Hostname: page-component-74656f9884-npj9w Total loading time: 0 Render date: 2026-11-08T08:35:21.612Z Has data issue: false hasContentIssue false
  • English
  • Français

Quid tres? On What Precisely Augustine Professes Not to Understand in De Trinitate 5 and 7

Published online by Cambridge University Press:  22 June 2007

Richard Cross
Richard Cross
Affiliation:
Oriel College, Oxford
Rights & Permissions [Opens in a new window]

Extract

It is generally held by systematic theologians that Augustine more or less radically shifts the understanding of the doctrine of the Trinity found in his Greek sources, and that, at least in part, this shift is a consequence of his failure fully to understand these sources. His own testimony of puzzlement provides some strong prima facie evidence in favor of such readings.

Information

Type
ARTICLES
Information
Harvard Theological Review , Volume 100 , Issue 2 , April 2007 , pp. 215 - 232
Copyright
© 2007 Cambridge University Press

It is generally held by systematic theologians that Augustine more or less radically shifts the understanding of the doctrine of the Trinity found in his Greek sources, and that, at least in part, this shift is a consequence of his failure fully to understand these sources. His own testimony of puzzlement provides some strong prima facie evidence in favor of such readings. Colin Gunton comments as follows, both on Augustine's apparent agnosticism and on his consequent misunderstanding of the Greek achievement:

It is difficult for [Augustine] to understand the meaning of the Greek hypostasis. One reason is that he can make nothing of the distinction so central to Cappadocian ontology between ousia and hypostasis: “I do not know what distinction they wish to make” (V.10). Certainly, it is unfair to say that he gets nothing of the point at all, for he goes on to say that, in view of the difficulty of translating the Greek terms into Latin, he prefers to say, with his Latin tradition, unam essentiam or substantiam and tres personas. Augustine at least realizes that different concepts are required if we are to express the distinction between the way in which God is one and the way in which he is three. It becomes clear, however, that the adoption of the correct Latin equivalents does not enable him to get the point, for, in a famous statement, he admits that he does not really see why the term should be used. ‘Dictum est tamen tres personae non ut illud diceretur sed ne taceretur’ (V.10, cf. VII.7: “this formula was decided upon, in order that we might be able to give some kind of answer when we were asked, what are the three”). … Moreover, Augustine reveals that he is unaware of what is going on when he makes it appear to be merely a matter of linguistic usage (forte secundum linguae suae consuetudinem, VII.11).1

Colin Gunton, The Promise of Trinitarian Theology (Edinburgh: T&T Clark, 1991) 39–40.

In this short piece, I want to ask more precisely what it is that Augustine professes not to understand. Answering this question involves some close textual reading of a somewhat dense kind, reading which is fundamentally logical in nature (as Augustine's text is at the moment when it raises the sort of question Gunton highlights). But the answer is important, since readings such as that offered by Colin Gunton have ecumenical consequences—that the Western tradition, thanks to Augustine's failures, fundamentally misunderstands genuine (i.e., Greek) trinitarian theology. Taking steps towards showing the contrary would thus be a matter of great ecclesiastical moment.2

Literature rebutting alleged divergences between Eastern and Western views of the Trinity is becoming increasingly widespread: see for example the various different approaches to the topic in Rowan Williams, “Sapientia and the Trinity: Reflections on De Trinitate,” Collectanea Augustiniana. Mélanges T. J. van Bavel (ed. B. Bruning, M. Lamberigts, and J. van Houlin; Leuven: Leuven University Press, 1990) 317–22; Michel René Barnes, “De Régnon Reconsidered,” Augustinian Studies 26 (1995) 51–79; Richard Cross, “Two Models of the Trinity?” Heythrop Journal 43 (2002) 275–94; David Bentley Hart, “The Mirror of the Infinite: Gregory of Nyssa on the Vestigia Trinitatis,” Re-Thinking Gregory of Nyssa (ed. Sarah Coakley; Malden, Mass.: Blackwell, 2003) 111–31 (many of the essays in Coakley's valuable collection can be read profitably in this context); Lewis Ayres, Nicaea and Its Legacy: An Approach to Fourth-Century Trinitarian Theology (Oxford: Oxford University Press, 2004) chs. 14 and 15. The best refutation of the fundamental historiographical paradigm upon which the older analyses tended to rely is André de Halleux, “Personalisme ou essentialisme trinitaire chez les péres cappadociens,” Revue théologique de Louvain 17 (1986) 129–55, 265–92. As far as I know, none of this literature includes a careful logical reading of the relevant portions of Trin. of the kind that I attempt here.

To try to see what Augustine professes not to understand, we shall have to clarify what precisely Augustine attempts to teach in the relevant parts of books 5 and (especially) 7 of De Trinitate [hereafter, Trin.]. In passing, I hope to provide some preliminary assessment of the wider issue of whether in fact Augustine misunderstood the distinction or distinctions between

and

that can reasonably be thought to have been proposed by the Greek theologians of the late fourth century. But the judgment will be provisional, in the sense that I do not have the space here even to begin an adequate treatment of the fourth-century Greek trinitarian tradition or traditions.

What Augustine Professes Not to Understand

In Trin. 5.8.9 Augustine sets up a problem that he leaves unresolved until the middle of book 7. The problem has to do with divine simplicity and its relation to the use of the term “person” in God.

3

I owe my understanding of the nature of the problem in book 5 of Trin. entirely to Joseph Jedwab—particularly the insight that “person” is not a relational term. He disagrees strongly with my account of Augustine's solution to the problem, which I outline in section 3 below.

According to the doctrine of divine simplicity,

Let us hold this above all, that whatever is said of that most eminent and divine greatness is said substantially. But whatever [is said] “to another” (ad aliquid) is said not substantially but relationally. “Of the same substance” is of such force in the Father, Son and Holy Spirit that whatever in the highest nature is said of each with respect to themselves is said not plurally but singly. For just as the Father is God, the Son is God, and the Holy Spirit is God, which no one doubts is said according to substance, nevertheless we do not say that the most excellent Trinity is three gods but one God.4

Augustine, De Trinitate 7.8.9, 1–6 (ed. W. J. Mountain; 2 vols; Corpus Christianorum: Series latina 50, 50A; Turnhout: Brepols, 1968) 1:215.

Through the rest of 5.8.9, Augustine goes on to apply this rule to “great,” “good” and a range of other non-relational divine attributes. In 5.8.10, Augustine considers the Greek term

. The concern, presumably, is that, like “God,” “great” and “good,”

is not a relational term, and Augustine goes on to make the same point about “person”:

They indeed say hypostasis, but I do not know what they intend to be the difference between ousia and hypostasis, such that most of us who treat these things in the Greek language are accustomed to say “one ousia, three hypostases,” which is “one essence, three substances (unam essentiam, tres substantias)” in Latin. But because our customary [Latin] usage has already obtained, [namely] that when we say “essence” is understood what is understood when we say “substance,” we do not dare to say “one essence, three substances,” but “one essence or substance”. … When it is asked “what three (quid tres),” human language straightforwardly labors under great lack. “Three persons” is said, not that it might be said, but that it might not fail to be said.5

Trin. 5.8.10, 43–51; 5.9.10, 7–11 (1:216–17). From now on I use “essence” for the Latin term essentia, “substance” for substantia, and “person” for persona.

The proper names of Father, Son, and Spirit are said relationally.6

Trin. 5.11.12 (1:218–20).

But the term “person” does not seem to be like this. This seems to allow the inference from one God to one person. To avoid Sabellianism,7

Trin. 7.9.10, 7–8 (1:217).

we have to say something; that that “something” is “person” is puzzling.

Later, in book 7, Augustine twice explicitly makes the point that God's subsistence itself cannot be relational:

If it is one thing for God to be, and another for him to subsist, just as it is one thing for God to be, and another for him to be Father or Lord (for what he is is said non-relationally (ad se), whereas “Father” is said in relation to the Son and “Lord” to the creature which serves him), so that he subsists relationally just as he begets relationally and is Lord relationally—so now substance will not be substance, because it will be relational. For just as “essence” (essentia) is named from being (esse), so we say “substance” from what it is to subsist. But it is absurd that substance should be said relationally, for everything subsists non-relationally. How much more does God—if indeed it is fitting that God should be said to subsist.8

Trin. 7.4.9, 11, 142–50 (1:260).

This yields the conclusion that substance is non-relational—and since, as we shall see below, being a person entails subsisting, it seems that “person” should be a non-relational predicate. Later, Augustine makes a similar point:

Therefore, as the substance of the Father is the Father himself, not as he is Father, but as he is, so too the person of the Father is not other than the Father himself. For person is said non-relationally (ad se), not in relation to Son and Spirit, just as he is called “God,” “great,” “good,” “just” and all other such things. And just as it is the same thing for him to be as to be God, great, and good, so it is the same thing for him to be as to be person. Why therefore do we not call these three together “one person,” as “one essence” and “one God,” but say “three persons” even though we do not say “three gods” or “three essences,” unless it be because we want some one word to serve for this meaning by which we understand the Trinity, so that we would not be entirely silent when asked “what three,” when we confessed there to be three?9

Trin. 7.6.11, 28–33 (1:262).

Neither “substance” nor “person” can be relational terms, for we do not use them as parts of two-place predicates (unlike genuinely relational terms such as “father” or “friend”).10

Trin. 7.6.11, 8–27 (1:261–62).

Now, the gist of all this is that, if “person” is non-relational, then the “one God, therefore one person” or “one essence, therefore one person” inferences should be sound. What Augustine tries to provide in book 7 is a way of understanding the term “person” such that these inferences are not sound, despite the fact that “person” is non-relational. The same problem arises too for the Greek usage, in which “substance” (

) corresponds to “person,” and “essence” (

) to “substance”:

For what must be understood to be said of persons according to our usage must be understood to be said of substances according to the Greek usage. For in this way they say “three substances, one essence,” just as we say “three persons, one essence or substance.”11

Trin. 7.4.8, 112–16 (1:259).

As we shall see, Augustine is not too concerned about the second of these, because he is not strongly committed to thinking of God as a substance. But the first is troubling, for it would again seem to license the disallowed inferences.

Nevertheless, another agnostic statement, parallel to the end of the passage from book 5 quoted above, makes it clear that the distinction between essence and person is hard to understand in both Latin and Greek, since there is no easy answer to the question “what three.” Having noted the divergence in usage between Greeks and Latins, Augustine continues:

When it is asked “what ones the three are (quid tria)” or “what three (quid tres),” we apply ourselves to find out some generic or specific noun under which we can include the three, and nothing comes to mind, because the supereminence of the divinity exceeds the faculty of customary usage. For God is more truly thought of than spoken of, and more truly is than thought of.12

Trin. 7.4.7, 10–15 (1:255).

There is no obvious word to use in reply to the “what three” question, and it is thus perhaps a matter of indifference that the Greeks use “substance” (

) whereas the Latins use “person” (persona).

What is the force of the hesitancy in these passages? We should certainly keep in mind a series of warnings Augustine repeats throughout the book to the effect that we are “speaking of things that cannot be uttered,”13

Trin. 7.4.7, 1–2 (1:255).

and that his motivation for wanting to do so was merely the necessity of responding to heretical views of God14

Trin. 7.4.7, 1–3 (1:259).

—and thus (in accordance with the apophatic statements just quoted) he is not saying something for the sake of saying it, but because it has to be said. But it seems to me that, having embarked on this course of action—speaking of the unutterable for the purpose of refuting Arians15

Trin. 5.3.4, 1–5 (1:208).

and Sabellians16

Trin. 7.4.9, 125–27 (1:259).

—Augustine's rhetorical reserve masks a greater theological confidence, and I shall try to demonstrate this in the next section of my paper.

The Argument in Trin. 7.4.7–7.6.11

Trin. 7.4.7–7.6.11 forms a unit. The first three chapters of book 7 deal with the way in which all three divine persons are identified with the divine wisdom, and there is an obvious break between Trin. 7.6.11 and 7.6.12. Trin. 7.6.12—the last section of the book—deals with some of the issues raised in the earlier parts of the book, but in it Augustine adopts a less argumentative rhetorical voice and attempts to make his points not by reason but by scriptural exegesis. The portion of text I am concerned with focuses on a very particular problem, and Augustine deals with it using the tools not of biblical exegesis but of technical logic. Scripture is used rarely—indeed, never in support of the technical trinitarian aspects of the argument. The general problem Augustine addresses in this portion of Trin. is related to the one I described in the previous section. The term “essence” is common to all the divine persons and likewise common is the term “person.” This is a function of the fact that “person” is not a relational term (as we saw in the previous section). Why then, having asserted that there is one essence, can we not therefore assert that there is also one person? Conversely, why, having asserted that there are three persons, can we not therefore assert that there are also three essences? Specifically, the question that arises at the very beginning of the section is about the legitimacy of talking about three persons at all (I quoted the relevant passage in the last section); Augustine's way into the pair of general questions is by close examination of this specific question. The context, it should be noted, is a return to the concerns first raised in book 5 about the linguistic divergences between Greek and Latin traditions. Despite the various agnostic protestations noted above, it seems to me that, by the end of book 7, Augustine has offered a reasonably clear answer to the questions raised. The effect of the argument is cumulative, and it relies on showing that the divine essence is common to the persons in a way in which divine personhood is not. This provides a principle on the basis of which Augustine can resist the disallowed inferences, even on the assumption that “person” is a non-relational term.

After noting the difficulty of the subject, and its fundamentally mysterious nature, Augustine begins his discussion of the legitimacy of the “three persons” formulation by considering the case of three human beings, Abraham, Isaac, and Jacob, and observes that we can speak of these human beings in the plural “by a specific noun (nomine speciali)”—namely, “man”—or “by a generic noun (nomine generali)”—namely, “animal.”17

Trin. 7.4.7, 16–22 (1:255).

Augustine goes on to note that the plural use of a specific noun entails the plural use of a generic one—if there are three men, then there are three animals—whereas the plural use of a generic noun does not entail plural use of a specific noun—it is not the case that there being three animals requires there to be three of a given species of animal, since most arbitrarily chosen groups of three animals will include animals of different species.18

Trin. 7.4.7, 37–45 (1:256).

This observation, clearly true, is also traditional; compare, for example, Porphyry, developing some insights from Aristotle's Categories:

Now given what each of genus and species is, and given that whereas the genus is one the species are several (for the division of a genus is always into several species), the genus is always predicated of the species, and all the higher things are predicated of those beneath them. But the species is predicated neither of the genus immediately above it nor of higher ones. For it does not go both ways.19

Porphyry, Isagoge (ed. Adolfus Busse; Commentaria in Aristotelem Graeca 4/1; Berlin: Reimer, 1887) 6, 24–7, 4; trans. Paul Vincent Spade, Five Texts on the Mediaeval Problem of Universals (Indianapolis: Hackett, 1994) 6 (par. 33).

Up to this point, the discussion is simply about names, and Augustine studiously avoids making any metaphysical claims about real universals. But on the face of it he goes on to do so:

Abraham, Isaac and Jacob have in common that which is man (commune habent id quod est homo), and for this reason are called three men; horse and ox and dog have in common that which is animal, and for this reason are called three animals. So too we call three laurels three trees; laurel, myrtle and olive we merely call three trees or substances or natures. And so three stones three bodies; but stone, wood and iron we merely call three bodies or by any other generic noun by which they can be spoken of.20

Trin. 7.4.7, 45–55 (1:256).

A number of points call for comment. The philosophical background here is again something like the so-called “tree” of Porphyry from the Isagoge [Isag.]. (Whether or not Augustine was directly acquainted with the Isag. is unknown, but the Porphyrian patterns of argument to which he appeals were common enough in late antiquity.) Porphyry argues that each one of Aristotle's categories can be divided from most general to most specific,21

Isag. 4, 15–20; Spade, Five Texts, 4 (par. 21).

and the example he offers is the division of the genus of substance:

Substance itself is a genus. Man is the most specific and the one that is only a species. Body is a species of substance but a genus of animate body. Animate body is a species of body but a genus of animal. Again, animal is a species of animate body but a genus of rational animal. Rational animal is a species of animal but a genus of men. Now man is a species of rational animal, but no longer a genus—of particular man. Instead it is a species only. Everything prior to individuals and predicated immediately of them is a species only, no longer a genus.22

Isag. 4, 25–5, 1; Spade, Five Texts, 4 (par. 23).

Each species is distinguished from others of the same genus by what is called a “specific difference:”

One thing is said to differ from another when it is distinguished by a specific difference. For example, man is distinguished from horse by a specific difference: the quality rational. … It is in accordance with the differences that make a thing other that there arise the divisions of genera into species, and that definitions are given, since definitions are made up of genus and such differences.23

Isag. 8, 15–17—9, 2–4; Spade Five Texts, 7–8 (pars. 41–42).

Porphyry's discussion concerns language, but it might also concern more than that: real genera and real species, shared extramental universals. And the same could be said of Augustine's discussion. It clearly concerns language (and certainly does in the second half of the last Augustinian passage just quoted), but it might also concern more than that. The impression that it does is reinforced by the claim that three human beings “have in common that which is man,” and three animals “that which is animal.” Does Augustine here mean to suggest that there is an extramental object, a real universal, common to things in a species, and another one common to things in a genus? If so, then the specific universal would be one of the possible referents of the noun “man,” and the generic universal one of the possible referents of the noun “animal.” Whether or not this is what Augustine intends is not known to me, though his intentions in any case make no difference to the overall argument I am attempting to develop here. Supporting a more conceptualist or nominalist reading of Augustine on the issue of genera and species is the thought that (as we shall see) Augustine treats “person” as a generic noun, exactly parallel to “animal”; it is perhaps unlikely that he thought that this committed him to the position that one possible referent of the noun “person” is a real universal, one that in the case of God would be common to the three divine persons as a sort of second essence or nature over and above the divine essence that the three persons somehow share. If Augustine does take a nominalist line on the reality of genera and species, then he has a principled way of distinguishing the commonality of generic terms such as “person” from the real commonality of the divine essence. But in any case, however we construe the commonality of genera and species, Augustine thinks that there is a quite different way in which the divine essence is common to the persons. Briefly, as we shall see below, Augustine holds that genera and species are divisible into their particulars in a way that the common divine essence fails to be. The divine essence, in effect, is indivisible and common. Whenever I refer to “specific” or “generic commonality,” I shall presuppose that the kind of commonality highlighted is divisible commonality.

The argument thus far has been that, according to Augustine, the use of plural count nouns requires some kind of specific or generic commonality. The question that then arises is this:

Since therefore the Father, Son and Spirit are three, let us ask what three (quid tres) they are, what they have in common.24

Trin. 7.4.7, 56–57 (1:256–57).

The issue is what kind of things are persons, and this involves the identification of something that the persons have in common. Augustine makes some progress in this discussion, since he proposes some such features, and this gets him as far as seeing that “person” is a genus term. He does not know what the relevant species term is; it is for this reason, presumably, that he uses the term “person” reluctantly: it would be better if there were a term that could be used to express something specific common to the persons—it would be better, in other words, if there were a species term that we could use. Still, if the term is common, then it is either species or genus:

What three (quid tres)? For if three persons, then that which a person is is common to them (commune est eis id quod persona est). Therefore if we respect the custom of usage, this noun is specific or general to them.25

Trin. 7.4.7, 65–67 (1:257).

Augustine answers this by making some distinctions on the basis of his earlier use of the Porphyrian tree: namely, that three things of the same species (things that, as Augustine puts it, lack “difference/diversity of nature”26

Trin. 7.4.7, 67–68, 69, 73 (1:257).

) are ipso facto of the same genus and that three things of the same species ought to have both specific and generic names.27

Trin. 7.4.7, 67–72 (1:257).

The three divine persons lack diversity of nature and, thus, should have a specific name. In fact, however, there is no such name. “Person” is a generic term, predicable of man and God:

Here indeed, where there is no diversity of essence, it is necessary that these three also have a specific name: which, however, is not found. For “person” is a generic noun, in as much as a human being can also be called this, although there is such a great distinction between man and God.28

Trin. 7.4.7, 72–76 (1:257).

Whatever its meaning, its categorial status is clear: “person” is a generic term since it applies across species, and Augustine leaves unsolved the aporia of the correct specific term for Father, Son, and Spirit. (This lack is perhaps not surprising: there could well be specific terms predicable of divine persons, but such terms will be irreducibly proper to the divine persons, and perhaps not knowable without revelation. And the Bible is notably silent on such technical matters.)

It is clear enough that Augustine believes himself to be able to provide some sort of account of the meaning of this generic word. As we shall see below, he denies that the meaning can include the concept of individuality. But, as Augustine makes clear a bit later on, in Trin. 7.5.10, it certainly includes the notion of subsistence:

29

Trin. 7.5.10, 24–26; 7.6.11, 1–8 (1:261).

failing to be in something as in a subject.30

Trin. 7.5.10, 3–5 (1:260). I return to this important passage below, for Augustine does not want the relevant notion of subsistence to include being the subject of accidents.

The notion of subsistence here is one, however, that Augustine is careful about, for paradigm cases of things that subsist are substances. “Substance” is said non-relationally. In this respect it is just like both “essence” and “person.” But the notion of substance implies being a subject of properties—“standing (stans) under (sub)” greatness, wisdom, omnipotence, goodness, and so on, in the case of God. God is therefore more properly talked of as essence, not substance, since he does not “stand under” his attributes, but is identical with them.31

Trin. 7.5.10, 1–26 (1:260–61).

Equally, “three persons” is a better formulation than “three substances” for just the same reason.32

Trin. 7.5.10, 1–2 (1:261). The critical edition includes this sentence as 7.6.10, linking it with the next section. In fact, the later division of the text at either of these points tends to obscure the evident continuity of the argument here.

Still, the notion of non-inherence (subsistence) does not entail that of “standing under” properties (substance); for this reason, persons can subsist while not being substances. Augustine's concerns about the Greek usage, then, might focus on the thought that, if the divine persons are substances (as the Greeks suppose), then, counterfactually, they must stand under the divine properties. But he is clearly not too concerned about this fundamentally technical clarification, acknowledging that different usage does not necessarily signal different metaphysical commitment.33

Trin. 7.6.11, 2–8 (1:261).

Perhaps his complaint about the Greek distinction between substance and person,

and

(“I do not know what they intend to be the difference”) relates either to the lack of any previous attempt known to him to define

and to assign to it any appropriate categorial analysis, or to the Greek attempt, rejected by Augustine, to analyze

in terms of individuality. Augustine does not say, and I do not know. (I return to this in section 3 below.) What is clear is that Augustine is puzzled by the fact that “person” is a generic term and that there is no specific term known to him to refer to Father, Son, and Spirit; this puzzle affects the Greek account as much as his own. Augustine is happy to talk about “three somethings (tria quaedam)” in this context; presumably “something (quoddam)” is a higher genus for “person.”

34

Trin. 7.4.9, 125 (1:259).

With something like this account of the word “person” presupposed, Trin. 7.4.8 begins the attempt to answer the general question posed at the beginning of the unit of text under discussion here. If having something in common is sufficient for countability, then, since the divine essence is common, surely we should be able to treat “God” as a count noun too:

In relation to the generic term itself [viz. “person”], if we say “three persons” because that which a person is is common to them … why do we not also say “three gods”? For certainly, since the Father is a person, and the Son a person, and the Holy Spirit a person, there are for this reason three persons. Therefore since the Father is God and the Son is God and the Holy Spirit is God, why are there not three gods? Or since, on account of the ineffable conjunction, these three are together one God, why are they not also one person such that we could not say “three persons,” even though each singular one we call “a person,” just as we cannot say “three gods” even though we call each singular one “God,” whether the Father, or the Son, or the Holy Spirit?35

Trin. 7.4.8, 77–78, 81–89 (1:257–58).

The answer to this question is scriptural—namely, that “three gods” is contradicted by the Bible, whereas “three persons” is not.36

Trin. 7.4.8, 90–97 (1:258).

This does not dispose of a closely related theological puzzle, however:

Why is it not therefore also lawful to say “three essences,” which likewise Scripture neither says nor contradicts? For if “essence” is a specific name common to the three, why are they not called “three essences” as Abraham, Isaac and Jacob are called “three men” on the grounds that “man” is a specific noun common to all men? If however “essence” is not a specific noun but a generic one, since man, cattle, tree, constellation and angel are called “essence,” why are these not called “three essences” as three horses are called “three animals” and three laurels “three trees” and three stones “three bodies”?37

Trin. 7.4.8, 97–106 (1:258).

If “essence” were a specific or generic noun, we would have to say “three essences,” just as we say “three persons,” on the assumption that “person” is a generic noun. Augustine will eventually draw the conclusion that the divine essence is common to the persons, but not in the divisible way that a genus or species is common to three individuals or particulars.

Now, that God is one essence, Augustine thinks, is easy enough to show. Since it is the same thing for God to be God as for him to be (esse), then if there is just one God there will just one essence (essentia, “be-ing”).38

Trin. 7.4.9, 139–41 (1:259–60).

But this raises the question about “person,” since “person” is said non-relationally, and accepting one (non-relational) essence seems to run the risk of forcing an acceptance of one (non-relational) person too. The relevant passage is the third quoted in section 1 above. The reply is in two stages. First, the divine essence is neither the genus nor the species of the divine persons. (If it were, then “essence” would be a count noun, as are all specific and generic nouns according to Augustine.) Secondly, the divine essence is a numerically singular thing shared by the three divine persons (it is “common and the same” in all of them). And this justifies our saying “one essence,” and thus provides a basis for the biblical usage of “one God,” just as the generic status of the term “person” justifies the phrase “three persons.” What Augustine needs here is an account of the essence that distinguishes it from any genus or species. And this is precisely what he goes on to provide.

The first of these two stages considers in turn the possibility that the divine essence could be a genus and the persons species or individuals of it, and the possibility that the divine essence could be a species and the persons individuals of it. One result of this stage of the discussion is that the definition of “person” cannot include the concept of individuality: the persons are not individual instances of some genus or species.39

Interestingly, Augustine never treats Father, Son, and Spirit as individuals of the genus person, but while it is easy to think of reasons why he would not want to treat of them in this way, he says nothing that is sufficient to prevent the conclusion that the persons are in fact individuals of this genus. After all, person is clearly a genus, and Augustine is happy enough with the claim that there are three persons—which looks like a claim that the persons are three individuals of the genus person. Nevertheless, it is clear enough from my discussion thus far that Boethius's famous definition of “person” as “individual substance of rational nature” is markedly un-Augustinian. Boethius, De persona et duabus naturis contra Eutychen et Nestorium 3 (ed. C. Moreschini; Bibliotheca scriptorum graecorum et romanorum teubneriana; Munich: Suar, 2000) 214.

The possibility that the divine essence could be a genus and the divine persons species of it is attributed by Augustine to “some (nonnulli)” theologians.40

On the Cappadocian view that the divine essence could be a genus or species, and the persons individuals, see section 3 below.

Augustine notes that, if this possibility is correct, then the Porphyrian rule of inference from many cases of a species to many cases of a genus does not obtain, for there would be three persons (species) and yet just one essence (genus), contrary to the established laws of logic.41

Trin. 7.6.11, 33–41 (1:262–63).

If the divine essence is a genus and the persons individuals, then two distinct logical rules are violated. First, in a manner analogous to the case just considered, the soundness of the inference from many individuals to many cases of a genus is broken (since there would be three individual persons and only one God, whereas in standard cases such as Abraham, Isaac, and Jacob we can infer both three men and three animals, for example).42

Trin. 7.6.11, 41–51 (1:263).

Secondly, we would be left with an anomalous case of a genus with no species (thus fracturing the pattern of division established in the Porphyrian tree).43

Trin. 7.6.11, 53–55 (1:263).

The divine essence is therefore not a genus.44

Trin. 7.6.11, 55–56 (1:263).

Secondly, the divine essence is not a species to which individuals belong. The reason is simple, and Augustine's discussion of it provides the key evidence for my assertion, made above, that the genera and species have some kind of commonality compatible with divisibility. A species is divided into its particulars—“subdivided,” as Augustine puts it—as (for example) the species man is into many men.45

Trin. 7.6.11, 56–60 (1:263).

But the divine essence is singular, and like any singular item is indivisible (just as an individual human being cannot be “subdivided into other singular men”).46

Trin. 7.6.11, 60–65 (1:263).

There is also clearly a sense in which different created individuals can be said to be the same essence. But this seems to be an informal sense of sameness—Augustine prefers the non-technical term “nature” in this context—that Augustine mentions merely to introduce his own preferred analogy, that of common matter (to which I return in a moment):

Do we say “three substances, one essence,” or “three persons, one essence” just as we say that three men of the same sex, of the same bodily constitution, or of the same mind are one nature (for there are three men and one nature)? And this is in some way similar, for the ancients, who spoke Latin before they had these words (which came into use only recently)—namely, “essence” or “substance”—used “nature” to say them. Therefore it is not according to genus and species that we say these things, but according to the same common matter, as it were.47

Trin. 7.6.11, 65–76 (1:263–64), my italics.

The final sentence marks the conclusion of the argument about genus, species and individual that has occupied Augustine from the beginning of Trin. 7.4.7. The divine essence is neither a genus nor a species. The point of the preceding sentences is that “essence” need not be understood in any technical way, as implying genus or species; it could be equivalent to “nature,” and this latter term (presumably) can be used to pick out not merely the divisible commonality of genus or species, but also some kind of indivisible commonality: the commonality of something identical in the things that share it. And this indivisible commonality explains why the “three essences” usage is illegitimate—the second stage of Augustine's explanation of the unsoundness of the inference “three persons, therefore three essences/three Gods.”

The example of such commonality that Augustine develops and, with strong qualifications and reservations, accepts is that of matter:

We do not therefore say these things according to genus and species, but as it were according to the same and common matter.48

Trin. 7.6.11, 74–76 (1:264).

The analogy is not all that helpful, though it certainly has the advantage of appearing in Augustine's predecessors—most notably in this context Basil of Caesarea.49

See, e.g., Basil, Contra Eunomium 2.4 (PG, XXIX, 577C).

But the key thing about it, as far as I can see, is that matter is (in some sense) the same in all the things that are made of it:

Just as, if three statutes are made from the same gold we should say “three statues, one gold,” we should not say that the gold is a species and the statues individuals.50

Trin. 7.6.11, 76–79 (1:264).

The great benefit of the analogy, as Augustine sees it, is that “gold” is a mass noun, not a count noun, and there is thus no valid inference from “three statues” to “three golds,” as it were. And the illegitimacy of such an inference obtains in the case of the divine essence too: we cannot argue from three divine persons to three divine essences. In neither case, that of gold or of the divine essence, are we talking about a genus or species. The disadvantage of the analogy, according to Augustine, is that there are things other than statues that are made of gold (whereas only men fall under the definition of “man”). Both points, positive and negative, are made in the following passage:

Therefore, although in the three statues we should rightly say “three statues, one gold,” we do not however speak such that we understand gold to be a genus and the statues species. Neither do we call the Trinity “three persons or substances, one essence and one God” in this way, as though the three subsisted from one matter, even if whatever that is is unfolded in the three: for there is nothing else of its essence other than this Trinity.51

Trin. 7.6.11, 90–97 (1:264).

I take it that the denial made in the second sentence here is not that the analogy to matter is unsound tout court; we should not see Augustine here as rejecting the model, merely as providing strong restrictions on its applicability.

The remainder of the section, up to the end of Trin. 7.6.11, makes some concluding clarifications, for the most part further circumscribing the applicability of the analogy to gold. “Gold” is a mass noun, and as such admits of quantitative variations such as “more” or “less”; “essence,” however, is not actually a mass noun, despite its similarities to such nouns in other respects as just outlined. One divine person is no more of the divine essence than three are—the essence, it seems, is fully in, and coordinatively common to, each divine person.52

Trin. 7.6.11, 107–14 (1:265). This claim allays one remaining fear that the analogy might cause, namely, that the sameness between the divine persons is merely specific, not numerical. On the numerical indivisibility of the divine essence, see e.g., Epistola 120.2.7 (Epistulae 31–123; ed. A. Goldbacher; Corpus Scriptorum Ecclesiasticorum Latinorum, 34/2; Vienna: Tempsky; Leipzig: Freytag, 1898) 710, 15–20.

And we should not claim that the persons are “from” the essence in the way that statues are made “from” gold. The preposition “from” implies that there can be more things that remain to be produced, and there can be no more divine persons.53

Trin. 7.6.11, 98–102 (1:265).

(Comparing “essence” to the looser and less technical “nature,” Augustine notes parenthetically that we can say “three human beings from one nature” since there can always be more human beings, again reinforcing his claim that “essence” predicated of the divine persons is not a species noun.54

Trin. 7.6.11, 102–7 (1:265).

)

Augustine's view, then, is that each divine person includes the common divine essence, and that each divine person is distinct from any other divine person by a relation. The essence cannot be a genus or species of the persons, since genera and species are both divided into their particulars. The divine essence is fully real (as evidenced by the analogy to matter), and yet undivided—identically the same singular thing in each divine person. About this, it seems, Augustine is not overly puzzled, though he knows that the best he can offer is a highly-qualified analogy from the material realm. As he puts it,

[The animal man] cannot think except of mass and space, whether small or great, as phantasms, like the images of bodies, fly around in his mind.55

Trin. 7.6.11, 114–17 (1:265).

Still, it seems that Augustine has answered his question about the way in which to prevent inferences of the “one God, one person” or “three persons, three gods” kind, and that this question no longer holds particular problems—even on the assumption (accepted by Augustine) that “person” is a non-relational term. There are created analogues to the one essence and three persons in God (e.g., one gold, three statues), and in any case the rejected inferences would obtain only on the understanding (which Augustine has shown to be mistaken) that the divine essence were a genus or species. Count nouns are, in Augustine's analysis, either generic or specific terms, and “God” and “essence” in this context are neither—or at least, they have enough in common with mass nouns to render any plural usage illegitimate. (This is the case notwithstanding that we can say “one God,” “one essence”—something not possible with a mass noun. We can, of course, meaningfully speak of the same matter in such contexts, just as we can meaningfully speak of the same divine essence.) Augustine's best shot at an answer is an analogy, the example of common matter, and (as I have been arguing) the reason for this is that both “God” and “essence” have in common with mass nouns such as “matter” the fact that plural usage is illegitimate.

I take it, then, that Augustine has offered some kind of an answer to the fundamental puzzle and that, given the qualifications that he himself provides, there is no evidence that he believes the answer to be unsuccessful. But he confesses himself stumped by another question: “Quid tres”—“what species are the three persons?” He knows that the persons belong to a genus, and he has a pretty good idea as to what this genus is (it is the class of things that do not inhere). But he is puzzled by the lack of an appropriate species word here and obviously hesitant to use a merely generic word (“person”) in the absence of a species word.

What, If Anything, Augustine in Fact Misunderstands

It seems, then, that Augustine is reasonably clear on the things he finds hard to understand, and what is odd is that, contrary to the suggestion made by Colin Gunton and quoted at the beginning of this article, none of them in fact have anything to do with Augustine's interpretation of the Greek-speaking theologians. Here I would like to consider whether there is any other evidence, at least from book 7 of Trin., to support Gunton's claim that Augustine misunderstands the Greek theologians. As I indicated above, the observations I shall offer are limited in extent, and provisional, since a fully satisfactory treatment of the Greek traditions on the question of the distinction between person and essence, on the failure of the inference from three persons to three gods, and on the definition of the term

, would require much more space that I can allocate here. The issue is made somewhat harder by the fact that we cannot be sure precisely which Greek theologians Augustine read. Irénée Chevalier has shown that Augustine read the pertinent sections of Gregory of Nazianzus and Didymus the Blind, and perhaps of Basil of Caesarea and Epiphanius of Salamis.

56

See Irénée Chevalier, Saint Augustin et la pensée grecque. Les relations trinitaires (Collectanea Friburgensia 24; Fribourg: Librairie de l'Université, 1940) 98.

The Greek writer who deals most conspicuously with some of the issues discussed here, however, is Gregory of Nyssa, and I shall allow my discussion to include this Gregory too, since, from a conceptual point of view, he seems to be the most obvious source for, or target of, some of Augustine's remarks.

I have argued that the aporia left by Augustine in book 7 of Trin. is that our talk of three somethings or three persons in God is not wholly satisfactory, and that the reason for its unsatisfactory nature is that “something” and “person” are genus terms. We do not know what species the persons have in common, and we do not know what word we could use to refer to this species. Clearly, this puzzle is not dependent on any particular concerns about translation from Greek, but Augustine nevertheless makes explicit that he finds the equivalent Greek term

equally puzzling, not least because the Greek writers use it without providing any definition of it (“I do not know what they intend to be the difference between

and hypostasis”). This is quite right: none of the Greek writers just mentioned offers any definition of the term

(though they certainly know how they want to use the term).

57

In a recent study, Lucian Turcescu has argued that Gregory of Nyssa formulates the most sophisticated understanding of “person” and “individual” in antiquity. But on Turcescu's reading, what this amounts to is, first, having a range of synonyms for

; secondly, noting that persons are individuals (as opposed to universals); and thirdly, maintaining that they are collections of properties existing in “communion” with other persons. See Turcescu, Gregory of Nyssa and the Concept of Divine Person (American Academy of Religion Academy Series; New York: Oxford University Press, 2005) 116–17. Even if a definition of the word “person” could be formulated using these tools (and it is hard to see how it could, since the third conjunct would require use of the definiendum), Turcescu's study makes it abundantly plain that Gregory himself does not tie all these threads together into one clear and explicit definition. For Turcescu, “communion” fundamentally involves the inseparability of the divine persons, spelled out in terms of the origination relations between them. That Augustine would accept this without any difficulty goes without saying. Indeed, to judge from well-known passages such as the following, he would go further: “[The Holy Spirit is that] by which the begotten is loved by the begetter and [by which the begotten] loves his begetter.” Trin. 6.5.7, 4–6 (1:235); “That ineffable conjunction of Father and Image is not without enjoyment, without love, without happiness. Therefore that delight, pleasure, felicity or happiness, if it can be worthily expressed by any human voice … is the Holy Spirit in the Trinity.” Trin. 6.10.12, 29–33 (1:242). The persons are not merely contemplated together by us; they contemplate each other too.

Augustine seems to have understood this failing pretty well and to have made it his task in book 7 of Trin. to rectify it. The insight that “person” is a non-relational word—an insight that certainly seems to be an innovation of Augustine—perhaps makes the philosophical problem harder, but it does not change its fundamental nature. For on any plausible understanding of “person” in this context, there are problems preventing the inference from one God to one person. Furthermore, Augustine's concern is a technical point about the word “person” and its cognates. Augustine certainly does not deny that the persons are distinct by relations or that “Father,” “Son,” and “Spirit” are relational words.

Beyond this, however, things get a bit more complicated. For Augustine explicitly rejects the view that the divine essence could be a genus or species, and the persons species or individuals, and he does so for two reasons: the divine essence is neither genus nor species, and “person” is a genus noun. The second of these claims—making “person” in effect a natural kind—seems to be his own suggestion, and it certainly provides a suitable philosophical underpinning for the legitimacy of the “three persons” formulation (there is some divisible kind under which we can count the persons). On the first, there is some prima facie divergence from the Greek tradition. Basil of Caesarea uses the analogy of a common genus and an individual instance of a species to explicate the relation between the divine essence and a divine person.58

See Basil, Epistola 236.6 (ed. Yves Courtonne; 3 vols; Collections des universités de France; Paris: Belles Lettres, 1957–66) 3:83–84.

Still, Basil elsewhere is perfectly happy to use the analogy of matter and things made from the matter, as I pointed out above. What Augustine seems to be doing is simply restricting the range of valid analogies, presumably on the grounds that seeing the essence as a genus or species would make it impossible to resist the inference from three persons to three gods. Gregory of Nyssa encounters just this problem with the analogy to a species, and, in a notorious passage, draws the opposite philosophical conclusion to that drawn by Augustine. Since the analogy is good, it must follow that our understanding of species nouns in general—as count nouns—is mistaken. Gregory makes the metaphysical point in the strongest possible terms:

There are many who have shared in the nature [of man] … but the man in them all is one.59

Gregory of Nyssa, Ad Ablabium, in Opera Dogmatica Minora. Part 1. (Gregorii Nysseni Opera III/1; ed. F. Müller; Leiden: Brill, 1958) 40, 17–18, 19; for the whole discussion, including the merely semantic claims about the proper use of the term “man,” see 40, 15–42, 12.

So despite the well-known Greek use of the analogy between universal and particular, seeing the divine essence as a species (something explicitly rejected by Augustine), it is clear that even the person who uses this analogy most proficiently is fully aware of its problems. And the problem is precisely that which leads Augustine to reject the analogy: on the face of it, genera and species are divisible into particulars; the divine essence is indivisible. Gregory agrees, at least in certain texts, that the divine essence is coordinate and indivisible.60

See e.g., Gregory of Nyssa, Ad Ablabium, 41, 2–7. For a good discussion of this, see Johannes Zachhuber, Human Nature in Gregory of Nyssa: Philosophical Background and Theological Significance (Supplements to Vigiliae Christianae 46; Leiden: Brill, 2000). Gregory's stress on this indivisibility of the divine essence makes him reluctant to think of the divine persons as (in any strict sense) individuals; on this, see Richard Cross, “Gregory of Nyssa on Universals,” Vigiliae Christianae 56 (2002) 372–410, esp. 405–8.

Viewed in this way, it is hard to see the divergence as implying anything more than an expression of a different set of philosophical choices (here about the divisibility/indivisibility of species), resulting in a different selection of appropriate analogy.

What of Gregory of Nazianzus, at least some of whose works were definitely known to Augustine? Like Augustine, Gregory is far more hesitant than his two Cappadocian friends to use the analogy to three human beings—indeed, he explicitly restricts the relevance of the analogy by arguing that species such as man are merely concepts, a move that would be exactly paralleled by Augustine's analysis if we adopt the conceptualist or nominalist reading of Augustine's view of divisible genera and species mooted above:

The community in this case [viz. that of three men] has a unity seen only in thought, and the individuals are different from each other, divided in time, capacities, and power.61

Gregory of Nazianzus, Or. 31.15, 4–7, Discours 27–31 (Discours théologiques; ed. Paul Gallay; Sources chrétiennes 250; Paris: Cerf, 1978) 304 (my italics).

But, again, this is a concern about the divisibility/indivisibility of species, and certainly consistent with Gregory's professed acceptance—at least in certain passages—of a view of the divine essence that makes it coordinate and indivisible.62

The key passage is Gregory of Nazianzus, Or. 31.14, 2–9 (Gallay, Discours théologiques, 302); see too, e.g., Or. 30.11, 7–10 (Gallay, Discours théologiques, 246).

Conclusion

Overall, there seems to be plenty of evidence against Gunton's claim quoted at the beginning of this article. Indeed, even the small amount of material presented here suggests that there is considerable divergence between the various Greek theologians—even those customarily given a single label (“Cappadocian”)—such that we should not think of the late fourth-century Greek Nicene tradition as theologically homogeneous or unitary. It seems to me that there is no evidence at all in favor of any theologically significant version of Gunton's divergence claim, but demonstrating that would be the work of another article. What Augustine does not understand—and presents the Greeks as not understanding—is what species the divine persons have in common, since “person/

” is (on Augustine's analysis) a genus word. His hesitancy about the word “person” is simply a function of his reluctance to use a genus word without an available species word. And this is a technical philosophical concern that has no bearing on the theological question.