Three Models of the Trinity

A few months ago I wrote a post exploring an aporetic triad related to the Christian doctrine of the Trinity. In this post I want to approach the topic from a different angle by considering three different models of the Trinity. By “the Trinity” I mean a tri-personal version of monotheism such as is depicted in the image to the left. More specifically, any model of the Trinity must affirm at least the following: (1) There eternally exists exactly one God (i.e., one concrete instance of the divine nature). (2) There eternally exist three fully divine persons: Father, Son, and Spirit. As fully divine each of these persons fully exemplifies all of the intrinsic perfections proper to the divine nature and thus “is God.” (3) These three divine persons are mutually distinct, and this distinction is not merely nominal or conceptual. Thus the Father objectively “is not” the Son or the Spirit and the Spirit objectively “is not” the Son.

Three Models of the Trinity

Now, as I said, I want to discuss three models of the Trinity. These models differ by what they take the foundation of the Trinity to be. Here they are:

• Monarchial trinitarianism – the Father is foundational. He is the singular source or archē of the Trinity. The Son and Spirit are eternally caused to exist by the Father and thus are ontologically derivative. In causing the Son and Spirit to be, the Father shares his entire concrete divine nature with them. For this model (I shall argue below) the Father “is God” by identity whereas the Son/Spirit each “is God” by predication. Because identity is symmetric, we can thus truly say “God is the Father” but we cannot truly say “God is the Son” or “God is the Spirit.”
• Latin trinitarianism – the divine essence is foundational. This model presupposes absolute divine simplicity. Thus, God is (=) the divine essence. The three trinitarian persons are relations of the divine essence to itself. In another post I point out that for this model the Father, Son, and Spirit can be mapped on to the three equivalence relations (reflexivity, symmetry, transitivity) entailed by the identity relation. Each of the persons “is God” by predication and not by identity.
• Trinity monotheism – the Trinity as a whole is foundational. Neither the Father nor the divine essence is ontologically prior to the Son and Spirit. Rather, the Trinity is an irreducible 3-persons-in-1-nature package deal. So God is (=) the Trinity. Each of the persons “is God” by predication and not by identity.

DISCLAIMER: After publishing this blog post, I received forceful push-back on Facebook from my friend Scott Williams, a philosophical historian whose academic expertise is on the Trinity. He thinks my statements of the above models, particularly the first two, are overly simplistic to the point of representing straw-men. I agree that as historically developed both monarchial trinitarianism and Latin trinitarianism are considerably more complex than my simplified descriptions above indicate. But my goal here is not to present historical models, which are always complex and messy, and sometimes just plain incoherent, but only simplified conceptual models where the focus is on what is most fundamental in the Trinity.  

Side note #1: The term “trinity monotheism” was apparently coined by Brian Leftow in a paper that he originally published in 1999. Leftow uses it to label a model that he then critiques. J. P. Moreland and William Lane Craig defend the model in the first edition (2003) of their Philosophical Foundations for a Christian Worldview. According to their definition (p. 589), trinity monotheism “holds that while the persons of the Trinity are divine, it is the Trinity as a whole that is properly God.” That agrees well with my characterization, but I do not intend my use of the term here to line up precisely with either their account or Leftow’s. I’m simply using the phrase as a convenient label for the view that the Trinity as such is foundational.

Some challenges facing these models

(a) Monarchial trinitarianism

Monarchial trinitarianism is the official model of Eastern Orthodoxy and the Eastern Church generally, though the fully developed model as articulated by the Cappadocian fathers (Gregory of Nazianzus, Basil of Caesarea, and Gregory of Nyssa) is considerably more complex than my simplified description suggests. This model is arguably implicit in the Nicene Creed, the opening line of which (“I believe in God the Father almighty”) can easily be read as implying that God is (=) the Father. In any case, the fundamentality of the Father is deeply rooted in the Christian tradition and supported, explicitly or implicitly, by many venerable authorities.

The main challenge facing this model is that it conflicts with the idea that aseity (being foundational and thus underived) is an essential divine perfection. If aseity is an essential perfection, then only the Father (who, on this model, is foundational) qualifies as fully divine, and the Son and Spirit, who are ontologically derivative, and not fully divine. It follows that the Son and Spirit do not fully share in the Father’s divine essence and so the three are not homoousias (of the same nature), but at most homoiousias (of a similar nature). If this reasoning is correct, then monarchial trinitarianism is an incoherent model of the Trinity.

The standard response to this worry is to maintain that aseity is not an essential divine perfection and so not a natural divine property but only the personal property of the Father. The way this is usually argued is to say that aseity is a negative extrinsic property—the property of not being dependent on anything else—and so not an intrinsic perfection, one that a fully divine person must have in order to be fully divine. An extrinsic property is one that something has in virtue of how other things are, if there are any other things, and not in virtue of how that thing is in itself, i.e., intrinsically. Thus, for example, if I have the intrinsic property of thinking about President Trump, then he has the positive extrinsic property of being thought about by me. The property is extrinsic to him because it makes no intrinsic difference to him whether I’m thinking of him or not. Likewise, if I’m not thinking about President Trump, then he has the negative extrinsic property of not being thought about by me. This negative property would be true of him, however, even if I never existed. The thought then is that aseity is a negative extrinsic property. It’s true of the Father not in virtue of anything in the Father or in anything other than the Father but simply because the Father is the ultimate source.

I’m not persuaded by this response for two reasons.

First, if God is (=) the Father, then anything that is a necessary personal property of the Father is also an essential property of God. And since the Son and Spirit lack aseity on this model, it follows that they are not homoousias with the Father. One cannot reasonably reply that aseity is a merely contingent personal property of the Father like being Creator is because on this model aseity, i.e., being underivative, is precisely what makes the Father Father. It’s part of his personal essence and therefore, if God = the Father, it’s part of the divine essence as well. Ah, but is it the case that God = the Father? Given the fundamentality of the Father, I think this follows. Here’s why: If the Father is fundamental, then there is an explanatory moment where the Father exists and the Father alone. In that moment if we consider God and the Father with full concreteness, that is, without abstracting anything out of them like the Father’s personal properties or the divine nature, then it seems that the Father and God are numerically identical.

Second, aseity, rightly understood, is not merely a negative extrinsic property. This is because, to be God, a being must not merely not depend on anything else but must be foundational. The “a” in “aseity” is not an “alpha privative,” a negative prefix such as one find in words like asymmetric and atypical. It comes, rather, from the Latin prefix ab- meaning “from.” To have aseity is thus not merely not to exist from another but rather to exist from oneself (a se). This makes it a positive and intrinsic divine perfection. It means that God contains within himself (i.e., intrinsically) the ultimate sourcehood of everything that is or that ever could be. To be God is thus essentially to be the sole foundational being. If that’s right, then for the Son and Spirit to be fully divine and homoousias with the Father, they must be foundational just as the Father is.

Here are two addition problems for this model.

In the first place, causing to be seems to be an inherently temporal process. The reason for this is simple. If A causes B to be, then there is an explanatory moment in which A exists without B and then a subsequent explanatory moment in which B exists. These two moments are mutually incompatible because they cannot be co-realized since in the first moment B does not exist and in the second moment B does. Obviously B can’t both exist and not exist. So the explanatory sequence here cannot be “merely logical.” In a purely logical sequence successive terms have to be mutually compatible, as the premises and conclusion of a valid argument are. If the sequence cannot be merely logical, then it must be temporal. Because the two moment cannot be co-realized, they can only be realized in succession—in series, not in parallel. Hence, the Father’s causing-to-be of the Son and Spirit implies that the Son and Spirit are not co-eternal with the Father, contrary to what the model claims. As Arius famously put it, “there was a time when the Son was not.”

In the second place, the Father’s causing of the Son and Spirit to be is an agential faculty or “energy” that only the Father has. But, as Gregory of Nyssa argued in his letter to Ablabius titled “On Not Three Gods,” the reason why the three persons of the Trinity are not separate gods is because they share the same concrete instance of the divine nature and thus share the same divine faculties; they have a common divine mind, will, and agency. Peter, James, and John, in contrast, represent three concrete instances of human nature and thus each has his own mind, will, and agency.

But in the case of the Divine nature we do not similarly learn that the Father does anything by Himself in which the Son does not work conjointly, or again that the Son has any special operation apart from the Holy Spirit; but every operation which extends from God to the Creation, and is named according to our variable conceptions of it, has its origin from the Father, and proceeds through the Son, and is perfected in the Holy Spirit. For this reason the name derived from the operation is not divided with regard to the number of those who fulfil it, because the action of each concerning anything is not separate and peculiar, but whatever comes to pass, in reference either to the acts of His providence for us, or to the government and constitution of the universe, comes to pass by the action of the Three.

But note that Gregory restricts the scope of his argument to “every operation which extends from God to the Creation.” The restriction is mandated by monarchial trinitarianism, but it undermines the preceding sentence and his overall argument because it means that the Father does have faculties–namely the ability to generate the Son and Spirit–which the Son and Spirit lack. Their generation is something the Father does “by Himself.” So I don’t know how Gregory can have it both ways here. Either the Father has personal faculties that He doesn’t share with the Son and Spirit, in which case it’s not clear why it doesn’t follow that there is more than one “god,” or the Father, Son, and Spirit truly do share the same faculties, but then the Father can’t generate the Son and Spirit because He would need their co-operation to do so and they can’t co-operate in an operation that explains their own existence.

For the above reasons, I don’t think monarchial trinitarianism as standardly developed is a tenable approach to the Trinity. There is, however, a non-standard interpretation of monarchial trinitarianism—I’ll call it modified monarchianism—that may avoid some of the problems albeit while incurring some new challenges. The idea is this: In the first explanatory moment, God is unipersonal. In a subsequent explanatory moment, this unipersonal God self-differentiates into Father, Son, and Spirit. Let’s call the pre-differentiation unipersonal God “FATHER” and his post-differentiation correlate “Father.” In these terms, modified monarchianism affirms the monarchy of the FATHER but not the monarchy of the Father, as the two are not identical. (The FATHER is God by identity, but the Father is only God by predication.) Because this is an internal self-differentiation, the Son and Spirit are not created persons. Their origin is not ex nihilo but ex Deo and, indeed, ex PATRI. They eternally exist latently, as it were, in the “bosom” of the FATHER, and are then expressed or brought forth. This internal self-differentiation is not modalism because these personal distinctions are intrinsic to God, not extrinsic. Aseity proper belongs to the FATHER alone and not to the Father, Son, or Spirit considered individually. Since the Father, Son, and Spirit are on the same ontological plane, they have equal claim to divinity and are homoousias. Considered collectively, they constitute the FATHER and so in that sense are fully divine, sharing the FATHER‘s mind, will, and agency. So far modified monarchianism may sound reasonably promising, but the model also faces a number of questions that don’t have obvious answers. For example, why and how does the FATHER self-differentiate? Why doesn’t this self-differentiation result in a quaternity of FATHER, Father, Son, and Spirit? And could this self-differentiated trinity at some point re-coalesce? If not, why not? Because of questions like these I’m quite hesitant to endorse modified monarchianism even though I find it an intriguing proposal.

(b) Latin trinitarianism

Like monarchial trinitarianism this model has a substantial historical pedigree. Unlike monarchical trinitarianism this pedigree is mainly post-Nicene, having been introduced by Augustine in his book de Trinitate. Since then Latin trinitarianism has become the official position of the Roman Catholic Church, though as fully developed by various scholastics it is considerably more complex than my brief description suggests. As noted above, this model is predicated upon absolute divine simplicity, the view that everything in God is (=) God. So God = God’s essence = God’s attributes = God’s activities. Everything in God is just one big identity.

The obvious challenge for this model is how to get a robust Trinity from an absolutely simple God. If God = God’s essence, then the only way to find any differentiation in God is to find it in different relations of the divine essence to itself. Thus, for example, the identity relation (=) contains within itself three logically distinct relations: reflexivity (a=a), symmetry (if a=b, then b=a), and transitivity (if a=b and b=c, then a=c). Each of those relations presupposes the identity relation (=) to be stated, but they are not identical since we can imagine cases where, for example, symmetry holds but transitivity fails. (Case in point: person x has slept with person y. This seems to be symmetric, but is obviously not transitive. If x sleeps with y and y sleeps with z, it doesn’t follow that x sleeps with z.)

But here are three major problems with this model:

First, mere relations aren’t persons. Persons are subsistent individuals of a rational nature that can bear or stand in relations to other persons. To say, as Aquinas does, that the trinitarian persons are “subsistent relations” seems like an obvious category error. Persons have a categorical existence—they either exist or they don’t—not a merely relative existence. In their existing, persons relate, but their existence is not reducible to their relations. As Moreland and Craig put it, “on no reasonable understanding of person can a person be equated with a relation. Relations do not cause things, know truths or love people in the way the Bible says God does” (p. 586).

Second, this model seems to collapse into modalism, the view that the Father, Son, and Spirit are merely nominally or conceptually distinct. Thus, if everything in God is (=) God, then the distinctions between the Father, Son, and Spirit must be a merely extrinsic. The distinctions must be projected onto God from outside, as it were. They can’t be distinctions that exist intrinsically within God. This is because, if everything in God = God, and if these persons are each in God, then the Father = God = the Son = the Spirit. In short, given absolute divine simplicity there’s no way to maintain any stable distinctions within God. Like a black hole, the identity relation collapses everything into undifferentiated equality. Hence, the trinitarian distinctions collapse into unitarian modalism.

Third, this model is predicated on absolute divine simplicity, which is incompatible with Christianity on many different levels. Not only is it incompatible with orthodox Christology, but it leads to a dilemma both poles of which are disastrous for Christianity. Here’s the dilemma: Information about the specific details of creation is either intrinsic to God or not. On the first option, if the information is intrinsic to God, then by absolute simplicity that information is identical to God’s absolutely necessary divine nature. But then that information is absolutely necessary as well. Hence, we get a modal collapse: everything in creation is as absolutely necessary as God is. On the second option, information about the specific details of creation is not included in God at all. This allow us to preserve the contingency of creation but leads to a providential collapse. God is essentially indifferent to creation. Creation makes no intrinsic difference to God and therefore no difference to God. He does not and cannot know or care about creation. Finally, splitting the difference by making some information about creation intrinsic to God and some not doesn’t help. It just makes the intrinsic stuff absolutely necessary while leaving God absolutely indifferent to the extrinsic stuff.

For these reasons, I don’t think Latin trinitarianism is a tenable approach to the Trinity.

(c) Trinity monotheism

The other models in different ways attempt to derive the “three” (persons) from a more fundamental “one,” whether that be one of the persons (namely, the Father or perhaps the FATHER) or the singular divine essence. The strategy in each case is reductive: to explain the threeness we reduce it to a prior oneness. The third model rejects any reductive approach. The Trinity is a fundamental unity-in-difference, oneness-in-threeness, and it is a mistake to try to get behind that to a deeper, more fundamental unity. The one concrete instance of the divine nature is essentially tri-personal, and since each trinitarian person is fully divine, each exemplifies the complete divine nature and so implicates not only the one nature but the other two persons as well. In short, the one nature entails the three persons and each person in turn entails both the one nature and the other two persons in an irreducible 3-in-1 package deal. Whether you start from the common nature or one or more of the persons, you get the whole package.

Like the other models, however, this one also faces some challenges.

The main challenge is to avoid tri-theism. As developed by Moreland and Craig, Trinity monotheism is a kind of social trinitarianism on which “God is a soul … endowed with three complete sets of rational cognitive faculties, each sufficient for personhood” (p. 594) and thus has “three distinct centers of self-consciousness, each with its own proper intellect and will” (p. 583). They illustrate their view with an analogy to Cerberus, the three-headed canine guardian of Hades in Greek mythology (p. 593), and they argue that unity of Cerberus’s soul is sufficient to establish that there is only one dog in Cerberus despite its having three distinct doggie minds.

With respect to Moreland and Craig, I don’t think this does nearly enough to alleviate the tritheism worry. As Gregory of Nyssa argued in “On Not Three Gods,” the reason why the Father, Son, and Spirit do not constitute three gods is because they share one numerically identical set of divine faculties, one divine mind, one divine will, one agential or energetic faculty. The three persons necessarily always know, will, and act as one, not merely because (as on Moreland and Craig’s view) they have three distinct sets of faculties that necessarily functionally align, but because they share the numerically same faculties. I believe Gregory would accuse Moreland and Craig of tritheism. Their Godhead is a three-headed triumvirate of functionally distinct albeit necessarily co-operating “gods.”

In my judgment, trinity monotheists should drop the “three distinct centers of self-consciousness, each with its own proper intellect and will” idea. To avoid the charge of tri-theism, they need to emphasize that the three persons share numerically one concrete instance of what we might otherwise think of as a generic or common divine nature. The Father, Son, and Spirit are not related to each other as Peter, James, and John are. The latter three share a common generic human nature in three concrete instances. To secure the concrete unity of the divine persons, the trinitarian monotheist can’t merely posit—Cerberus style—three distinct personal consciousnesses each with their own mind and will as constituents within a single “soul.” Rather, they need to make a plausible case that the one concrete divine nature requires three persons and that these persons are, necessarily and perichoretically united together, such that each of the persons requires the other two persons. For this I think we need to have, as Gregory would contend, shared faculties of intellect, will, and agency (energy). This ensures that they cannot know, will, or act independently and so are one God, not three gods.

Another reason to drop the “three distinct centers of self-consciousness” talk is that it’s not clear how there can even be three such distinct centers in God. Given divine omniscience, the perspectives of the three persons must necessarily coincide. Each of them, as omniscient, must have the singular, all-encompassing God’s-eye perspective on reality. This fits well with Gregory’s insistence that they share a common divine mind, but not with Moreland and Craig’s version of social trinitarianism. The most one can say is that Father uniquely knows “I am the Father,” the Son uniquely knows “I am the Son,” and the Spirit uniquely knows “I am the Spirit,” while sharing of all their non-first-person knowledge. Philosophers refer to uniquely first-person knowledge as de se knowledge. Perhaps that’s enough to distinguish the three persons. And maybe we can equate this de se knowledge with the unique personal properties of each person. But once we set de se knowledge aside it seems awfully redundant to think that they need three sets of mental faculties for their shared knowledge.

Relatedly, there is also a challenge of how to distinguish between the three persons. On certain versions of trinity monotheism it looks like the three persons are interchangeable with each other. Could the person that we call “Father” have been the “Son” or “Spirit” instead? What is the significance of these personal labels? Are those labels grounded in the immanent Trinity, i.e., the Trinity considered from the inside? Monarchial trinitarianism has a ready answer to this: what distinguishes the persons are causal relations of origin that eternally obtain between them. The trinitarian monotheist can’t give that answer. Because he takes the persons to be equally foundational, he can’t posit relations of origin within the immanent Trinity. He can, however, associate the personal labels with the “economy,” i.e., how God relates to creation. Thus, broadly speaking, the Father is the person who remains transcendent and non-localized. He anchors the transcendent Godhood of God in relation to creation.

Side note #2: That the Father anchors the transcendent Godhood of God explains why the Bible often refers to the Father as ho theos (“the God”) and why Jesus refers to the Father as “the only true God” (John 17:3). Jesus is not not excluding himself from the Godhead but, as Cyril of Alexandria puts it, “insofar as this true and natural Son became as we are, so he speaks of Father as his God, a language fitting to his self-emptying.” (On the Unity of Christ, p. 63)

The Son, in contrast, is the person who becomes immanent and localized in theophanies and ultimately in the incarnation. He is God with us, the mediator between the Father and creation. And, finally, the Spirit is the person who becomes immanent but remains non-localized so as to be God omnipresent throughout creation. He mediates all of creation to the incarnate, localized Son. Of course, all that could be right without addressing the immanent trinitarian distinctions at all. So the trinity monotheist either has to “bite the bullet” and admit that the persons are immanently and economically interchangeable—e.g., the Father could have become incarnate while the Son remained transcendent, in which case we would have called the Father “Son” and the Son “Father”—or provide some additional reason independent of the economy for thinking that the labels “Father,” “Son,” and “Spirit” map onto the immanent Trinity in a non-arbitrary way. Either way, this doesn’t strike me as a “deal-breaker” for the model, but it is an outstanding challenge.

On the whole, I think some version of trinity monotheism is probably right, but I don’t think the Moreland and Craig version of trinity monotheism is tenable. It doesn’t succeed in avoiding tri-theism. I sketch out a better version of trinity monotheism in the second-half of this post, based on the idea that God is essentially perfect love (1 John 4:8, 16) and that perfect love is a structured triadic relation.