Classical theism and the Trinity

    A couple of years ago, I wrote a post surveying many of the major Trinity theories in modern analytic theology, in which I argued that all of them end up being problematic in terms of both philosophy and Scripture. I also wrote another post in which I argued that the Trinity is an incoherent doctrine, since it’s impossible to logically parse the metaphysical language of “one ousia and three hypostaseis” in a way consistent with all the tenets of trinitarianism. However, after delving into the metaphysics of classical theism, I think there is a way to coherently model the Trinity along the lines of the interpretation that’s been common in the Western tradition since Augustine. In this post, I’ll be presenting this coherent Trinity theory and defending it against objections.

(Note: Just because the Trinity can be coherent doesn’t necessarily mean that it’s true; many coherent philosophical views are clearly false, such as solipsism. But both unitarians and trinitarians should be intellectually honest and interpret the other’s doctrine in the strongest possible way.)

Defining orthodox trinitarianism

    First, let’s look at the axioms that define an orthodox Trinity theory (OTT). Any OTT must be at least minimally consistent with the following ten claims:

1. There is one God.

2. The Father is God.

3. The Son is God.

4. The Holy Spirit is God.

5. The Father is not the Son.

6. The Father is not the Holy Spirit.

7. The Son is not the Father.

8. The Son is not the Holy Spirit.

9. The Holy Spirit is not the Father.

10. The Holy Spirit is not the Son.

In other words: there is one God, and there are three persons that are each individually the one God but are not each the other persons. However, the meaning of these axioms is disputed, especially the meaning of “God” and “is.” When we say, “There is one God,” what exactly are we saying that there is one of? And when we say that each of the persons individually “is” God, what relation are we referring to?

    The standard problem here is that if “is” is interpreted as a relation of numerical identity, then the ten axioms presented above are contradictory. Numerical identity is a transitive and symmetrical relation, which means that if A = C and B = C then A = B. (“Transitive” means that if A = B and B = C, then A = C. “Symmetrical” means that if A = B, then B = A.) But this means that if the Father = God (axiom 2) and the Son = God (axiom 3) then the Father = the Son (contradicting axiom 5). Therefore, if the ten axioms are self-consistent, “is” must refer to a relation that is non-transitive and/or asymmetrical. Each Trinity theory seeks to solve this Problem in its own way.

One and three-self OTTs

     In modern analytic theology, OTTs are primarily divided into “one self” and “three self” theories. (See the Stanford Encyclopedia of Philosophy article on the Trinity.) From that SEP article, a self is defined as “a being who is in principle capable of knowledge, intentional action, and interpersonal relationships.” One-self OTTs seek to show that the ten axioms of trinitarianism are consistent with God being a single self, while three-self OTTs seek to show that the axioms are consistent with God being and/or having three selves.

    One-self OTTs are clearly consistent with the first axiom: “one God” refers to the single great Self which is God. It gets a bit dicey with the other axioms. In order to solve the Problem, one-self OTTs typically interpret the Father, Son, and Holy Spirit as modes of God: “God-qua-Father,” “God-qua-Son,” and “God-qua-Spirit.” This seems to solve the Problem (there is a sense in which modes of a self, such as “me-qua-son” and “me-qua-student” are the same as me but not the same as each other). But are these modes essential to God or accidental to him? If they’re essential to him, their existence appears to be a brute (unexplained) fact, since nothing can explain its own essence and nothing outside of God can explain his essence either. And if these three modes are merely accidental, extrinsic ways that God relates to us, this may be nominally consistent with the ten axioms, but historically it’s been seen as “modalism” which is not authentically trinitarian.

    Three-self OTTs are consistent with axioms 5-10, since different selves clearly aren’t the same as one another. But they run into a problem with axioms 1-4. In order to say that three numerically distinct selves are each individually “God,” three-self OTTs typically interpret “one God” (axiom 1) to refer to a single universal “God-ness” which the three persons instantiate. But in other cases where there are multiple instantiations of a single universal (e.g., “dog-ness”) we say that there are multiple things (“dogs”). It’s a very unusual use of language to say that three instances of “God-ness” is one “God.” And if “God” does refer to “God-ness,” it’s strange to say that each person is “God-ness” – we wouldn’t say that every dog is “dog-ness.”

    Other three-self OTTs, rather than relying on an unusual use of language, rely on an unusual system of logic or an unusual theology. For example, relative identity OTTs use the concept of ‘relative identity,’ according to which the three persons can be “identical-qua-God” but not “identical-qua-persons.” Most philosophers deny that this concept is even coherent. William Lane Craig has an OTT wherein “God” refers to the Trinity itself, and the three persons are proper parts of God. Most theologians consider this unorthodox, and it relies on an unusual mereology (metaphysics of parts) that allows us to say that a proper part “is” the whole, which seems contrary to the very concept of proper parts (parts that aren’t identical to the whole).

Monarchical trinitarianism

    Another OTT is the monarchical view of the Trinity, which is a common Eastern Orthodox view. This view has been defended analytically by Beau Branson (2022), and is also argued by Metropolitan John Zizioulas, Fr. John Behr, and other prominent Eastern Orthodox figures in various places. According to monarchical trinitarianism, “God” properly refers to the Father because he is the One who is absolutely uncaused and unoriginate, while the Son and the Spirit are “God” insofar as they have the same nature as the Father but not as uncaused and unoriginate (since they originate from the Father). This is argued to be more consistent with the 4th-century creeds that proclaim “one God, the Father,” and also to be the view that the Cappadocian Fathers held (e.g., Gregory of Nyssa, Ad Petrum 4; Contra Eunomium 2.5). Basil of Caesarea even declared, “There is one God because there is one Father” (Contra Sab. et al. 3-4).

    This view is obviously consistent with the first axiom: “There is one God” is true because, as Basil said, there is one Father and the Father is numerically identical to God. It’s also consistent with axioms 5-10 since the Son and the Spirit are different selves than the Father (like the three-self OTTs), and with axiom 2 since the Father is numerically identical to God. But it’s harder to see how it can be consistent with axioms 3 and 4, that the Son is God and the Spirit is God. Even though they have the same nature as the Father, “God” properly refers to the Father as uncaused and unoriginate, and the Son and the Spirit “are” God in a different sense than the Father “is” God. This solves the Problem at the expense of interpreting “is” differently between axioms 2 and 3-4.

Classical theism and the Trinity

    All of the OTTs surveyed above manage to interpret the ten axioms in a way that solves the Problem, but each of them comes with significant theological and/or philosophical costs. What if we add the classical theistic doctrines that God is simple (without proper parts) and purely actual? At first, this seems to make the Problem worse, since it means God is really identical to everything intrinsic to him. But if we consider this more closely, it might actually give the trinitarian a way out of the dilemma and a coherent model of the Trinity that combines all the best aspects of the other OTTs.

    First, let’s consider what it means for God to know himself. Humans are made of multiple parts – at the very least, our nature (humanity) and the matter that individuates us from other humans. Our act of thinking is separate from both of these things. So when we think about ourselves, there are at least three things: our matter, our nature, and our act of thinking which is related to our nature. If non-material beings (like angels) exist, then their thinking about themselves is slightly more simple: it only requires their nature and their act of thinking. But when God knows himself, because he is absolutely simple, there are no parts: just God’s nature which is related to itself. His act of thinking is really the same as his nature. God is “thought thinking itself,” as Aristotle said (Metaphysics 12.7, 9).

Figure 1.
 A representation of a human knowing themself (A); an angel knowing themself (B); and God knowing himself (C). “M” refers to matter, “N” refers to nature, “AT” refers to an act of thinking, and “G” refers to God. Based on figs. 1-4 of Koons (2018).

    Thus, if classical theism is right about God’s simplicity and pure actuality, he must fundamentally be a relational act directed toward himself. (He certainly can’t be a relational act directed toward anything else, since he exists a se, self-sufficiently.) This can be understood as God thinking himself, knowing himself, loving himself... if he’s simple, then all these acts are really the same in him. But if God is a binary relation of knowing and loving directed toward himself, we can distinguish three different relational properties here: being lover, being beloved, and being lover and beloved. This in turn gives rise to three of what are called “qua-objects” in modern analytic philosophy: God-qua-lover, God-qua-beloved, and God-qua-lover-and-beloved (Koons 2018, 344-47).

    The classical theistic OTT identifies these qua-objects with the three persons of the Trinity (God-qua-lover = the Father; God-qua-beloved = the Son; God-qua-lover-and-beloved = the Spirit). This is consistent with the first axiom, “There is one God,” because classical theism affirms that God is one and absolutely simple. It’s also consistent with axioms 2-4, since each qua-object is really the same as God. (It couldn’t be otherwise; there are no proper parts of God such that a qua-object could be only part of him.) But each qua-object isn’t the same as the others (axioms 5-10), because each one has a relational property that the others don’t, namely, being lover, being beloved, or being lover and beloved.

    This model of the Trinity, like the one-self OTTs, is able to use qua language to describe the persons of the Trinity. However, it’s able to avoid the dilemma of making the persons either accidental or a brute fact, since in this model the persons arise out of God’s relational nature. Like three-self OTTs, this model affirms the distinctness of the persons, without relying on any unusual use of language. Along with monarchical trinitarianism, it’s able to affirm the processions of the Son and the Spirit from the Father (unlike some Protestant OTTs), and actually relies on these relations, thereby preserving the Father’s place as the only uncaused and unoriginate. But unlike monarchical trinitarianism, it’s able to say that the Father, Son, and Spirit “are” God in the same sense (really the same as the divine nature).

    This view is arguably the one that Western thinkers like Augustine and Aquinas held, or at least, it’s very similar to theirs. Augustine held that the Trinity is analogous to “the mind, and the knowledge by which it knows itself, and the love by which it loves itself” (De Trinitate 9). “[W]hen the mind knows itself and loves itself, there remains a trinity: mind, love, knowledge” (De Trinitate 9.5). Aquinas argued that the persons of the Trinity are what he called “subsistent relations” (ST 1.40.1-2), which we would call qua-objects, and he associated them with the intrinsic relationality of God’s nature (ST 1.27-28).

Objections to the model

    A few objections could be raised against this model of the Trinity. First, no qua-object is numerically identical to its basis, so in what sense can we say that each of the persons “is” God? We already know because of the Problem that “is” can’t be interpreted as numerical identity. Robert Koons (2018, 347-49) came up with a relation of “real distinctness,” according to which qua-objects aren’t “really distinct” from their basis, but different qua-objects are “really distinct.” This could be seen as an ad hoc solution, but I think it’s quite intuitive: it seems that there must be a relation such that my qua-objects (e.g., “me-qua-son” and “me-qua-student”) are really the same as me but not the same as each other. I really am a son and a student, but my son-ness and student-ness aren’t the same. In an analogous way, God could be Father and Son and Spirit, but his Father-ness and Son-ness and Spirit-ness are distinct.

    Another objection: how could it be that God’s qua-objects are different persons but ours aren’t? I’m a son and a student, but my son-ness and student-ness can’t relate to each other as persons. The difference is that God’s qua-objects are intrinsic and essential to him, but my son-ness and student-ness are based on my relations to extrinsic things. Furthermore, God’s qua-objects are fully God, since he is simple without distinct parts and attributes, whereas our qua-objects refer to individual aspects of our composite being. Even if we think about our act of thinking, this doesn’t make us tripersonal, because this thinking is merely accidental to us and (as composite beings) our acts of thinking are distinct from our nature.

    A third objection: if the Father is God-qua-lover and the Son is God-qua-beloved, does this mean that the Father isn’t loved or that the Son doesn’t love? No, because God is simple and his qua-objects share all of his essential attributes, including loving God and being loved by God (Koons 2018, 346). However, each does so in their own way. For example, God-qua-beloved loves-qua-beloved God, but he doesn’t love-qua-lover nor does he love-qua-lover-and-beloved. As long as the relation of love within God is both irreflexive and symmetric, the distinctness of the persons and their love for each other is preserved.

    Finally, how could it be that there are only three qua-objects that are persons? If we say that God-qua-lover and God-qua-beloved are persons, why aren’t God-qua-omniscient and God-qua-creator also persons? First of all, the reason why God-qua-lover and God-qua-beloved could be seen as persons is because they describe the relata of an irreflexive relation. (To love and to be loved are conceptually distinct, even in the case of self-love, which makes this relation irreflexive.) To be omniscient isn’t a relation, much less an irreflexive relation, so God-qua-omniscient is just numerically identical to God. To be creator is a relation, but it’s a relation to something extrinsic to God, so God-qua-creator is also just identical to God. A simple God has only one intrinsic, binary, irreflexive relation, which logically leads to exactly three qua-objects that can be considered as persons.

xkcd 2591 “Qua”

Conclusion

    The axioms of orthodox trinitarianism appear to entail a contradiction, if “is” is interpreted as strict identity. There are various Trinity theories that modern analytic theologians have put forth to solve this problem, all of which come with their own difficulties. But if God is simple, then everything intrinsic to him is really the same as him, which makes it easier to understand how there could be multiple persons that are God. And if God is simple and intrinsically relational, as classical theism affirms (“thought thinking itself,” or in more properly Christian terms, “love loving itself”), then we can construct a logically coherent model of the Trinity.

No comments:

Post a Comment

Resurrection and metaphysical anthropology (part 1 of 2)

    Is bodily resurrection something desirable, or even a metaphysical possibility? According to nearly all Christians, it is. But this view...