Not being a historian of medieval philosophy, I can’t vouch for the historical accuracy of Denys Turner’s account of Thomas Aquinas’s thought. But as a philosopher appreciating careful and creative thinking about Christian teachings, I can and do recommend Turner’s book as an impressive achievement. His Aquinas is a lucid and rigorous philosopher, following every argument wherever it leads, showing enormous ingenuity and unflinching integrity. Unfortunately, Turner’s discussion also shows—though he wouldn’t agree—that in most cases the final result is not deeper understanding but unintelligibility.

Given the space available, I’ll focus on just three key Christian doctrines where Aquinas’s arguments lead to perplexing conclusions: immortality, creation, and the nature of God as both one and triune. Turner covers much else, including stimulating discussions of Aquinas on grace, divine love, the Incarnation, and transubstantiation, as well as an ingenious view of how Aquinas the philosopher and theologian relates to Aquinas the saint. But, throughout, the specter of unintelligibility—or, if you like, “mystery”—looms large.

Turner emphasizes that there is an important sense in which Aquinas’s view of human beings is materialist. It is not a reductive materialism: Aquinas doesn’t claim, for example, that consciousness itself is nothing but a flux of neurons. Aquinas insists that there are irreducibly nonmaterial aspects of our existence. But he rejects the idea that a human being is a soul, only contingently connected to a body; rather, our identity as persons requires embodiment. The “soul” is an Aristotelian form, animating the body, not a Platonic substance temporarily “inhabiting” the body. The person I am is an animal and must have a body.

Aquinas does have an argument that he thinks shows that the soul is immortal and so will exist forever after death. But, as Turner points out, “Thomas says that even though human souls survive death, souls thus separated from their bodies are not persons.” How could they be, since a person must be embodied? Therefore, according to Aquinas, since I am a person, I do not survive my death. When my body dies, I cease to exist.

How can this be consistent with the Christian promise of eternal life? Aquinas’s answer is that eternal life is achieved through the body’s resurrection and reunion with the soul, which has survived continuously, but not as me—or any other person. Showing that this answer is even possible requires some difficult threading of philosophical needles. First, of course, Aquinas must develop a cogent argument that the soul is immortal, a particularly difficult matter, since he has defined it as not a substance in its own right but the life-principle of an animal body.

Aquinas, however, maintains that, despite its intimate connection to a material body, the human soul is immaterial by virtue of its capacity for knowing material things. The nub of his argument is that an act of knowing must itself be “neutral” regarding the object known. For example, an act of hearing a sound must itself be soundless (otherwise, its own sound would interfere with the sound it is hearing). Similarly, an act of seeing a color must itself be colorless. Accordingly, an act of knowing a material thing must itself be immaterial.

The argument is not compelling, not least because of its questionable analogy of intellectual knowing with sense perception. But even if we grant the soul’s immateriality, its strong dependence on the body still suggests that it will not be able to exist without a body. Immateriality does not imply immortality. Further, even if the soul is immortal, does it make any sense to say that the person resulting from the strange process of resurrection could actually be me? If we ask why I’m the same person as the infant born years ago, there are two sorts of answers. One refers to the body: my present body results from a continuous series of changes beginning with the infant’s body. Another answer refers to psychological states: my current psychological state (memories, feelings, etc.) can be traced back through a continuous series of psychological states to those of the infant.

There’s room for considerable philosophical disagreement about which of these conditions—or which combination of them—makes for personal identity. But Aquinas’s account of resurrection violates both of them. There is neither bodily continuity nor psychological continuity between me and the resurrected person. Aquinas is right that a Platonic identification of the human person with an immaterial soul does not fit well with what we know, philosophically and scientifically, about human beings. But the Platonic view makes the doctrine of immortality relatively easy to understand, whereas, at best, Aquinas’s more materialist Aristotelian view makes it extremely difficult to understand.

The problem of understanding is even more difficult when we turn to the claim that God is creator of all things, a claim Aquinas thinks he has established by natural reason. The basis of the claim, especially as developed by his Third Way in the Summa, is that every contingent thing (any thing that might not exist) comes into or remains in existence only through God’s creative power. A first problem is that, as Turner deftly shows, this view of creation conflicts with one common response to the argument from evil against the existence of God. That argument claims that an all-good, all-powerful God would have been both able and willing to eliminate evil from the created world. The free-will defense responds that this need not be so, since creatures (like ourselves) capable of free choices are a great good, but one that may lead to freely chosen evil actions. God could, of course, cause all of our actions to be good ones, but then these actions wouldn’t be free. It may, therefore, be that the value of free choice offsets the bad results of evil actions, so that God could not have this good in his creation without allowing the creatures to make evil choices. This shows the consistency of a divine creator with the reality of evil.

But, according to Turner, for Aquinas the free-will defense is a nonstarter. Everything, including our free choices, must be caused by God’s creative power, whereas the free-will defense assumes that if an action is caused by God it isn’t free. As Turner puts it: “It is too easily taken for granted that to speak of God causing my free actions is necessarily a contradiction. For Thomas far from it: worse, to say my actions are free only insofar as God does not cause them presupposes a plainly idolatrous conception of the divine causality.” Rather, for Aquinas, “our free actions are the direct creation of the divine will.”

It’s not just that we lose one of the main responses to the problem of evil. More important, on Aquinas’s view, our concept of causality collapses when it is applied to divine creation. An action directly and entirely caused by another agent is by definition not my action. You can say the words “God is the total and direct cause of my free action,” but in this claim both “free” and “cause” lose contact with what those terms mean. If the words are not contradictory, they are meaningless.

Finally, there is Aquinas’s account of the nature of God. Here the discussion takes two directions: an analysis of the truth, knowable by natural reason, that God is one; and an analysis of the truth, knowable only by revelation, that God is triune.

The oneness of God would pose no special problem if it meant that God is the only one of his kind. But Aquinas denies this formulation on the grounds that God does not belong to any kind—as, for example, an eagle is a kind of bird. According to Aquinas, this is because if God did belong to, say, the kind divine being, then there could be other instances of this kind—since, as a matter of logic, it’s always possible that there could be many instances of a kind. But there doesn’t just happen to be one God; there cannot be more than one. (One way to show this might be to note that God is nothing but total perfection, so that for a being to differ from the one God would mean that it was imperfect and so not a divine being.) It follows that God does not belong to any kind.

This result alone is a major obstacle to any real understanding of God. For it means that when we say God is “wise,” is “good,” or “loves” us, this cannot mean that he is wise, good, or loving in the same sense that, for example, we are. It is not a matter of his having, even to a maximal degree, the kind of feature a human being who is wise, etc., has. The terms we use to describe God cannot mean what we mean by them.

Aquinas’s response is to say that in speaking of God we use these terms “analogously.” But analogies are useful only because we can say in what respects they hold and in what respects they don’t. If I say, “my love is like a rose,” I don’t mean that she has thorns and grows in the ground, but that, for example, she has a certain sort of delicate beauty. This, however, implies that she is an instance of the kind beauty. This can’t be what we mean when we say God is good or beautiful, since God belongs to no kind. So it remains fundamentally unclear what our talk about God means. (Turner’s failure to discuss Aquinas on analogy is one point on which I would fault his excellent book.)

But Turner does discuss in detail another problem that arises from Aquinas’s understanding of God’s oneness: showing that there is no contradiction in the central Christian doctrine of the Trinity. According to the doctrine, there are three divine persons, the Father, the Son, and the Holy Spirit, but there is only one God. In other words, each person fully possesses all the attributes of God, but still there is only one God. How can this be? If there are three persons and each of them is fully God, how can there not be three Gods?

The standard formula to avoid contradiction is that the three divine persons “share” the same divine nature. But this won’t work because it amounts to saying that the divine nature corresponds to a kind of being in which Father, Son, and Spirit share. Thus, the kind divine being would have these three instances. This can’t be true, for two reasons: God, as we have seen, cannot belong to a kind; and, if there were three beings that were instances of the kind divine being, then there would be three Gods, not one.

According to Turner, Aquinas (like Augustine) “saw that what gets in the way of Trinitarian orthodoxy is the troublesome word ‘person.’” As long as we take “person” to have the standard meaning of “an individual substance of a rational nature” (to use Boethius’s classic formulation), the claim that three persons are God must imply that there are three Gods. To avoid contradiction, says Aquinas, we must change the meaning of “person” when we apply the term to God.

In particular, we need to make the astonishing assertion that a divine person is not a substance (an independently existing individual) but a relation. Instead of saying, as the traditional theology does, that the Father generates the Son, we must say that the Father is the relation of generating. And rather than saying that the Son is generated by the Father, we must say that the Son is the relation of being generated. We might be able to entertain the idea that a person could be a relation. We might, for example, coherently think that the person I am is the complex relation composed of all the relations among the different temporal stages of my existence. But here we still have my various temporal stages as the things that are related to one another (the relata). In the case of God, what are the relata?

The only possibility is that the persons of the Trinity themselves are the relata (introducing anything else as relata will make the divine persons derivative from—and dependent on—something else). But then we have slipped back into the polytheism of three different Gods, each a subsisting individual related to the other two. This is why Aquinas must take the radical step of claiming that the divine persons are relations that do not relate anything. Turner puts it this way: “there is nothing here but the relatings, no somewhats doing the relating.” He goes on to say, “The language strains,” but he also suggests that, although “bent and twisted,” the language does not break: it allows Aquinas to avoid “gross inconsistency” in asserting the doctrine of the Trinity.

It may be that the bending and twisting of language avoids inconsistency. But at best it does so by leaving the doctrine of the Trinity with no meaning at all. As Turner also says, “the meanings of ‘person’ and ‘threeness’ [and, I would add, ‘relation’] migrate off the semantic map of our secular vocabularies” and into a realm of “utter unknowability” or “unutterable mystery.” Turner insists that “for Thomas, the doctrine of the Trinity is not incoherent nonsense.” But, although it’s not random babbling, it provides no meaning that removes the apparent contradiction of the doctrine. How could it, since, as Turner says, it presents the doctrine as “utterly unknowable”? If Aquinas had shown that the doctrine is not self-contradictory, then we would at least know that much about it.

Turner, however, tries to celebrate this bleak result. Speaking not only of the doctrines I’ve discussed but also of Christian teachings on Christology and the Eucharist, he says, “It is characteristic of the logic of these core doctrines that theirs is a vocabulary, a way of speaking of God, that, in the end, witnesses to its own depletion.” Theological language begins from concrete experiences (moral choices, love for Jesus, receiving the Sacraments), but at the “last moment” when that language “reaches into God” it loses “that grip on its object that the secure logic of affirmation and negation” provides. At this point, Turner turns to Wittgensteinian language: “For Thomas the ladder of language that ascends to God rests securely on its foundations in our human worldly experience; but the ladder being climbed, God is reached only when it is ultimately cast away: and that is the ‘mystical,’ as Wittgenstein somewhat opaquely puts it.” According to Turner, this is a positive development. It is a “conceptual ground-clearing” that opens up space for a “positive theology” that, “no longer entangled in the logician’s dilemma,” can move unimpeded to an “understanding of the relation between creature and Creator.”

Here, however, Turner fails to face up to the challenge to theological language that his brilliant exposition of Thomas’s thought poses. For, as we have seen in the three examples of immortality, creation, and the divine nature, what Aquinas has “cleared away” are not the obstacles to a theological understanding of these doctrines but rather the very possibility of such an understanding. He leaves us with an analysis that avoids self-contradiction only by taking us beyond the limits of meaning, unable to judge whether resurrection, creation, and the Trinity even make enough sense to be judged true or false.

The conclusion to draw from this is not that we should simply reject the doctrines. The solution lies rather in the principle of the unity of truth, which Aquinas asserts as a general proposition but fails to apply in the case of the “mysteries” of faith. According to the principle, when two apparent truths conflict, they cannot both be correct, but there is no reason that an apparently revealed truth must trump an apparent truth of natural reason. A well-grounded truth of reason may require us to modify or reject what we had thought was a revealed truth. If we find, as he did, that careful formulation of what we think a revealed truth means leads to absurdity, we should then conclude that the truth does not mean what we thought it did. The task then is to reformulate, not to try to convince ourselves that our failure to understand is a higher form of understanding.