There are three divine Persons,Footnote 1 Father, Son, and Holy Spirit. Each is distinct from each of the others, and each of them is God. There is only one God. Any doctrine of the Trinity must make consistent sense of those three propositions. The Nicene settlement entails the further claim that the Persons share a single divine nature or ousia; this shared nature is an important part of what makes it the case that the three together are one God. The classical doctrine of divine simplicity implies that each Person is identical with that Person's divine nature; furthermore, there is only one divine nature and therefore one God. This has been suggested as a solution to the problem noted above (see Ayres (Reference Ayres2004), 378). This, however, has the unfortunate consequence that, in view of the transitivity of identity, each Person is identical with each of the other two Persons, which is unacceptable. If no solution can be found for this difficulty (and I think there is none), divine simplicity cannot provide the solution for our problem.
Recently several philosophers have suggested a solution to the problem in terms of the metaphysical notion of constitution. This notion involves complex issues, some of which will be considered later on, but the fundamental idea is fairly simple. Consider the often-used but useful example of a statue made of gold. There is the statue, but there is also the lump of gold of which the statue is made. One's first thought may be that the lump of gold is identical with the statue, but this cannot be the case: not if we are thinking in terms of the strict, Leibnizian concept of identity. The lump may exist either before or after the statue (for instance, if the statue is beaten into another shape), and in any case their persistence-conditions are different: the lump of gold is such that it can persist if its overall shape is radically changed, whereas such a change will destroy the statue. So we say that the lump constitutes the statue but is not identical with the statue. Once we have the basic idea of constitution in hand, many examples become evident. A piece of cloth, dyed in a certain way, constitutes a national flag, given that there is a nation that recognizes objects with that colour-pattern as its flag. A quantity of water, frozen into a certain shape, constitutes an ice cube. A piece of paper, manipulated according to certain legal conventions, constitutes a marriage licence. And so on. The appeal of constitution in the trinitarian context is that it supplies a notion of a kind of unity between distinct items that is weaker than identity but that offers the prospect of unifying the divine Persons in a way that can solve the problem we are addressing.
Solutions to the trinitarian problem in terms of constitution are, naturally, controversial. In some instances, objections are tied to a general rejection of the notion of constitution – a response that is not unexpected, in view of the controversial status of constitution itself. Responding to objections of this nature would consume more space than is feasible here. In any case, constitution remains a viable option in contemporary metaphysics. However, Brian Leftow has formulated more specific objections to constitution in the trinitarian context, objections that do not require rejecting constitution in other situations (Leftow, Reference Leftow2018). His targets in this discussion are Michael Rea and Jeffrey Brower, Christopher Hughes, and me. Our applications of constitution to the Trinity are significantly different, so I will not undertake here to draw conclusions about the proposals of Brower and Rea and of Hughes, but I will defend my own proposal. That proposal can be stated succinctly as follows: The divine nature, the divine mind/soul, constitutes the divine trinitarian Persons when it sustains simultaneously three divine life-streams, each life-stream including cognitive, affective, and volitional states (see Hasker (Reference Hasker2013), 243). I will now summarize a (rather long) passage in which Leftow criticizes this proposal:
What constitutes the divine Persons is a Soul, which is not a Person and is not God. ‘[T]he divine soul helps make up a divine Person, as bones help make up human persons’ (368). The Soul could sensibly become an object of veneration (as the bones of dead saints are venerated) – indeed of divine-level veneration, which is very close to worship. If the Soul does not have a mental life it could not strictly be an object of worship, as worship is directed at a being that is supposed to understand that the worship is offered to it. In that case, we have an object of divine-level veneration other than God or the Persons, which is worrying. If the Soul does have a mental life, in consequence of the Persons having mental lives, the Soul could quite literally be an object of worship. But then the Soul is an additional object of worship, along with God and each of the Persons. This would amount to a divine Quaternity, which is unacceptable. (see Leftow (Reference Leftow2018), 368–370)
Furthermore,
Hasker runs into a final problem because he holds that only the whole Trinity is properly God.Footnote 2 On any orthodox account of the Trinity, God is triune by nature, not by accident or deliberate choice. Hasker agrees. For him, it's the nature of the divine soul to generate the Persons and so the whole of God. This makes it the ultimate reality in a thicker sense. It's bad enough to have something lie deeper in reality than God, but should anything generate God? (ibid., 370)
This critique makes some incorrect assumptions about my view. Taking the last point first, the constituting entity need not, and in this case does not, generate the constituted entity. (The lump of gold does not generate the statue.) On my view, as for the trinitarian tradition, it is the Father who generates the Son and the Holy Spirit.Footnote 3 God, or the Trinity, is the entire complex of Father plus Son plus Holy Spirit. The Father is fons trinitatis.
The other challenge is more complicated. Leftow states that ‘the divine soul helps make up a divine Person, as bones help make up human persons’. The reference to bones, however, is misleading. It is difficult to think of bones as anything other than parts of persons. The Soul, however, is not a part of God, and Leftow carefully refrains from saying this. But the ‘helps to make up’ relation is exceedingly vague; too vague to support any reasonable inference concerning its implications. Nor is constitution sufficiently similar to proper parthood to sustain such inferences. We have to look specifically at the constitution relation itself. Leftow is worried that the ‘worshipability’ or ‘venerability’ of the Soul may compete with that of the Persons and of God. First of all, the Soul as such does not perform mental acts apart from the Persons it constitutes, so it is not, properly, an object of worship. But it does not, in general, make sense to add two entities – one that constitutes, and the other that is constituted – together to get two instances of some property they have in common. This applies to instances of constitution quite generally. If the statue and the lump of gold each weighs 178 pounds, the weights cannot be added to make 356 pounds. If a railway ticket is constituted by a piece of cardboard, each of the piece of cardboard and the ticket has the property of conferring the right to take a railway journey, but one can't add the two together and conclude that the holder is entitled to two such journeys! When Leftow, in his own illustration, admires the marble but not the sculpture that has been carved from it, there are not two objects of sensory contemplation, one beautiful and the other ugly. Rather, there is just one object that he is assessing in two different respects, in one case favourably and in the other case not. And the divine Soul is not an additional divine entity that competes with the Persons and the Trinity for the veneration or worship that might be directed to it. The soul is each Person, and it is the Trinity; the ‘is’ in each case being the ‘is’ of constitution and not the ‘is’ of identity.
I conclude, then, that Leftow's criticisms fail to inflict any serious damage on my use of the constitution relation to explicate the doctrine of the Trinity. At this point, however, a more detailed discussion of that relation is called for. I begin with a definition of the constitution relation, paraphrased from Lynn Rudder Baker: Suppose x has F as its primary kind, and y has G as its primary kind. Then x constitutes y at time t just in case
(i) x and y are spatially coincident at t;
(ii) x is in G-favourable circumstances at t;
(iii) necessarily, if an object of primary kind F is in G-favourable circumstances at t, there is an object of primary kind G that is spatially coincident with that object at t; and
(iv) it is possible for x to exist at t but for there to be no object of primary kind G that is spatially coincident with x at t (paraphrased from Baker (Reference Baker2000), 43).
An important concept in this definition is that of a ‘primary kind’. An item's primary kind supplies the answer to the question ‘What sort of thing is it?’ Baker acknowledges that she does not have a general theory of primary kinds, but she suggests a consideration that would lead us to say whether a case is one of constitution (i.e. the appearance of a new object with a new primary kind) or of mere property acquisition. If x constitutes y, then y has whole classes of causal properties that x would not have had if x had not constituted anything. (Compare the powers to elicit certain types of responses for the mere piece of cloth and for the national flag.) And the ‘G-favourable circumstances’ are the circumstances in virtue of which an object of one primary kind, F, constitutes an object of a second primary kind, G.
This, however, is a definition of material constitution, the constitution of one physical thing by another physical thing. Clearly, some modifications will be needed if the concept is to be applied to the Trinity. One point to note concerns the time-relations: since God, and the divine Persons, are eternal (either timeless or omnitemporal), the time-references are superfluous; on the other hand, it does no harm to retain them. A more serious difficulty arises with regard to the requirement of spatial coincidence. The divine nature and the divine Persons are either non-spatial or present at every point of space, but their being this does not ensure that they coincide in the required way, unless some stipulation is made guaranteeing that nothing else is either non-spatial or omnipresent, and such a stipulation might be hard to justify. (Abstract objects, for example, are usually considered to be non-spatial.) I propose, then, that for the requirement of spatial coincidence we substitute the requirement that x and y have all their parts in common. The sharing of parts is generally taken to be inherent in the constitution relation; any two physical objects that coincide spatially will have all their parts in common, and vice versa. And Baker indicates that this would be true on her definition.Footnote 4
Yet another, more serious, difficulty arises. This difficulty concerns the final clause of the definition, ‘it is possible for x to exist at t but for there to be no object of primary kind G that is spatially coincident with x at t’. This requirement is satisfied in the case of ordinary examples of constitution: the water could exist without constituting an ice cube, the cloth could exist without constituting a national flag, and so on. But where the Trinity is concerned, this will not work: the existence of God as Trinity is metaphysically necessary, so there is no possibility of the divine soul or essence existing without the trinitarian Persons.Footnote 5
The way I have previously dealt with this problem (Hasker (Reference Hasker2013), 243–244) is to modify the relevant clause of the definition: the requirement will be, not that the constituting item should be able to exist without constituting the divine trinitarian Persons, but that its existence should be conceivable apart from that of the Persons. Whether this is so, however, may be debatable: can we coherently conceive of the divine nature without supposing it to be the nature of Father, Son, and Holy Spirit? This of course depends on what we should take the concept of the divine nature to be, and this in turn may be controversial. But even if this manoeuvre succeeds, we are left with the fact that we no longer have the same concept of constitution applied to the Trinity as is applied to ordinary cases of material constitution; at best, we have an analogous concept. This might, to be sure, be the best solution that is available, and perhaps it could be considered satisfactory.
Let us however ask, why is it that, in ordinary cases of constitution, it is possible for the constituting object to exist without constituting anything? Once asked, that question has a ready answer: in order for the constitution to occur, some causal activity is required. The piece of cloth must be coloured in the right pattern before it can constitute a national flag. The water must be frozen in a cubical shape in order to constitute an ice cube. The piece of paper must be signed by two individuals who wish to become married in order to constitute a marriage licence. (If the paper had been printed up as an example of a new typeface, and the names on it were fictitious, it would not be a marriage licence.) In most cases the causal activity will involve some modification of the constituting object, but this need not always be the case: there have no doubt been a number of instances in which a piece of cloth already in existence came to constitute a national flag as a result of an official action designating it as such. And since this causal activity is contingent, it is possible for the constituting object to exist without it, and therefore without there being a constitution relation.
Now in the application to the Trinity the contingency of the constitution relation disappears, as has been noted. But the need for causal activity does not disappear. In order for the divine mind/soul to constitute the Son, the divine nature must sustain the Son life-stream, and so also for the other two Persons. We inevitably lack any detailed account of what this ‘sustaining’ amounts to: we do not have, and presumably will never have, anything corresponding to a ‘neuroscience of the divine nature’. But the sustaining is a form of causal activity, and this is required just as much in the trinitarian case as in ordinary cases of constitution. So now we can substitute for the original:
(iv) it is possible for x to exist at t but for there to be no object of primary kind G that is spatially coincident with x at t
the following:
(iv*) in order for there to be an object of primary kind G that shares all its parts with x at t, a specific form of causal activity is required, the form of activity depending on the natures of F and G.
And now we are able to state in full the revised definition of constitution:
Suppose x has F as its primary kind, and y has G as its primary kind. Then x constitutes y just in case
(i*) x and y have all their parts in common at t;
(ii*) x is in G-favourable circumstances at t;
(iii*) necessarily, if an object of primary kind F is in G-favourable circumstances at t there is an object of primary kind G that has all its parts in common with that object; and
(iv*) in order for there to be an object of primary kind G that shares all its parts with x at t, a specific form of causal activity is required, the form of activity depending on the natures of F and G.
This definition of the constitution relation applies without modification both to ordinary cases of material constitution and to the constitution of the trinitarian Persons. In fact, it can be argued that the revised clause (iv*) is superior even in the ordinary cases, since it penetrates to the reason why, in those cases, the constitution relation is contingent.Footnote 6
As regards the Trinity, then, we have the following situation. We begin, as any orthodox doctrine of the Trinity must begin, with the Person of the Father. (References to sequence must be understood in terms of logical and/or causal dependence; all the events and states of affairs mentioned will be eternal, either atemporal or omnitemporal.) The divine nature is, in the first instance, the nature of the Father; the creed of the council of Nicaea says that the Son is ‘of the essence of the Father’ (ek tēs ousias tou patros). So the divine essence eternally, and necessarily, sustains the Father life-stream. In view of the great desirability of there being more than one divine Person, the Father eternally and necessarily causes it to be the case that the divine nature sustains also the Son life-stream, and so there is the Son as well as the Father.Footnote 7 And since the mutual love of Father and Son could only be perfected in their common love of a third Person, they bring it about that the divine nature sustains also the Spirit life-stream. So we have a Trinity of three divine Persons, each of them constituted, and the entire Trinity constituted, by the divine nature.Footnote 8 I submit that this is an appealing – and clearly, a fully orthodox – picture of the metaphysics of the tri-personal God.
A comparison: the constitution view and Leftow's Trinity
At this point it may be interesting to compare the trinitarian metaphysics put forward in this article with Leftow's own view of the Trinity (Leftow, Reference Leftow2004). Briefly stated, Leftow holds that there is a single divine person, namely God. God, however, lives his life in three discrete life-strands, the life-strands of Father, Son, and Holy Spirit. God, be it noted, lives his life only in this triply divided way; there are no mental divine acts that are not the acts either of the Father, or the Son, or the Holy Spirit. The Father's mental acts are those mental acts performed by God in the Father life-strand, and so also for the Son and the Spirit.
A little reflection reveals in this a marked similarity to William Craig's schema according to which there is a single divine soul, God, that is ‘endowed with three sets of rationalFootnote 9 . . . faculties, each sufficient for personhood’.Footnote 10 There is, however, a crucial difference between them. For Craig, each of the three ‘sets of rational faculties’ has a distinct personal subject; in possessing such a set of faculties, the divine soul constitutes a person, in quite an ordinary sense – that is, a centre of consciousness, capable of intentional acts that are cognitive, affective, and volitional. In consequence, there are three divine persons; the divine Persons are persons! Craig, in other words is a Social trinitarian. For Leftow, on the other hand, there is only one divine person, God, who lives his life in each of the three divine life-strands as Father, Son, and Holy Spirit. The Father, for example, is God living the Father-life-strand, and so also with the Son and Holy Spirit. Leftow is, quite distinctly, an ‘anti-Social’ trinitarian.Footnote 11
It may seem evident that the choice between these two ways of viewing the situation will depend in large part on what we initially take the doctrine of the Trinity to be: on how we construe the classical confessions, as well as the relevant biblical texts. But we can at this point make some preliminary observations, based simply on the metaphysical coherence and plausibility of the respective schemes. As has been shown, Craig's schema fits well with the revised definition of constitution: the divine Persons are constituted by the divine nature, insofar as the latter supports the three sets of divine rational powers. Note especially that the Persons satisfy Baker's requirement for a new ‘primary kind’: by constituting the Persons, the divine nature brings into play ‘whole classes of causal properties’, namely all those made possible by the capacity of the Persons to perform cognitive, affective, and volitional mental acts; properties which would not otherwise exist.
Where Leftow's Trinity is concerned, we cannot speak of constitution. This is so not merely on account of his expressed aversion to the notion, but because there is no new object – no ‘y’ to be constituted by ‘x’ – in his view of the Trinity. Rather, there is just the one divine person, God, living his life in one of his three discrete life-strands.
As we turn to the creeds, we encounter persistent, and probably irresolvable, disagreement between Leftow and me. We have discussed elsewhere our reasons for understanding Christian orthodoxy as supporting either Social or anti-Social trinitarianism; little would be gained by rehearsing all those arguments here. Nevertheless, I cannot forbear pointing out that Leftow's scheme endangers what many have taken to be a central commitment of the doctrine, the assertion that the Father is not identical with the Son or the Holy Spirit. On his view, to be sure, it would be misleading for the Father to say, ‘I am the Son, who died on the cross.’ This would naturally be taken to imply that the Father is God living the Son life-strand, the life-strand which included crucifixion, and this of course is not true. But the Father can truthfully say, ‘I am the person who, while living the Son life-strand, suffered and died on the cross.’ The Father and the Son are indeed the same identical individual, although that person is designated severally ‘Father’ or ‘Son’ according to the different roles he plays.Footnote 12
Finally, we turn to Scripture, the revelational source of the doctrine of the Trinity. Here I do not think there can be much disagreement about which of the two views corresponds more closely with a natural understanding of the biblical text. Jesus, the Son, prays to the Father, submits his will to the Father's will, and speaks eloquently about the Father as if about a person distinct from himself. The Father is, one might say, less empirically evident, but he supports the Son, sometimes with a voice from heaven, sometimes also with special manifestations, such as in the Transfiguration. Leftow argues, based on a time-travel story, that it is coherent to suppose that someone might confront himself in another life-phase without knowing that it was himself. No doubt this is so, but few if any before Leftow have thought this a natural way to read the Gospels. Especially poignant are the words of desolation from the Cross, ‘My God, why have you forsaken me?’ I have pointed out that, on Leftow's view, this should come out as ‘Why have I-as-Father forsaken myself-as-Son?’ To this Leftow responds, ‘It should not seem odd to us that someone would feel that he had forsaken himself if he was not aware that it was he himself who (he felt) was doing the forsaking’ (Leftow (Reference Leftow2012), 322). (Note that this episode occurred at the very end of whatever developmental process our Lord may have undergone with respect to recognizing his own divine identity.) So we have, on Leftow's own account, two incompatible views concerning Jesus's relationship to the Father: Jesus's own view, and Leftow's. We might want to consider which of the two is more likely to be correct.
