine nature.
From (c), (d), and transitivity:
(e) Christ=the incarnate divine nature.
But (e) is false, as the divine nature is only part of Christ.
In diagrammatic form:
This is, in structure, precisely the same paradox as the one concerning Tibbles: both involve identity statements, from which is derived a further identity statement, x=y, which conflicts with the previously established truth that x is merely a proper part of y.
That is not to say, of course, that they are parallel in all respects. van Inwagen sets up the original paradox as one concerning material objects, and his purpose in so doing is to argue against a thesis of material structure: that there are arbitrary undetached material parts. The divine nature, in contrast, is standardly supposed to be non-material. And whereas Tib is an arbitrary undetached part, the divine nature is anything but arbitrary. Thus the first way out of the Tibbles paradox (viz. banning arbitrary undetached material parts) is not available in the case of the Incarnational paradox. In addition, one of the principles at work in the Tibbles case, that there cannot be two numerically distinct but wholly spatially coincident objects, has no obvious application in the Incarnational case, so rejecting that principle cannot help us here. But we have independent, theological reasons for accepting (b), which is the counterpart of the assertion that Tib on Tuesday is identical to Tibbles on Tuesday. For if the divine nature that becomes incarnate is not the Second Person of the Trinity, then, against orthodoxy, it is not the Son who becomes incarnate. It would also raise difficulties for the Trinity, since it implies four persons, or three persons plus one nature, in one godhead.
There is potentially a more radical way in which the cases of Tibbles and Christ diverge. The Tibbles paradox crucially involves identity through time. In presenting the incarnational version, I preserved this feature, talking of the ‘pre-incarnate’ and ‘incarnate’ divine nature. But what if the divine nature is timeless? (See, e.g. Helm (Reference Helm1988), Leftow (Reference Leftow2002).) Then, arguably, it makes no sense to say that it persists through time. However, even if we take out the temporal structure of the paradox, a problem remains. The question concerns the relation between Christ and the timeless divine nature. On the one hand, it should be identity, since Christ is the incarnate Second Person (and we can timelessly identify the divine nature with the Second Person). On the other, the divine nature is merely part of Christ, since Christ also has human parts.
The effect of these dissimilarities is not to undermine the Incarnational paradox, but rather to restrict the number of manoeuvres we can make in attempting to avoid it. However, it might be thought that there is an orthodox alternative to (c) that does not have a counterpart in the Tibbles case, such that we do not have to base our rejection of (c) on a controversial philosophical position such as mereological essentialism. The suggestion is this: Christ is not, despite the wording of the Chalcedonian definition, held by the orthodox to be identical to the Son. Rather, the Son takes on flesh, and a new person comes into existence, namely the composite of divine and human elements. Thus, Christ is uniquely related to the godhead, and so is genuinely divine. But it would be wrong to say that Christ was identical with a member of the divine Trinity.
Now, as a matter of fact, orthodox statements of the doctrine of the Incarnation do present it in terms of identity, rather than some other relation between Christ and God the Son:
No relationship is adequate to unite the Son of Mary with the Son of the Father except the relationship of numerical identity. (Williams (Reference Williams1968), 523)
The core claim of the traditional Christian doctrine of the Incarnation, the fundamental and most distinctive tenet of the Christian faith as defined by the Council of Chalcedon (AD 451), is the claim that the person who was and is Jesus of Nazareth is one and the same individual as God the Son, the Second Person of the divine Trinity – a literal statement of absolute, numerical identity. (Morris (1986), 18–19)
… the doctrine expresses, so far as human words permit, the central belief of Christians that God himself, without ceasing to be God, has come amongst us, not just in but as a particular man, at a particular time and place. The human life lived and the death died have been held quite literally to be the human life and death of God himself in one of the modes of his own eternal being. (Hebblethwaite (Reference Hebblethwaite1987), 1–2)
Christians hold that God became incarnate – somehow became the man Jesus of Nazareth. (Leftow (Reference Leftow2002), 273)
There are good reasons for taking this line. If we say that the Second Person of the Trinity is not Christ, but Christ, as it were, replaced the Second Person, then the Second Person as a person ceased to exist. Thus, during the Incarnation, the Second Person of the Trinity was not in existence. This is not an acceptable result. If, to preserve the Second Person, we say that the divine nature was a person in its own right, though only part of the composite Christ, then we have Christ as two persons – the Nestorianism condemned by Chalcedon. Or at least we have to abandon the principle that no part of a person is a person, and would therefore be hard put to explain why the human part of Christ is not a person in its own right.
Perhaps there is another possibility. Suppose, in place of identity between the divine nature and the Second Person, we substitute a constitution relation: the divine nature constitutes, but is not identical to, the Second Person. The two go on to have different histories, for whereas the Second Person is identical to Christ, the divine nature is merely part of Christ. But this is a dangerous step to take. First, not everyone recognizes a distinction between constitution and identity (see e.g. Noonan (Reference Noonan1993) ). Second, the paradigm examples of constitution are material in nature: a piece of gold, for instance, may constitute a statue. But since the relation between the divine nature and the Second Person is not (presumably) a material constitution, then we have to say that the relation between them is merely analogous to the way in which a piece of gold constitutes a statue, and that means a partial abandonment of the original goal of providing a metaphysical account of the Incarnation. Third, the move endangers, or at least complicates, the doctrine of the Trinity, for in addition to the three persons of the Trinity we would have to recognize, as separate objects, the natures that constitute those persons.
What moves are we left with? Assuming that we do not wish to reject the transitivity of identity, there appear to be two options left: reject strict numerical identity over time and replace it with a temporal parts ontology, or assert the relativity of identity. Let us consider these in detail, beginning with relative identity.
Relative identity and regress
The central idea of relative identity, recall, is this. If we say ‘x is identical to y’ (in the sense of being one and the same individual, not merely similar), what we say is incomplete. We should report the matter like this: ‘x is the same F as y’, where ‘F’ stands for some sortal term. So we say ‘Tully was the same man as Cicero’, ‘Hesperus is the same planet as Phosphorus’. So far, this simply respects Locke's point that different kinds of thing have different identity criteria, but the really radical idea in relative identity is that x might be the same F as y, but not the same G as y. It might, for instance, be the same lump of matter as y, but not the same artwork. Let L be a lump of bronze before it is fashioned into a statue. L is then melted and poured into a mould. Call the resulting statue S. L is not the same statue as S because L is not a statue at all. But S and L are the same lump of bronze. The identity of S and L is thus relative to a particular kind.
Applying the idea of relative identity to the Incarnational paradox requires us to qualify the three components of the paradox as follows:
(a)* The pre-incarnate divine nature is the same nature as the incarnate divine nature.
(b)* The pre-incarnate divine nature is the same person as the Second Person of the Trinity.
(c)* The Second Person of the Trinity is the same person as Christ.
But now we can no longer derive, from (a)* and (b)*, the conclusion that the Second Person of the Trinity is identical to the incarnate divine nature. We cannot even derive a relativized equivalent, namely that that the Second Person of the Trinity is the same nature as the incarnate divine nature. We might, of course, think that this was in any case true, but we could still not derive, from this together with (c)*, the further disastrous conclusion that Christ is identical with, or is the same person or the same nature as the incarnate divine nature.
The fact that relative identity offers a way out of the paradox (and, indeed, out of the Tibbles paradox) surely provides a boost to this generally ill-favoured theory.Footnote 7 If we take this route out of the problem, however, we face regress. The standard abstract schema for relative identity is this:
x is the same F as y, but not the same G as y.
What we have here is a conjunction:
(a) x is the same F as y,
and,
(b) x is not the same G as y.
The idea is radical precisely because we take it that it is very same thing that both stands in the same F-relation to y and also fails to stand in the same G relation to y. But the relative identity thesis does not have the resources to express this idea – in fact, it prevents us from doing so, for talk of the ‘very same thing’ is an appeal to absolute (non-relative) identity, and if the relative identity thesis is correct, there is no such thing. So we have to treat the ‘x’ in (a) and the ‘x’ in (b) as expressions for things whose relation needs to be defined. We therefore have to say something like this:
x in (a) is the same F as x in (b).
But here again, we cannot say or imply that the very same thing that stands in the same F-relation to y also stands in the same F-relation to something which fails to stand in the same G-relation to y. We have to say, instead, that what stands in the same F-relation to y also stands in the same F-relation to something that stands in the F-relation, etc. But we cannot avoid use of the term ‘also stands’, or some equivalent expression, and this is nothing other than an expression of identity, which therefore has to be construed as relative to a sortal. In expressing this, however, we never manage to eliminate the identity-implying phrase, which means that we are embarked on regress. Moreover, the regress is not merely just one of propositions, each trivially entailing the ones lower down. It is a regress of real relations.
It is sometimes said that identity is not a relation. But relative identity has to be a relation, and wherever we find an x and y standing in different relative identity relations to each other, we have to posit a complex of other relative identity relations linking the relata of the first set of relations. The regress is therefore a metaphysical one, and we should avoid it if at all possible.
A perduring Son?
And so, having cast a dark and fearful eye on the alternatives, we turn to the remaining response to the paradox of the composite Christ: maintain that the pre-incarnate Son perdures as Christ. In other words, we take the Son and Christ to be different parts (or rather, different collections of temporal parts) that are intimately joined together. Insofar as this is an abandonment of strict numerical identity through time, it might seem to be completely at odds with Chalcedon, which insists on the identity of Christ and the Son. But Chalcedon does not, in any detail at any rate, provide us with a metaphysical theory of the Incarnation. What it provides is a framework within which we have to construct our own metaphysical theory. The perdurantist account of persistence through time does not, note, force us to give up ordinary talk of identity. What the theory does is to provide the truth-conditions of such talk. When you say that Peter in 1950 is identical to Peter in 2008, what you say is true precisely in virtue of the relations (causal, spatio-temporal, qualitative, etc.) that link the temporal part that exists in 1950 with the temporal parts that exists in 2008. Thus, the perdurantist view, as applied to the Incarnation, allows us to continue to say that Christ is identical to the Son, and also that the pre-incarnate divine nature is identical to the incarnate divine nature, where this is ‘ordinary identity’ talk about persistence, not ‘strict identity’ talk. But when we add the third side of the square, and say that the Son is identical to the pre-incarnate divine nature, we invoke a quite different kind of fact from the kind that underlies the first two identity statements. Since this fact is not about persistence, we can speak of strict identity here. It is this difference between the two kinds of statement that prevents us from drawing the conclusion that Christ is identical to his divine part.
There is, then, a coherent, though not uncontroversial, philosophical response to the compositional paradox of the Incarnation. But is it theologically acceptable? Will those who insist on the identity of Christ with God the Son be content with this watered-down version of identity over time? Two worries present themselves (no doubt there are more). The first is that perdurantism forces on us a view of God the Son as temporal, which then faces the orthodox arguments put forward in favour of divine timelessness. This might, perhaps, be open to question. Could we not imagine a composite consisting of a timeless God the Son together with the collection of temporal parts that is Christ, and so joined together by intimate relations of a causal or quasi-causal kind (so that, for instance, Christ has a certain nature because of the Son's properties) that we could talk of Christ and the Son being one and the same? Well, perhaps.Footnote 8
Another worry is that perdurantism allows branching objects in a way that strict identity does not. Consider an object A undergoing fission into two objects B and C. A cannot be strictly identical with B and C, since that would imply, given the transitivity of strict identity, that B are C were identical, which they are not. Strict identity therefore requires uniqueness: Peter in 1950 can be strictly identical with at most one person in 2008. But perdurance is rather different. A can be said to perdure as both B and C. B and C, that is, are parts of a branching collection of temporal parts that has A as one of its pre-branching parts. Perdurance, then, does not require uniqueness. And this, arguably, makes it inadequate for the theological role it is called on to play. Why were there not multiple incarnations? If the Son could have been made man, not just in Bethlehem, but in many places all over the world at that time, why did it happen just once? The answer, suggests Hebblethwaite, is that multiple incarnations are ruled out metaphysically, by the logical properties of identity. The Son could only be identical with at most one human being (Hebblethwaite (Reference Hebblethwaite1987), 49–52). But if we replace strict identity with perdurance, this response is no longer available to us. The Son could have perdured as many human beings, each quite independent of each other.
All this may well incline us to abandon the picture of the composite Christ for some other model. Which one? Earlier, I alluded to the ‘two-minds’ view, according to which Christ's mental life is divided into two streams of consciousness: the human, limited mind, and the divine, unlimited mind. Between them there is an ‘asymmetric accessing relation’ (Morris (1986), 103), whereby the divine mind is aware of all the contents of the human mind, but not vice versa. I do not have the space to discuss this account in detail, but I have three questions for those attracted to it. First, is this a genuine alternative to the composite model? It is certainly possible to construe the two-minds account as presenting different parts of Christ, though that is not how it is standardly presented. But, and this is the second question, if the two minds are not different parts of Christ, what exactly is the relation they stand in to each other? We might suppose the divine mind to contain the human mind (Morris, ibid., Swinburne (Reference Swinburne and Godfrey1989), 65).Footnote 9 This raises the third question: if the human mind is a subset of the divine mind, how is it possible for the human mind to entertain thoughts that the divine mind knows to be false (such as, for instance, ‘It is possible for me to sin’)?
Conclusion
I have suggested that, if they want to retain the orthodox understanding of Christ as identical to the Second Person of the Trinity, the only way for proponents of the composite Christ to avoid paradox is to accept perdurantism. That might be too high a price to pay, theologically (and perhaps also philosophically). If so, we shall have either to revise the orthodox understanding, or to abandon the idea of human and divine parts, and look elsewhere for an interpretation of the phrase ‘qua God, Christ had all the properties of divinity; qua man, he did not’, perhaps in terms of the ‘two-minds’ account, insofar as this is a genuine alternative to the composite account.
Let me end with some less abstract ruminations, which take us back to the starting point of this paper. Doctor Jekyll sought to divide his conflicted self into two monolithic selves. In this, he admits, he was only partially successful: only Mr Hyde emerged with a uniform nature. But perhaps even this is overstating the matter. Hyde is not entirely indifferent to Jekyll's concerns: he shows a certain desire to keep up civilized appearances. And it is Hyde that seeks to be transformed back into Jekyll. He is certainly no primordial creature from hell. In his introduction to the Penguin edition of The Strange Case of Dr Jekyll and Mr Hyde, Robert Mighall remarks that:
Jekyll attempts to make an absolute division between the respectable and the disreputable, the righteous and the libertine, the social and the sensual/sexual. But he fails. Ostensibly because of a mistake with his chemicals; but the experiment also fails because the divisions Jekyll imagines and attempts to solidify were impossible to sustain. It is not only the chemicals that are ‘impure’, the differences he considers to be absolute are also decidedly mixed and confused. (Stevenson (Reference Stevenson and Robert2002), xxv.)
An absolute division into two completely opposite natures, then, may not be possible: each affects, and is affected by, the other. How then, are we to imagine the divine and human parts of Christ? Do they run on parallel lines, never meeting or interacting? Or does each infuse the other with its own perspectives, thoughts and desires? The theory of the composite Christ has to keep the two parts as separate as possible. But a thorough interaction between the two natures may be essential to the unity of person insisted upon by Chalcedon.Footnote 10
