Augustine's de Trinitate 5 and the problem of the divine names "father" and "son".

Weedman, Mark

WHEN AUGUSTINE MEMORABLY ENCOUNTERED the "Books of the Platonists" and came to agree that God was immaterial, he was merely entering the mainstream of Latin theology. For at least two generations before Augustine's conversion, Latin Pro-Nicene theologians had accepted as a default position that God was immaterial--which is not to say that they had worked out all the implications of what that meant. In particular, the Latin theologians had difficulty explaining the relation between the Father and the Son in categories that preserved God's immateriality. These theologians initially thought they could account for the Father-Son relation by using the names "Father" and "Son" to explain how the Son could be equal to the Father: on analogy with human fatherhood, the divine Father gave birth to a Son who was distinct from but equal in nature to the Father. This approach had much to recommend it, as it was consistent with the traditional Latin way of describing the Father-Son relation, and it seemed to preserve both the unity and distinction of the Father and Son. But as the controversy progressed, it became increasingly clear to its proponents that this approach had deep flaws, because it depended on a human, "material" analogy--that of human fatherhood--to describe an immaterial relation; if the relation was also immaterial, then it would require "immaterial" categories to explain it.

The question that Augustine sets out to answer in De Trinitate 5, then, is how the unity and diversity of the Father and Son can be spoken of in a way that preserves the immateriality of that relation. He does so by abandoning the analogy with human fatherhood. He argues that some names for God, such as "Father" and "Son," tell us only about relations within God but nothing about God's substance. Thus the primary theological task when "explaining" the Trinity is not to show how a material relationship can describe the divine relations, as Augustine's predecessors had done. Instead, Augustine suggests, theologians must determine which names are substantial and which are relational.

Scholars have paid comparatively little attention to Augustine's theory of naming in De Trinitate. (1) However, not only does Augustine's De Trinitate 5 resolve a long-standing theological conundrum, but the way Augustine develops his theory on this question also shapes his trinitarian thought as a whole. In figuring out how to "dematerialize" the relation between the Father and the Son, Augustine also recognized that the same categories he used to articulate this relation could also contribute to a trinitarian grammar, as well as establish hermeneutical rules for properly interpreting Scripture and other texts. Augustine's approach to the names "Father" and "Son" in De Trinitate 5, therefore, provides a lens by which to explore the entire scope of his thought.

My intent here is to assess Augustine's thought in De Trinitate 5 by situating it within the context of Latin Pro-Nicene trinitarian theology. I argue that we can best understand Augustine's argument in De Trinitate 5 as the culmination of a process that began as Latin ProNicenes began to abandon the notion that they could express the relation between the Father and the Son by using an analogy with human fatherhood, an analogy that failed because it was too "material" to express the relation between the immaterial Father and Son. I begin by surveying the history of the father-son analogy in Latin Pro-Nicene thought, looking first at Phoebadius of Agen, who represents a traditional Latin response to the Homoians, then turning to Hilary of Poitiers, who first accepts the "name" motif and then gradually abandons it, and ending with Ambrose of Milan, who represents the transition between Hilary and Augustine. I then turn to Augustine's solution to this problem, especially in De Trinitate 5, where he develops the notion that some names should be applied to God's substance and some to God's "relations." I conclude by considering areas in which the theory of the relations in De Trinitate 5 opens avenues of inquiry into the rest of Augustine's trinitarian theology.

NAMING, SUBSTANCE, AND ANTI-HOMOIAN POLEMIC IN LATIN PRO-NICENE THEOLOGY

Augustine's predecessors in the West had used the names "Father" and "Son" to explain both their unity and their diversity: Father and Son are two different or distinct names, indicating that there are two persons. But because fathers give birth to sons who share the father's nature, the Son of the divine Father must also be divine and so one with the Father. Opponents of the Nicene position, however, quickly realized that this logic was flawed because it relied on a material analogy to explain an immaterial relation. Among other problems, the analogy implies that the Son was born in time and is subordinate to the Father, since in human, "material" experience, fathers must necessarily exist prior to the birth of their sons and have authority over their offspring. Likewise, if we assume that names reveal nature--a position that a number of theologians across the theological spectrum took--then we have an added problem, because the names "Father" and "Son" name two distinct substances. (2) Thus, the Homoians argued, the analogy between divine and human birth is actually more suited to a subordinationist understanding of the Father-Son relation, which means that it is not at all sufficient to explain the substantial unity between the Father and the Son.

In response to these attacks on the father-son analogy, Latin ProNicenes began to reconsider their use of the father-son analogy, and we can identify a process in which Latin theologians attempted to explain the relation between the Father and the Son without using material categories or analogies. I begin this discussion by examining the evidence of Hilary of Poitiers. (3) Hilary is important because he illustrates how and when ProNicenes began to transition from a view that names reveal nature to a more conventional view of the names "Father" and "Son." Hilary's first mature, polemically engaged attempts to refute the Homoians takes the notion that the names "Father" and "Son" refer to the Father's and the Son's natures as something central to Pro-Nicene thought. It did not take Hilary long, however, to turn away from that view, and so to begin a process that culminated in Augustine's De Trinitate 5.

Hilary and the New Direction

In his De Trinitate 7 Hilary argued that the name "Father" refers to the nature of God; when we call God "Father," we are saying that he is Father by nature. This insight has a number of implications for Hilary's trinitarian theology. If God is Father by nature, he suggests, then he is eternally Father, but to be eternally Father, he had to have a Son from eternity. Thus Hilary thinks he can use the substantial quality of the names "Father" and "Son" to preserve both the unity and distinction of the Father and Son: they are united because both are eternal and fathers give their natures to their sons, but they are distinct because they have

different names, Hilary recognizes that there is a potentially serious problem for Pro-Nicenes in this line of reasoning because sons are subordinate to their fathers, which would imply that God the Son was subordinate to God the Father, and Hilary continually insists that the analogy works only if we strip away any materialistic connotations. Still, for most of his anti-Homoian polemic, this is Hilary's primary argument for the eternity of the Son and the substantial relation between the Father and the Son. (4)

A potential source for Hilary's emphasis on the natural quality of the names "Father" and "Son" is Basil of Ancyra. (5) Basil places the names "Father" and "Son" at the center of his anti-Homoian polemic. So, for example, Basil makes much of Matthew 28:19, "baptizing them in the name of the Father, and the Son, and the Holy Spirit." Basil argues that this passage forces us to recognize the "natural meaning of the names":

For he did not say, "Baptizing them in the name of the Incorporeal and the Incarnate," or, "in the Immortal and the One who has suffered death," or, "in the Unbegotten and the Begotten," but, "in the name of the Father, and the Son, and the Holy Spirit," in order that, adhering to the natural meaning of the names ... we may know the Son to be like the Father, of whom he is the Son. (6)

Basil recognizes that these names do create a difficulty because the name "Father" implies passions, i.e. a father generates a son through an act of passion, and so when we speak of the Father and Son, Basil believes, we must eliminate the possibility of passion. (7) Nevertheless, by removing passion from the equation, Basil is convinced that attending to the idea that the names reveal nature helps us recognize how the Son can be "like according to substance" with the Father. The key to Basil's argument is his assertion that "we also hear the names in the nature, and a father always begets a son like himself, and we may understand the Father to be the cause of an essence like his." (8) Just as human fathers produce sons who share the father's nature, so too with the divine Father and Son. If the Father is a father by nature, then he must have a Son who shares that nature.

Basil himself had anticipated some of the problems with this argument, i.e., that the analogy also implied both "passion" on the part of the Father and the subordination of the Son to the Father. Before writing the twelfth and final book of De Trinitate Hilary also came to realize that his Homoian opponents were using a version of his Father-Son argument for their own purposes, and that they were doing so in ways that were even more effective than his own approach. (9) And so in De Trinitate 12 he begins to distance himself from the Father-Son argument. In this book Hilary acknowledges that the Homoians used the names "Father" and "Son" to argue that if the name does give access to the divine nature, the only way to understand that name, and so to understand the divine nature, is by way of the normal understanding of that name.

To grasp what the name "father" tells us about the divine Father, we must look at what it means to be a human father. Thus, according to the Homoians, human experience dictates that whatever experiences birth has a beginning in time. This is, in a sense, the reason for a birth--to bring about the beginning of the one who was born. (10) If we accept that the Son was "born" of the Father, then we must, on the basis of the father-son analogy, also believe that this birth necessarily occurred in time. In this way, then, Hilary's claim that the birth of the Son indicates his eternal generation does not stand the test of human logic.

Hilary responds to this attack in two ways. First, he argues that analogies between human experiences and the divine nature are limited because, while they help us understand certain aspects of the divine nature, they do not reveal to us anything necessary about that nature. (11) We know, for example, that neither the Father nor the Son is a creature; thus we cannot assume that the ordinary experience of a human birth applies to the divine birth. The analogies to human generation can tell us some things about how the divine Father and Son relate, but they do not hold for everything about the divine relations. (12) This point coheres with the way Hilary argued throughout De Trinitate; the potential for taking the father-son analogy "materially" is obvious, which is why Hilary repeatedly insists that the analogy works only if you strip away its materialist connotations. What he seems to have realized here, however, is that by removing material considerations from the analogy, we also limit our access to knowledge about the divine nature--which is a problem, because the entire argument depends on the ability of the analogy to tell us something about the divine nature: just as human fathers give birth to sons who share their father's nature, so too with the divine Father and Son. The Homoians had convinced Hilary that he could not have it both ways: either the analogy works from its "created" perspective, or it does not, to strip away its material connotations is also to limit its ability to tell us anything useful about the divine nature.

As a result, Hilary has to rethink completely his argument for eternal generation, and he does so largely by finding language to replace the father-son analogy. Since he cannot use the father-son analogy, he decides to turn to what we can say about God as one. In other words, Hilary is going to bypass the analogy altogether and go directly to the divine nature. He turns to Philippians 2, for example, to argue that because the Son is in the form and image of God, he is equal to God, and he possesses "all these attributes that are proper to God the Father, through the fullness of the Godhead in himself." The primary attribute of God is eternality, which Hilary here defines as infinity: when we contemplate God, we find that God is "forever drawing away from the contemplation of our infinite perceptions," so that in our contemplation we grasp nothing but the notion that "God always is." (13) If the Son shares this attribute, therefore, then the Son can have no beginning because "to be and not to be are contradictory to each other." (14) "From Him who is eternal, nothing else comes except what is eternal." (15)

I do not want to give the impression, however, that Hilary has abandoned the father-son language, and even in this section he constantly reverts to it. But what is now missing is an argument from analogy. Hilary no longer argues that the Father, by virtue of being a father, eternally gives birth to a Son, on the analogy of human fathers and sons, and we even see Hilary slipping between the language of birth, which is analogical, and that of generation, which is of a different order of discourse altogether. (16) Instead, the basis of his argument now proceeds from his conception of God's infinity and his notion that if God generates from within the divine infinity, then the Son cannot have a beginning.

Ambrose and the Relation between Father and Son

Ambrose's first foray into the trinitarian controversy came at the emperor's command to write a book on the trinitarian faith--indeed his early trinitarian theology displays little awareness of the current controversies. By the later books of De fide, however, Ambrose exhibits a much clearer conception of what is at stake in the Homoian controversy, including a better sense of how Latin Pro-Nicenes can respond to the Homoians. (17) Ambrose's treatment of eternal generation in De fide 4 is indicative. Like Hilary, Ambrose grounds his understanding of the Son's generation in the Father's eternity. The Homoians had asked that, if it is true that the Father and Son are one nature, why does not the Son also beget? Ambrose's answer is that to beget does not belong to God's "power" (potestas) (which Ambrose equates with substantia) but to what is proper to his function as Father. (18) Thus the Son can share in everything that belongs to the divine substance and not have to beget, because begetting (or being begotten) belongs to the category of "property" (proprietas), not to potestas. (19) This is not precisely Augustine's "relation" argument, but it is close. Ambrose's intent is to separate the property of fatherhood and begetting from the level of substance; he does so by employing a philosophical category that describes something integral to the being of the Father, but that ultimately reveals nothing of the divine substance. Ambrose lacks the technical precision of Augustine's relational language, but they are pursuing the same goal. (20)

Ambrose concludes the work with a comment that is more agnostic than constructive: "We know the fact of distinction, we know nothing of the hidden mysteries; we pry not into the causes, but keep the outward signs safe within us."2l This is an interesting answer on a couple of levels, not least because it might be a very early form of Augustine's doctrine of relations. However, to describe the relation between the Father and Son this way opens Ambrose's theology to the possibility that the Father chose to beget the Son in time. Ambrose has to address the question of will, namely, is the begetting a question of the Father's will? Ambrose equivocates, refusing to admit that the begetting was either by will or by necessity. Instead, he locates the generation in the Father's eternity: the Father does not have to choose to beget any more than he chooses to be good (22) In an ironic twist, Ambrose then returns the materialist argument back on the Homoians by noting that if the Father wills to beget, then he must have done so out of a desire to have all the "human" experiences that surround human begetting--a notion that would have been as distasteful to a Homoian as to a Pro-Nicene! (23)

We can acknowledge that Ambrose's theology lacks some technical precision and still note that he follows Hilary in pushing Latin Pro-Nicene theology in a different direction. Like the later Hilary, Ambrose is concerned to deemphasize theological inquiry into the substance of God, which includes denying that the names "Father" and "Son" point us towards substance. If Hilary's contribution to this impetus was to remove those names from the level of substance, Ambrose's contribution is more subtle, but no less important. Ambrose establishes for the Latins that there are divine attributes and actions that do not necessarily reveal anything about divine nature. Those actions that do not reveal the divine nature he categorizes as proprietas, those that do reveal the divine nature he calls potestas. The actions associated with fatherhood and sonship belong to proprietas. This move not only deemphasizes the substantial character of the names "Father" and "Son": it also pushes trinitarian inquiry away from God's substance altogether.

USING THE TRADITION: AUGUSTINE ON BIRTH AND ETERNAL GENERATION

From Hilary on, therefore, Latin Pro-Nicene theologians understood that God was immaterial, but they initially attempted to explain the relation between the Son and the Father by using material categories and so were unable to convey how the immaterial Father and Son related to each other. As a result, they had to develop immaterial categories for describing that relation. There is a parallel here between the account of his conversion in the Confessions, in which Augustine moves from a material conception of God to an immaterial one, but it is important not to confuse his conversion story with the story of how Latin trinitarian theology developed.24 By the time he wrote De Trinitate 5 Augustine was a fully engaged Pro-Nicene polemicist. Building on Ambrose Augustine also recognized that the father-son analogy would have to be entirely recast by firmly denying that the names reveal substance; thus the materialist connotations of the analogy no longer apply. Neither Hilary nor Ambrose managed to explain the Father-Son relation adequately, however, so to find the Latin Pro-Nicene solution to this problem, I turn to Augustine.

The Names "Father" and "Son" in De Trinitate 5

The precipitating cause of Augustine's De Trinitate 5 is an argument put forward by the Homoians about the substantial difference between the Father and the Son. (25) The Homoians maintain that everything said about God is said "substance-wise." This means that the Father is "unbegotten" in his substance, while the Son is "begotten" in his substance, because the words "unbegotten" and "begotten" reveal the substance of God. In the same way, the names "Father" and "Son" tell us something about who the Father and Son are in their substance: the Father is father by substance and the Son is son by substance. Therefore, the Father and the Son have different substances. (26)

Augustine responds to this attack by denying that everything said about God relates to substance. Some things are said by way of reference to something else, he argues, so that when we call the Father "Father," we are only describing his relation with the Son, just as to call the Son "Son" pertains only to the Son's relation with his Father. In other words, the Father is called "Father" only as a way of delineating how the Father is not the Son and the Son is not the Father. Neither name relates to the divine substance, however. So, although it is one thing to be Father and another to be Son, there is no implied difference in substance between the two, because the names "Father" and "Son" reveal only their relation. By denying that the names refer to substance, Augustine attempts to remove the possibility of using the names to distinguish the Father's substance from the Son's. Augustine's argument allows them to have different names and the same substance.

Augustine anticipates at least two potential problems with his relational argument. The first is the old anti-Nicene argument that the Father-Son relationship indicates that the Son must have been generated by the Father in time, since fathers give birth to sons in time. Augustine counters this argument by asserting that the relation between the Father and Son is eternal, so the relation exists eternally. (27)

Augustine spends more time on the second potential problem. He moves from the names "Father" and "Son" to the related issue of what it means to call God "unbegotten." This move seems tricky for a Pro-Nicene like Augustine. The Homoians admit that "begetter" and "begotten" might well be relational terms in the same way that "father" and "son" are. "Begetter," in fact, is a synonym for "father," and "begotten" is a synonym for "son." The term "unbegotten," however, applies only to the Father, because there is no equivalent term for the Son; thus "unbegotten" is not relational, which means that "unbegotten" refers to the Father's substance. And because we cannot call the Son "unbegotten," the Homoians continue, we cannot claim that the Father and the Son share the same substance. (28)

Augustine counters this argument by attempting to reclaim "unbegotten" as a relational term. He begins by acknowledging that the Homoians may have a point and that the actual referent to "begotten" is not "unbegotten" but "begetter." Therefore, the relational terms that best correspond to Father and Son are most properly understood to be "begetter" and "begotten." (29)

This still leaves the question of how to understand "unbegotten." To resolve this Augustine insists that when we say "unbegotten" (ingenitum), we really mean "not begotten" (non genitum). There is nothing wrong with saying "unbegotten," but using "not begotten" instead helps us recognize the relational intent of both terms. If we call the Son "begotten" relationally, then our identification of the Father as "not begotten" must also indicate his relation to the begotten, so that neither "unbegotten" nor "not begotten" reveals anything about the Father's substance. (30)

Augustine's point is that using "unbegotten" is dangerous because it seems to be a formal name for God, so that using it might mislead people into thinking that it is a name that applies to God's nature. But because no one thinks "not begotten" is a proper name for God, its use would be preferable, especially as a way of defining what we mean by "unbegotten." Thus "begotten" and "unbegotten" are relational terms. They reveal nothing of the divine substance, but tell us only how the two Persons relate to each other. (31)

Using the Tradition: Augustine on Birth and Eternal Generation

To show the extent to which Augustine's response to the name question infiltrates his trinitarian theology as a whole, I turn to his Contra Maximinum. (32) This work, written in the late 420s, is a difficult text to analyze because in much of it Augustine uncritically repeats earlier Latin Pro-Nicene formulations. (33) The pro forma character of this debate is interesting in part because it proves that Augustine knew the older polemical tropes. This text is useful, however, also because it shows, despite its reliance on the older polemics, Augustine employing the new Latin Pro-Nicene polemical tradition he had helped establish.

In book 2, for example, Augustine replies to Maximinus's contention that when Pro-Nicenes say that the Son is born of the Father, they mean it materially, in the same way that bodies are born of bodies. (34) Augustine initially responds by denying that this is what Pro-Nicenes mean when they talk about the "birth" of the Son.

Flesh is born of flesh; an offspring of the flesh is born of the substance of flesh. Remove from the discussion any corruption: cast aside from the light of the mind any carnal passions, and see the invisible reality of God through these things that have been made. The Creator gave to the flesh the ability to generate flesh, gave to parents the ability to generate true offspring from the substance of the flesh so that the offspring have one substance with their parents. Believe that he could, all the more, have begotten a true Son from his substance and could have one substance with his true Son, without any loss of spiritual incorruption and utterly removed from carnal corruption. (35)

Augustine then insists on the importance of substance and thus of homoousios, because without an account of the substantial relation between the Father and the Son, he cannot see how it is possible to avoid saying that the Son is "from nothing." The Homoians, who want to distance themselves from Arianism proper, deny that the Son is "from nothing." But according to Augustine, this position creates a logical difficulty because there are only two options for understanding the Son's begetting: either the Son was born from the Father's substance, or he was born from nothing. The latter is untenable to both Pro-Nicenes and Homoians, while the former is what Pro-Nicenes already believe. (36)

That this entire argument could have been lifted from Hilary's early anti-Homoian work and that Augustine returns again and again in Contra Maximinum to the idea that the Son was born from the Father's substance indicates the pervasiveness of this theme in Pro-Nicene polemics against Homoian theology. That Augustine uses the father-son motif is, again, surprising since he had distanced himself from it in De Trinitate 5. There are passages in the Contra Maximinum, however, that also demonstrate Augustine's awareness of how his Latin Pro-Nicene tradition had changed--indeed, how he had changed it. In Contra Maximinum 13, for example, Augustine examines the question of how the Son is wise in relation to the Father. Maximinus apparently took Romans 16:27 ("God who alone is wise") as proof that the Father and the Son did not share a common wisdom. Augustine has a ready answer for this: since the Romans passage does not say that the Father alone is wise but that God is wise, we can understand the passage to be talking about the entire Trinity. (37) While this response is straightforward enough, Augustine has more trouble with the next phrase in the Romans passage. The passage goes on to say that God is wise "through Jesus Christ." This phrase seems to imply that God becomes wise only by begetting Jesus Christ. However, Augustine argues, we should not take it that way, because God the Trinity is wise substantially, so that it would be impossible for the Son to make the Father wise through begetting the Son. Unlike rational creatures, Augustine concludes, wisdom does not come to the divine substance as something accidental; it belongs to the divine substance itself. (38)

Augustine's interpretation of the "through Jesus Christ" phrase is an abbreviated version of an argument he developed in the earlier De Trinitate. Indeed, in De Trinitate 6, Augustine interprets 1 Corinthians 1:24 ("Christ, the power of God and the wisdom of God") by employing the substance-accident distinction in a nearly identical way. He begins this book by noting that the 1 Corinthians text has a long history in Pro-Nicene thought, having been used by early Pro-Nicenes to prove that the Son was eternal with the Father: if the Son is the wisdom of God, and if God is never without wisdom, then the Son is eternal. (39) This argument creates a problem, however, because it seems to put the names "wisdom" and "power" in the same relational category as "son," which is problematic, because making "wisdom" relational suggests that the Father is wise only because he begets his wisdom, just as he is Father only because he begets the Son. (40) But if the Father is wise only in the begetting of his wisdom, or is Father only in the begetting of his Son, then he is not God without that wisdom or Son, which leads to the unsatisfying conclusion that the Son is indistinguishable from the Father. Thus we are left with modalism. (41)

Augustine's assumption is that the scriptural language makes it easier to distinguish between the Father and Son than to recognize their substantial unity. And when Augustine turns to John 1:1 in De Trinitate 6.3, we can get a sense of how the old logic no longer works: if we say that the Father is not God without the Son, then we must understand "in the beginning was the Word" as "in the Father was the Word." The Word would thus be an internal condition of the Father's being, with no necessary preexistence or distinction from the Father. Augustine goes on to suggest that there is an equally serious problem with the next phrase in the Prologue, "and the Word was with God." This phrase indicates that the Father is God and the Son is the Word, but if the Father and Son are God together, then there could be no "God" for the Word to be with, as it were. This rather unsettling realization leads Augustine to reflect on the "x from x" (e.g., "light from light") language in the creed, because that language seems to have some logical problems of its own. For one thing, they are not both together "God from God" or "light from light," because only the Son is from God. For another, if we have God only when we have both the Father and the Son, then there is no God for the Son to be from, as it were. Augustine's solution to this conundrum is to limit the scope of this language. If we recognize that the "x from x" language is intended merely to indicate the eternity of the Son with the Father, he argues, then we can more readily recognize which names belong to substance and which are strictly relational. The basic rule is this: if the name refers to something the Father and Son are together, then it refers to something substantial; if the name refers to something that cannot be said of both Father and Son together, then the name is relational. (42)

Augustine will refine his theory in book 7, noting, among other things, that everything called by a relational name is also something substantial. It is therefore not inconsistent to use both relational and substantive names when talking about God (7.2); thus it is acceptable to say that the Word is not the Father, but that the Father and the Son are together one wisdom (7.3). (43) As Ayres has shown, Augustine proceeds from the notion that "in God all qualities are identical with God's essence," which means that there is no way to dissociate any aspect of God from the divine substance, including the relations. (44) From this perspective, the task of trinitarian theology is to find a proper grammar for expressing the unity of the Son and the Father in a way that does not negate their difference. (45) Part of my argument here shows that the importance of divine simplicity in Augustine and Latin Pro-Nicene thought as a whole may have resulted from the process of answering the Homoians. The gradual move away from an emphasis on the names "Father" and "Son" forced the Latin theologians to articulate a more sophisticated account of what it means for God to be simple.

In other words, Augustine has established a kind of hermeneutic that allows us at once to make sense of the various names given to the Son, interpret key Scripture passages such as John 1:1, and understand creedal language from a properly Pro-Nicene perspective. What pertains to substance we can describe using "x from x" language; all other names are relational because they refer not to God's substance in itself but to what distinguishes the Son from the Father. (46) The value of this hermeneutic for Pro-Nicene thought is that it preserves the distinction between the Father and the Son while giving Pro-Nicenes language for expressing how the Son is eternal, shares the Father's substance, etc. This hermeneutic, in fact, suggests that Augustine limits the use of "x from x" language to articulating the coeternality of the Son with the Father, because eternal generation was a key point of emphasis in earlier Latin Pro-Nicene thought. Augustine does not, however, use the names "Father" and "Son" to demonstrate eternal generation, as Hilary had. Augustine allows those names to point us to the kind of begetter-begotten relation that obtains between the Father and the Son, but he also insists that we can express the substantial dimension of that relation only when we employ a proper grammar, such as "x from x."

Throughout this discussion Augustine is reluctant to speak about the divine nature, and this reluctance may be our best clue to how he used the name argument. He believes we must qualify whatever language we use for God, whether we use the names "Father" and "Son" or even creedal language such as "x from x." Augustine's reluctance extends to the second half of De Trinitate, books 8-15. He famously struggles in these books to find an analogy to explain how God can be inseparably three and one, finally arriving at "memory, intelligence, and will." At the end of De Trinitate, however, Augustine pulls back from the entire search for an analogy and reminds his readers of how dissimilar the mind analogy is to the reality of God's being:

To the memory, sight, and love of this supreme Trinity, in order to recollect it, see it, and enjoy it, [the reader] should refer every ounce and particle of his life. But I have sufficiently warned [the reader], so it seems to me, that this image, made by the Trinity and altered for the worse by its own fault, is not so to be compared to that Trinity that it is reckoned similar to it in every respect. Rather, [the reader] should note how great the dissimilarity is in whatever similarity there may be. (47)

The corporeal mind, even if it is--or perhaps because it is merely--the image of God, cannot show us who God really is. Our ultimate "sight" of the Trinity awaits the resurrection, when our hearts are purified and we are not weighed down by our corporeal bodies. (48)

The neo-Platonic overtones here are unmistakable, but it is important to note that Augustine is not forbidding the use of analogy. Just prior to the passage where he qualifies the usefulness of the analogy, he reaffirms that the search for analogies is legitimate, provided that analogies are kept under control. His instinct to pull back from the mind analogy, therefore, is not so much the result of a philosophical commitment as it is an instinct that became an important part of Latin Pro-Nicene thought. The Latin Pro-Nicene insistence on qualifying our use of the names "Father" and "Son" led Augustine to what we might call a "trinitarian apophaticism." This permits the search for knowledge of God in the created order, even while it always limits how deeply these analogies can penetrate the divine nature. (49) This reluctance to speak about God's nature may be what ultimately distinguished Latin Pro-Nicene trinitarian thought from that of its Homoian opponents. (50)

CONCLUSIONS

I have tried to describe the process by which Latin trintitarian theology moved from a naive materialism in describing the relation between the Father and the Son to a position that embraced the immateriality of God's nature in order to preserve the Father and Son's eternal, "substantial" relation. As a result of this process, attempts to describe the relation between the Father and Son became increasingly removed from the divine nature: first in Hilary's reluctance to use human analogies to describe the divine substance, then in Ambrose's distinction between the Father and Son's common nature and their "personal properties," to Augustine's theory that the only way to preserve the eternality of the Son is to deny altogether that the names "Father" and "Son" pertain to the divine nature.

This evidence highlights the degree to which Augustine was embedded in his tradition. That Augustine feels compelled at so many points to correct his tradition--and not just in order to make it more theologically viable, but to enhance its polemical effectiveness--is evidence enough of this point. But even more importantly, Augustine's trinitarian theology as a whole is shaped by his appropriation of Latin Pro-Nicene thought; the trinitarian grammar he establishes in De Trinitate 5 reemerges throughout De Trinitate. I have explored some of the connections between books 5 and 6 and the "trinitarian apophaticism" in book 15, connections that suggest the likelihood that a more thorough investigation of De Trinitate from this perspective would prove fruitful. Such an investigation awaits further research, but one potential line of inquiry would be to examine the relationship between Augustine's christological grammar and his trinitarian grammar. I have already suggested that these grammars are similar in form, and it seems likely that the correlation between the way Augustine speaks of God and the way he speaks of the incarnate Son reveals much about the shape of his thought in De Trinitate.

Perhaps the old thesis that "Augustine emphasized the unity" in his trinitarian theology is no longer credible; certainly the evidence I have presented here adds to the refutation of this approach to Augustine's thought. (51) The "name theology" that initially formed much of the foundation for Latin Pro-Nicene thought began--to use the old model--with the difference as indicated by the names "Father" and "Son." Pro-Nicenes were initially attracted to speculation about the natural quality of the names because they offered a way to express the substantial unity of the Father and Son in light of their difference. Though Latin Pro-Nicenes eventually realized that there were serious problems with the initial formulation of the "name theology," they never abandoned the project of explaining the unity in light of the diversity. Latin Pro-Nicenes, including Augustine, always knew that the Father and Son were distinct "somethings." The question was always how to articulate their substantial unity. The Pro-Nicene emphasis on the Son's eternal generation also reflects this concern for finding unity amid diversity. As with the "name theology," the question of eternal generation presumes that there are two persons, Father and Son, and the point of defending the Son's eternal generation is to unite the Son with the Father's substance.

(1) Scholars who have examined Augustine's theory of naming in De Trinitate 5 have tended to dismiss its importance. This dismissal may be partially related to the modern tendency to deemphasize the first seven books of the work; but even scholars who take books 1 to 7 seriously tend to dismiss the section on naming in De Trinitate 5. Edmund Hill, for example, claims that, while the distinction Augustine makes between "substantive and relative predication" may be his most lasting contribution to Western trinitarian theology, it is also "not his cardinal insight into the Trinitarian mystery" (Saint Augustine: The Trinity [New York: New City, 1991] 202 n. 14). Luigi Gioia takes this position even further, claiming that, had Augustine's opponents not introduced this topic, Augustine "would not have felt the need to dwell [on it] to the same extent" (The Theological Epistemology of Augustine's De Trinitate [New York: Oxford University, 2009] 149). Gioia's dismissal of Augustine's argument is surprising, given the relation between naming and epistemology in Pro-Nicene thought as a whole. Possibly what Augustine does in De Trinitate 5 supports Gioia's otherwise compelling perspective on the priority of epistemology in De Trinitate.

Scholars who have examined Augustine's thought in De Trinitate 5 have tended to approach it from a philosophical perspective, either exploring the Aristotelian roots of Augustine's use of accidental categories for the divine substance, or asking how well Augustine appropriated those categories. For a notable example of the latter, see especially Roland Teske, To Know God and the Soul (Washington: Catholic University, 2008) 93-111. Teske shows that Augustine articulates a rule for "distinguishing what is said of God relatively from what is said of God non-relatively" in a philosophically astute way (95). Teske deliberately avoids addressing either the theological background to this philosophical insight or its trinitarian implications, and so his work is only of peripheral interest to my purpose here. For more on the Aristotelian background to Augustine's categories of naming in De Trinitate 5 see Basil Studer, Augustins De Trinitate: Eine Einfuhrung (Paderborn: Ferdinand Schoningh, 2005) 136-37.

(2) Some scholars have remarked on an active debate in the East during the fourth-century controversies about whether names were accidental or natural. Eunomius in particular argued for the natural quality of names, and both Basil of Caesarea and Gregory of Nyssa argued that names are accidental. For discussion of the philosophical issues in fourth-century discussions of naming the divine substance, see Raoul Mortley, From Word to Silence, 2 vols. (Bonn: Peter Hanstein, 1986).

(3) Hilary was exiled to Cappadocia in 358, most likely for refusing to condemn Athanasius. While in exile he may have learned Greek: he certainly learned a new way of thinking about the Trinity. When he returned to the West around 361, he became one of the leading Pro-Nicenes in the West. His influence continued long after his death: in De Trinitate 6 Augustine cites Hilary as an authority. For an examination of Hilary's trinitarian thought both before and after his exile, see Mark Weedman, The Trinitarian Theology of Hilary of Poitiers (Leiden: Brill, 2007).

(4) Ibid. 136.

(5) Hilary did not use the names "Father" and "Son" in this way prior to his exile. While in exile he encountered Homoiousion thought, and some of his first polemical works from this period reflect that influence, especially his De synodis and large sections of De Trinitate. See Weedman, Trinitarian Theology of Hilary. As I show below, Hilary begins to distance himself from the Homoiousian perspective as he begins to realize its vulnerabilities to Homoian attack.

(6) Epiphanius, Panarion 73.3.2-3; Epiphanius, vol. 3, Panarion haer. 65-80: De fide, ed. Jurgen Dummer (Berlin: Akademie, 1985) 271 (hereafter referred to as Dummer); Engl. trans, in Epiphanius, The Panarion of Epiphanius of Salamis, 2 vols., trans. Frank Williams (New York: Brill, 1987) 2:436 (hereafter Williams).

(7) Epiphanius, Panarion 73.3.1; Dummer 271.

(8) Epiphanius, Panarion 73.3.3: Dummer 271; Williams 2:436.

(9) Hilary must have written book 12 only a short time after writing book 7, so the process I describe here must have happened very quickly. On Hilary's composition of De Trinitate, see Carl Beckwith, Hilary of Poitiers" on the Trinity: From De Fide to De Trinitate (New York: Oxford University, 2008).

(10) Hilary, De Trinitate 12.22 (Corpus Christianorum, Series Latina [hereafter CCSL] 62a.596).

(11) Ibid. 12.8 (CCSL 62a.585).

(12) Hilary addresses the question of analogies in two earlier places in De Trinitate (1.19 and 6.9). If, as Beckwith argues, Hilary added these reflections on analogy well after the original time of composition, we must suspect that this material reflects insights he came to late in the composition of the work--and in response to Homoian polemics against the father-son analogy. See Beckwith, Hilary 94.

(13) For Hilary's doctrine of divine infinity, see Mark Weedman, "The Polemical Context of Gregory of Nyssa's Doctrine of Divine Infinity," Journal of Early Christian Studies 18 (2010) 92-96.

(14) Hilary, De Trinitate 12.24 (CCSL 62a.597).

(15) Ibid. 12.25 (CCSL 62a.598).

(16) See ibid. 12.25 (CCSL 62a.598).

(17) For Ambrose's trinitarian theology and his engagement with the Homoians, see Daniel H. Williams, Ambrose of Milan and the End of the Arian-Nicene Conflicts (Oxford: Clarendon, 1995); and Christoph Markschies, Ambrosias yon Mailand und die Trinitatstheologie: Kirchen- und theologiegeschichtliche Studien zu Antiarianismus und Neunizanismus bei Ambrosius und im lateinischen Westen (364-381 n. Chr.) (Tubingen: Mohr, 1995).

(18) Ambrose, De fide 4.8.82 (Migne, PL 16.633): "Generatio enim paternae proprietatis est, non potentiae."

(19) Ibid. 4.8.87 (PL 16.634).

(20) Ambrose's argument is in some ways even closer to Gregory of Nyssa's than to Augustine's. Gregory articulates the difference between the Father and the Son by distinguishing the category of being (ousia) from that of power (dunamis). For Gregory, the Father, the Son, and the Holy Spirit are each "powers" of the divine being, which means that they are inseparable from that being without revealing it. Ambrose's terminology lacks Gregory's precision. By conflating substance with power, Ambrose struggles to find an adequate category by which to distinguish between the Father and the Son. Nevertheless, the fundamental methods of Gregory and Ambrose are the same: to answer Homoian and Eunomian attempts to subordinate the Father and the Son on the basis of the names "Father" and "Son," Pro-Nicenes increasingly came to separate the experience of fatherhood and sonship from the level of being or substance. For an important account of Gregory's use of ousia and dunamis, see Michel Barnes, The Power of God: Dunamis in Gregory of Nyssa's Trinitarian Theology (Washington: Catholic University, 2001) 220-59.

(21) Ambrose, De ride 4.8.92 (PL 16.635); Engl. trans, from Ambrose, A Select Library of Nicene and Post-Nicene Fathers of the Christian Church, 2nd ser., 14 vols., ed. Philip Schaff et al. (Grand Rapids, Mich.: Eerdmans, 1956-) 10:274.

(22) See Ambrose, Deride 4.9.104 (PL 16.636).

(23) See ibid. 4.9.106 (PL 16.637).

(24) It is true that the longer movement from Tertullian to Augustine involved a transition from a material to an immaterial conception of God. For the early manifestation of Latin thought, including its material view of the divine substance, see Jean Danielou, The Origins of Latin Christianity, trans. David Smith and John A. Baker (Philadelphia: Westminister, 1977). I would argue that we must locate the impetus for the Latin move to an immaterial conception of God to their involvement in the Homoian controversy. To describe successfully the Son's eternal generation, full equality with the Father, etc., while preserving his difference from the Father, Latin Pro-Nicenes had to find categories that allowed them to remove material considerations from their description of the Father-Son relation. My account of the development of Latin thought highlights the fact that Augustine was a Latin Pro-Nicene polemicist who drew from, even as he developed, the standards of Latin Pro-Nicene polemics that he inherited from Hilary and Ambrose.

(25) In addition to the bibliography cited above, readers should consult Lewis Ayres, Augustine and the Trinity (Cambridge: Cambridge University, 2011) 211-29 for a nuanced and crucial exploration of Augustine's thought in De Trinitate 5- 7. Ayres is interested in the implications of Augustine's theory of predication for his trinitarian ontology and his account of divine simplicity, how Augustine uses this theory to reject person and nature language, and how Augustine's theory develops his earlier trinitarian formulations. My purpose is to trace the process by which Augustine and other Latin Pro-Nicenes came to move toward my account here of how to predicate names to God, and I have chosen to emphasize a slightly different set of passages within books 5 to 7 that seem to highlight ways that Augustine has responded to his theological and polemical tradition.

(26) Augustine, De Trinitate 5.4 (CCSL 50.208); Engl. trans, in Hill, ed., Saint Augustine 191. For the identity of Augustine's opponents in De Trinitate as Homoian and not Eunomian or Arian, see Michel R. Barnes, "The Arians of Book V, and the Genre of De Trinitate," Journal of Theological Studies 44 (1993) 185- 95. Barnes has produced several important works on the polemical context of Augustine's trinitarian theology; besides his "Arians" article, see his "Re- reading Augustine's Theology of the Trinity," in The Trinity: An Interdisciplinary Symposium on the Trinity, ed. Stephen T. Davis, Daniel Kendall, and Gerald O'Collins (New York: Oxford University, 1999) 145-76; and "Exegesis and Polemic in Augustine's De Trinitate I," Augustinian Studies 30 (1999) 43-60.

(27) Augustine, De Trinitate 5.6 (CCSL 50.210).

(28) Ibid. 5.7 (CCSL 50.211).

(29) Ibid.

(30) Ibid. 5.8 (CCSL 50.214).

(31) Augustine's move to emphasize the relational quality of the names "Father" and "Son" mirrors a similar move in Eastern Pro-Nicene thought during the latter part of the fourth century. Basil of Caesarea, e.g., argues that the names "Peter" and "Paul" tell us only about the properties that characterize each individual. This means that the name "Peter" does not necessarily tell us anything about Peter's substance, only that he is not Paul. Peter and Paul share the same substance, but that fact is not indicated by their names. In the case of the names "Father" and "Son," then, we find that they tell us only about the individual properties that belong to each. As with Peter and Paul, the names "Father" and "Son" tell us nothing about the divine substance, which means that no one can introduce diversity into the substance of the Father and Son simply on the basis of their names (see Basil of Caesarea, Contra Eunomium 2.5). Basil's emphasis on the conventional, relational quality of the names "Father" and "Son" becomes a key component in Cappadocian trinitarian theology. Gregory of Nyssa, in fact, uses the relative value of names to discuss a variety of ideas and categories, including analogy, causality, epinoia (idea or conception) and language itself. For a full account of how Gregory intertwines these concepts, see Raoul Mortley, From Word to Silence, vol. 2, The Way of Negation, Christian and Greek (Bonn: Hanstein, 1986) 147-59; 178-91. Basil of Caesarea had access to a theory of naming that emphasized their natural quality, and like their Latin Pro-Nicene counterparts, Basil and his Cappadocian allies move to a relational view of the names, in large part because of polemical pressure brought against them by their anti-Nicene opponents, primarily Eunomius of Cyzicus. For the intellectual and polemical context of Basil's theory of naming, see Mark DelCogliano, Basil of Caesarea's Anti-Eunomian Theory of Names: Christian Theology and Late-Antique Philosophy in the Fourth Century Trinitarian Controversy (Leiden: Brill, 2010). This is not the place to assess the correlation between Cappadocian and Augustinian trinitarian theology, but there is in each case a common movement toward a nonmaterial, relative theory of the divine names.

(32) Maximinus, a Latin Homoian leader, flourished during the early part of the fifth century. The record of his debate with Augustine on the Trinity and Augustine's response are extant. For the story of Augustine's dealings with Maximinus and Latin Homoianism in general, see William A. Sumruld, Augustine versus the Arians: The Bishop of Hippo's Encounters with Ulfilan Arianism (London: Associated University, 1994), esp. 85-120; and Roland Teske, "Augustine, Maximinus, and Imagination," Augustiniana 43 (1993) 27-41.

(33) My thanks to an anonymous reviewer for pointing out how difficult Contra Maximinum is to read in the context of Augustine's polemical Trinitarian theology.

(34) Augustine, Contra Maximinum 2.14.2 (PL 42.771).

(35) Ibid. 2.14.3 (PL 42.771; Engl. trans, in Augustine, Arianism and Other Heresies, The Works of Saint Augustine, pt. 1, vol. 18, trans. Roland Teske [New York: New City, 1995] 282).

(36) Augustine, Contra Maximinum 2.14.4 (PL 42.72).

(37) Ibid. 2.13.1 (PL 42.769).

(38) Ibid. 2.13.2 (PL 42.770).

(39) Augustine, De Trinitate 6.1 (CCSL 50.228).

(40) Ibid. 6.2 (CCSL 50.229). 41 Ibid. 6.3 (CCSL 50.230).

(42) Ibid. 6.3 (CCSL 50.230).

(43) Augustine, De Trinitate 7.2-3 (CCSL 50.249-50).

(44) Ayres, Nicaea 377. See also Ayres, Augustine and the Trinity 221-27 for a full account of how Augustine develops the theme of "Wisdom" in these passages.

(45) Ayres, Nicaea 381.

(46) There is a deep similarity between this hermeneutic and the Pro-Nicene practice of assigning some statements and actions of the incarnate Son to his divinity and some to his humanity, so it is perhaps not surprising that Augustine articulates this christological hermeneutic almost immediately after this section in De Trinitate 6.5.

(47) Augustine, De Trinitate 15.39 (Hill, Saint Augustine 426).

(48) Ibid. 15.44 (Hill, Saint Augustine 429).

(49) John Cavadini locates Augustine's search for analogies as an exercise in education, one that focuses on developing the image of God that is embedded in the human nature. As Cavadini shows, although this search for analogies will ultimately end in failure, it is still educational and leads to a theological perspective that is "'neither apophatic nor kataphatic, but specifically trinitarian in its call neither to renounce language nor to accept its limitations as received" ("The Quest for Truth in Augustine's De Trinitate," Theological Studies 58 [1997] 429-40, at 440).

(50) Latin Pro-Nicenes' move away from analogies may also add credence to a growing scholarly consensus that Augustine's trinitarian theology was more apophatic than commonly believed. For an examination of his apophaticism, see Jean-Luc Marion, "Idipsum: The Name of God according to Augustine," in Orthodox Readings of Augustine, ed. George E. Demacopoulos and Aristotle Papanikolaou (New York: St. Vladimir's Seminary, 2008) 167-90.

(51) See Ayres, Nicaea 300.

MARK WEEDMAN received his Ph.D. from Marquette University and is now professor of biblical and historical theology at Crossroads College, Rochester, Minn. His area of special interest is early Christian theology, especially the trinitarian controversy. He has recently published The Trinitarian Theology of Hilary of Poitiers (2007); and "The Polemical Context of Gregory of Nyssa's Doctrine of Divine Infinity," Journal of Early Christian Studies 18.1 (2010). In progress is a work on a rereading of Augustine's Confessions that locates it as an extension of particular concerns in early Latin theology.