And the Son

Two dedications in as many days!

I dedicate this essay to Dr. Michael Liccione on the occasion of his birthday, 4 December 2006. May he celebrate many more!

There is an interesting dialogue going at among some of the (in my opinion) best bloggers on the so-called Filioque controversy. (See Sacramentum Vitae, Energies of the Trinity.) What follows is an essay that I wrote principally about Saint Anselm's treatment of this issue at the Council of Bari but also in defense of the Latin position on the Filioque. I do not see that Saint Anselm's argument can be anything but sound. It is certainly valid, and to dispute its premises would, in effect, be to beg the question against the West.

In 1978 Richard Cartwright gave a talk at North Carolina State University called “On the Logical Problem of the Trinity,” in which he argued that little progress, if any, had been made since the time of the Cappadocian Fathers of the fourth century in understanding the logical relations existing among the Persons of the Trinity and that this lack of understanding suggests that the doctrine of the Trinity is itself logical nonsense. Cartwright’s method in his presentation (subsequently published in his collected Philosophical Essays) was to take as a starting point the doctrinal elements of the Quicumque Vult, the so-called Athanasian Creed, a text of uncertain origins dating roughly to the latter half of the fifth century, and to show how these elements, when construed as delimiting logical relations, cannot be reconciled with one another on key points of logical consistency. Eight hundred and eighty years earlier, at the Council of Bari in southern Italy, Saint Anselm had adopted a remarkably similar methodology but with strikingly different results. Instead of finding the doctrine of the Trinity incomprehensible, he found it to be so clear and distinct as to lead inevitably to precisely that understanding that in the West had prompted the inclusion, in the Niceno-Constantinopolitan Creed used in the Mass, of the Filioque clause that had contributed to the Great Schism between East and West, traditionally dated to 1054 but in reality a long-simmering dispute over Church governance as much as theology. It is not entirely clear why what was so indubitably clear to Saint Anselm was so unfathomable to Richard Cartwright, though to the intervening nine centuries we may add in explanation the consideration that, while Saint Anselm summarized his work as “faith seeking understanding”, Richard Cartwright’s work can best be summarized as “putting the anal into analytic philosophy”. In this essay I intend to bridge this gap by putting the analytic into Saint Anselm’s faithful understanding. I intend to show that Saint Anselm’s treatise On the Procession of the Holy Spirit, written, as it seems, in an attempt to make the arguments he employed at the Council of Bari available to a wider audience, provides a detailed and logically rigorous defense of the consistency of the doctrine of the Trinity, a defense grounded in metaphysical principles that ought to have been both familiar and acceptable to his Greek opponents, since these principles were themselves ultimately of Greek philosophical origin.

The principal logical objection to the doctrine lies in the claim that, if we are to make any metaphysical sense of the Trinity, we are forced either to divide the substance of the Godhead or to confound the Persons of the Trinity. All the Credal formulations accepted by the Church agree that there is only one God and that this God is just one substance—ousia in the Greek Creeds, substantia in the so-called Athanasian Creed, which has only Latin versions in its earliest surviving forms. These same Creeds also agree, however, that God is three Persons, each Person being distinct from the other two. Since there is only one God who is one substance, then it is unclear in what way the three Persons are to be regarded as distinct from one another. The so-called Athanasian Creed says that “what the Father is, such is the Son and such the Holy Spirit.” If taken at face value this claim appears to suggest something like the identity of indiscernibles, since it appears to be saying that there is no property that is not shared by all three Persons. A simple and straightforward assessment of property relations as governed by the Law of Noncontradiction suggests that it is not possible for things to be both identical in this sense and distinct at the same time and in the same respect, so either the three Persons are distinct and we must divide the substance of the Godhead, or else they are identical in a traditional sense and we must confound them with one another, which is heresy. Since it is also heresy to divide the substance of the Godhead, things don’t look too good for the faith seeking understanding.

The treatise De processione Spiritus Sancti begins with a brief synopsis of what the Greeks and the Latins agree upon: there is but one God, God is also three Persons, and the relations that exist among the three Persons are in some ways symmetrical and in other ways asymmetrical. The symmetrical relations include a kind of modified identity: if God is the Father, then the Father is also God; if God is the Son, then the Son is also God. The asymetrical relations are intended to prevent the symmetrical relations from entailing heresy: although the Father is God and the Son is God and there is only one God, nevertheless the Father is not the Son and the Son is not the Father. The symmetrical relations are thus strictly logical in character, while the asymmetrical relations are theological in character but secondarily also logical, that is, they conform to certain formal principles.

Arguably the most important of these formal principles is what we may call the Principle of Opposition, a principle whereby the relations in the Trinity are kept from contradicting one another by limiting their scope in such a way that consistency is virtually guaranteed, since these relations will have logical entailments only “where this is not impeded by the opposition of relation”:

The unity should never lose its consequences except when a relational opposition stands in the way, nor should the relations lose what belongs to them except when the indivisible unity stands in the way (Quatenus nec unitas amittat aliquando suum consequens, ubi non obviat aliqua relationis oppositio, nec relatio perdat quod suum est, nisi ubi obsistit unitas inseparabilis, DPSS 1, p. 181.2-4 S).

This principle is itself of Greek origin and can be traced as far back as Gregory of Nazianzus, who clearly enunciates it in Orations 20, 31, 34, and 41: there is complete identity among the three divine Persons except for the relations of origin. I shall have cause to return to the conception of identity being deployed here presently, but first a word or two about the Principle of Opposition itself. According to this principle, the unity of the Godhead serves as a primitive; all consequences of unity as such are permitted unless they contradict other fundamental relations, and all fundamental relations are permitted their consequences except in cases where an entailment would contradict the unity of the Godhead. The fundamental relations are those that define the three Persons with regard to the manner in which they stand relative to one another as source of being: the Father stands as origin or source with respect to the other two Persons in two distinct ways, as begetter (of the Son) and as source of procession (of the Holy Spirit); the Son stands as the begotten of the Father; the Holy Spirit stands as that which proceeds from the Father. At this early point in the treatise Anselm does not specify in what sort of relation the Son stands to the Holy Spirit since he is sensitive to the danger of begging the question against the Greeks, but he does note that the Greeks agree both that the Holy Spirit is “God from God” (“Deus de Deo”) and that the Son is God, hence, because all sides agree that the Son sends the Holy Spirit, there is some sense in which the Holy Spirit is “from” God the Son. What the Greeks specifically objected to in this was the suggestion that the Holy Spirit was from God the Son as his origin or source of being—they insisted on a principle of monarchia according to which only God the Father could serve as an origin or source of being. In some circles monarchianism became a heretical movement that denied the subsistence of the Son, but in the sense intended here it is a perfectly orthodox concept and, indeed, it is not explicitly denied by Anselm; as we shall see, there is actually a limited sense in which he depends upon it.

Putting this point aside for the time being, it is worth taking a moment to examine the application of the principle of opposition to the relations that are agreed to by the Greeks. The Father begets the Son and the Son is begotten by the Father, but the Father and the Son are one in the sense that they are both God and there is only one God. The principle of opposition prevents us from saying that the Father is the Son or that the Son is the Father, even though this follows logically from the unity of the Godhead, because to say that the Father is the Son or that the Son is the Father would be to say that the begetter is also the begotten, which Anselm takes as even more incomprehensible than saying that the Father and the Son are both one and distinct. While distinguishing between degrees of incomprehensibility in this way may seem to us, well, somewhat incomprehensible, there are two points to keep in mind here. The first is that the Greeks do not disagree with the Latins that the Father and the Son are both one and distinct, or that the Father and the Spirit are both one and distinct, or that the Son and the Spirit are both one and distinct. What they disagree about is the precise nature of the distinction between the Son and the Spirit. So the strictly logical difficulty of saying that the begetter is also the begotten would be a far greater difficulty for them than the theological difficulty of comprehending the nature of the Trinity, which, for all parties involved, had been a matter of settled doctrine since the fourth century. Second, it is clear from examples that Anselm gives later in the treatise that the combined unity and distinctness of the Persons did not strike either him or the Greeks as all that perplexing, since he compares the situation to both complex theological analogies such as that given by Saint Augustine in his treatise De Trinitate, where he compares the Godhead to the unity of the three distinct elements in human thought, namely thinking subject, thought, and process of thinking itself, and to the homey example that apparently had its origin with the Greeks attending the Council of Bari, the unity of a spring, a stream flowing from that spring, and a pond into which that stream flows.

Let me return for a moment to the point that I mentioned in passing just now, the fact that the Greeks did not dispute the fact that the Spirit is from the Son in some sense. The polemical strength of Anselm’s treatise, as we shall see, turns on a metaphysical principle that the Greeks ought to have accepted but did not, but it is worth noting that the semantics of the doctrine tended to work against the Latins, who did not possess the rich theological language in which the doctrine of the Trinity had been defined at the Councils of the fourth century. In his own treatment of this debate in Addresses to One of his Pupils about What the Latins Allege, possibly written just prior to the Council of Bari or, at the latest, at about the same time as the De processione or shortly thereafter, Theophylact of Ochrid points out that the Latins had only one word to refer to multiple kinds of processions of the Holy Spirit. The Greeks did not deny that the Holy Spirit comes from the Son in some sense, but they were suspicious of the fact that the Latins used the very same word, procedere, to refer to the procession of the Holy Spirit from the Father and to the mode in which the Holy Spirit is from the Son. Theophylact points out that the Greeks use ekporeuesthai, khorêgeisthai, metadidosthai, and pempesthai to refer to modes in which the Holy Spirit is Deus de Deo. Of these it is the first, ekporeuesthai, that refers specifically to the sense in which he is ek monou tou Patros, “from the Father alone” and, hence, of these Greek terms only ekporeuesthai refers to the procession of the Holy Spirit in the sense of how he has his being, and it does so in a special way that is peculiar to the Spirit’s relation to the Father. The Latin term procedere, to the extent that it is a translation principally of ekporeuesthai, collapses distinctions that the Greeks believed to be matters of extremely important theological import.

Theophylact was extremely patient about the inability of the Latin language to do justice to the finial theological nuances of the doctrine of the Trinity, far more so than most Greek theologians of his day, who took advantage of theological excuses to promote their own anti-Latin Church polity. In his view it was important to allow for the particularities of local practices in the various Churches, especially in the East (he appealed to the case of the Bulgarian Church to illustrate how even in the East it is essential to insure the understanding of theological terms by the lay members of the community) but also, of course, in the Latin West, and many of the differences between East and West he ascribed to cultural differences that were due ultimately to differences in language and custom. Such things were quite permissible, in his view. He drew the line, however, when it came to matters that had been settled by the decision of an Ecumenical Council, and the text of the Creed fell under this rubric. With the insertion of the Fiolioque clause the Latin West had gone too far, and his treatise represents a systematic attempt to put a stop to that sort of innovation.

In this regard another principle underlying Anselm’s text is particularly significant, which I will call the Principle of Non-ampliative inference, which permits the addition to any body of credal statements another statement that follows deductively from the same body of credal statements. The polemical point underlying this principle is, obviously, that any statement that follows deductively from the statements already accepted will be non-ampliative, that is, it will not add anything new to the already existing body of credal formulae and, hence, will not count as a genuine theological innovation. Although it remains unstated there is a clear implication in Anselm’s text that the Latin position is that materials that do not constitute theological innovation may be harmlessly inserted into canonical texts, particularly if there are pastoral reasons for doing so, as the Latins believed to be the case in the matter of the Filioque clause. The Greek position, by contrast, was that the Filioque did count as an unsupportable innovation and that, whether or not it represented an innovation, it was nevertheless contrary to the dignity of Conciliar pronouncements to augment or supplement them in any way in their liturgical settings, even for pastoral reasons. It is this latter point that is the most decisive for the Greeks, because generally speaking the Greeks did not particularly object to non-ampliative theological statements. Gregory of Nazianzus, for example, says of the Persons in Oration 39:

Three individualities or hypostases, if any prefer to call them, or persons (prosôpon), for we will not quarrel about names so long as the syllables amount to the same meaning…

Anselm, in short, viewed his argument as simply carrying out an explication of what was already agreed to by all sides. From his perspective the liturgical change was a done deal whether or not it had been a good idea; his goal was merely to show that the sentiment expressed by the liturgical change was not heretical. But from the Latin point of view the liturgical change had been a good idea, because it was motivated by a concern over the Spanish adoptionist controversy. So on Anselm’s account the Filioque was (a) not an innovation but a logical consequence of Trinitarian doctrine and (b) an effective barrier against heretical innovation of a very dangerous kind.

Polemically Anselm seems to have regarded his task principally as a logical one, that is, he appears to take it for granted that if he can show how the procession of the Holy Spirit from the Son follows from Trinitarian doctrine accepted by all parties then the matter will be settled, at least for anyone who is rational. Liturgical matters can be left to the individual Churches, but doctrinal matters must be grounded in consensus, and consensus must be grounded in reason. In light of this it is curious that Anselm did not provide any analysis in the De processione of the concepts of identity and unity that are deployed in the doctrine of the Trinity as he understood it. According to the Principle of Opposition, the unity of the Godhead is a primitive, and this unity is to be understood in terms of the identity relations that exist between the three Persons and the Godhead itself: the Father is God, the Son is God, and the Holy Spirit is God, and there is exactly one God, so at the very least the Father is identical to God and the Son is identical to God and the Holy Spirit is identical to God. Beyond this the Principle of Opposition will not permit us to go, since the principle dictates that the Father may not be identical either to the Son or to the Holy Spirit, nor may the Son and the Holy Spirit be identical to each other—the three Persons must be distinct. And yet this distinctness of Persons is precisely what troubles one the most, from a logical point of view, in light of the unity of the Godhead. Identity relations as they are commonly understood would entail, from these starting points, that the Father, the Son, and the Holy Spirit are all identical to each other, and this suggests that there are not really three distinct Persons in the Godhead, but only three different aspects under which the concept of God can be understood, and this is clearly far too weak for Christian purposes, bordering, as it does, on the heretical view known as modalism.
Cartwright suggests that it is possible that Anselm endorsed, in De Incarnatione Verbi, a form of identity that was also defended by Peter Geach in his paper “Identity” (Rev. Met. 21 (1967) 3-12). On this view, we may formulate the identity relations of the Trinity in the following way:

If something is a God, then anything that is a God is the same God as it.

Cartwright describes this sameness relation as being similar to counting novels on a shelf:

We assign the number 1 to some novel on the shelf and to whatever on the shelf is the same novel; if, after this, no novel on the shelf remains unnumbered, there is exactly one novel on the shelf.

Clearly this description falls short of a definition of sameness since it appeals to the very notion under consideration as part of the description, but one gets the idea. However, it will be quite useless as a definition of identity in the Trinity since however one may like to count one’s books it is clear that there is a difference between type and token and tokens that are numerically distinct are distinct entities whether or not they are of the same type. Cartwright himself notes that this definition of sameness leads to doctrinal error if it is taken as an explication of the statement from the so-called Athanasian Creed that

What the Father is, such is the Son and such the Holy Spirit.

If one understands this in Geach’s sense, then we may infer that if the Father is a different Person than the Son, then the Son is a different person than the Son; or if the Father is unbegotten, then the Son is unbegotten. Clearly this will not do, and it is worth pointing out that while Cartwright cites the De Incarnatione by name in support of his attribution of this view to Anselm, he does not document precisely where Anselm allegedly adopts this view in that text, and there seems to be clear evidence to the contrary in that same text when Anselm says in section three that if we take the unity of the Persons to mean that the distinct Persons share literally every property then we will fall into the heresy of Sabellius and there will be one Person and not three.

Indeed, it seems to be precisely for these reasons that he introduces the Principle of Opposition in the De processione. To avoid seeming like an ad hoc solution to a rather serious problem, however, the Principle of Opposition needs to be supplemented with an account of sameness that preserves the unity of the Godhead while preventing the identification of the Persons with one another. Anselm accomplishes this with a principle that I take to be by far his most important and original contribution to this debate—what I shall call the Principle of Distribution. The principle is introduced in section 1 but is most clearly stated in section 15:

If we should consider the Father and the Son and the Holy Spirit in pairs, it is clear from what I have said that it is necessary either that one of the pair is from the other, since the other is not from that one, or that that one is not from the other, since the other is from that one. For example, if we should pair the Father and the Son, we perceive that the Son is from the Father, since the Father is not from him, and that the Father is not from the Son, since the Son is from the Father. And similarly, if we should consider the Father and the Holy Spirit, we find that the Holy Spirit is from the Father, since the Father is not from him, and that the Father is not from the Holy Spirit, since the Holy Spirit is from him. So also, if we should explore how the Son and the Holy Spirit are related to one another, we shall understand that the Holy Spirit is from the Son, since the Son is not from him, and that the Son is not from the holy Spirit, since the Holy Spirit is from the Son. Therefore, what I said before [in section 1] is evident, that the aforementioned relations, although they are in one thing, cannot let their plurality be absorbed in the unity nor can the unity let its uniqueness be absorbed in the relations.

This is a deucedly clever analysis, both from a theological and a polemical point of view. Polemically the argument ought to have appealed to the Greeks on account of their own clearly stated interest in the proper way to understand the expression “true God from true God”. Anselm suggests that the property of being either Deus or Deus de Deo is central to his own position regarding the status of the Filioque clause. Theologically the argument deploys a Platonic essentialism that also ought to have appealed to the Greeks if for no other reason than its manifest origin in Greek metaphysics but that apparently fell rather flat. The Godhead contains, as a matter of its essence, the asymmetrical relational property pair “from/not from”, and it is this fundamental relational property that defines the relations between the three Persons. Just as the Father begets the Son, the Son is begotten—that is, a non begetter, establishing the Father as the “from” and the Son as the “not from” elements of the relation when they are brought into comparison. Just as the Holy Spirit proceeds from the Father, the Father is a non proceeder, establishing the Father again as the “from” element and this time the Holy Spirit as the “not from” element in the relation. Since it is possible for the Son and the Holy Spirit to be considered as a pair, the full essence of the unified Godhead must still be present, otherwise it is not really a unified substance shared by all three Persons. If the “from/not from” asymmetrical relational property is an essential element of that substance, and if that substance is truly a unified singular, then it is necessary that the property will be present in any pairing of Persons, including the pairing of the Son and the Holy Spirit. Since the Son is evidently “not from” the Holy Spirit, we can infer that the Holy Spirit is “from” the Son, and not merely through the Son as by an instrument, per Filium, as the Greeks insisted, but from the Son as source, proceeding from him just as he proceeds from the Father, since the relational property must apply in the same way to every pairing. This point would have been very difficult for the Greeks to accept, and they could not have tried to apply Gregory of Nazianzus’ semantic equivalence rule here even if they had wanted to, since this is manifestly not the same sort of Double Procession that had been advocated by Saint Augustine and provisionally accepted by such Eastern Fathers as Saint Cyril of Alexandria. It was, in fact, in order to nuance the facts about Double Procession that the several Greek technical terms for modes of procession had been developed in the first place, and ex hypothesi those terms did not carry the same meanings as required by Gregory. Taken individually, then, every Person of the Trinity is in fact the same with respect to having the disjunctive property “God who is either from or not from God as origin or source”; taken as pairs, the disjunct collapses into the respective Personal relations, either “God who is from God as origin or source” or “God who is not from God as origin or source”. This is a sameness relation that preserves both unity and distinctness.

Cartwright objected to Geach’s proposal on the reasonable grounds that it does not prevent us from inferring that there are three Gods, since on Geach’s account the distinctness of the Father from the Son and the Holy Spirit, and the distinctness of the Son from the Holy Spirit, does not preserve the unity of the Godhead: no matter how similar the content of the various editions of the novel I own, each copy is a distinct entity, and they do not have every property in common if for no other reason than they necessarily have an ordering on my shelf. While there is one sense in which my seven copies of Mister Dog can be counted as only one novel, there is an equally plausible sense in which they can be counted as seven novels of precisely the same kind. This is particularly true if you are a nominalist, and the kind term, “anything that is an edition of the novel Mister Dog”, is merely a name for the various particular entities bearing the title "Mister Dog by Margaret Wise Brown" or its equivalents, such as Monsieur Chien, Herr Hund, and the rest. For a Platonist such as Saint Anselm, however, kind terms refer to genuine substances, substances that are singulars in Ernst Tugendhat’s sense of sui generis particulars. There is only one God in the same way that, for Plato, there is only one Form of the Good. According to classical Platonism, there may be material particulars that have a share in the Forms by means of a rather vaguely defined participation relation that permits the predication of the Form name of the particular, but the Form is, strictly speaking, a singular and not a universal in the nominalist sense. In a similar way God is not a kind of entity of which the three Persons are instantiations, but rather just is the three Persons in singularity. The entities that are the Persons are not to be distinguished from one another by some principle of individuation, such as physical matter in the case of sensible particulars or intelligible matter in the case of the objects of such sciences as mathematics and geometry. Rather they can be distinguished only by virtue of the logical and theological relational properties that we have been discussing, and these properties are not properties that belong to the Persons as particulars in the way that “this particular shade of paleness” belongs to Socrates, but are rather properties that belong to the Godhead itself as a unified, singular entity.
Since the Persons are not constituted in their very being by any principle of individuation in the way that material objects are, they cannot be distinguished from one another at all unless some property of their substance does the job, and in this case that is precisely what the Principle of Distribution does. The worry now is not so much that the properties of the Persons are incompatible with the properties of the Godhead, but that Anselm’s solutions to those properties are ad hoc. In particular, the concept of sameness tempered by the Principle of Distribution together constitute an uncanny metaphysics to modern ways of understanding. The somewhat desperate claim sometimes made in this context, that the doctrine of the Trinity is supposed to be mysterious, is an obvious non-starter for the philosopher, but it is a non-starter for the theologian as well. On this point it will be salutary to quote from Cartwright again, who makes some strikingly apposite remarks (p. 193):

It will be said that a philosopher is trespassing on the territory of the theologian: the doctrine of the Trinity is a mystery, beyond the capacities of human reason, and hence the tools of logic are irrelevant to it. The objection is based on a misunderstanding. The doctrine of the Trinity is indeed supposed to be a mystery. That simply means, however, that assurance of its truth cannot be provided by human reason but only by divine revelation. It is to be believed “not because of the natural light of reason, but because of the authority of God who reveals it” [Vatican I]. But a mystery is not supposed to be refutable by human reason, as if a truth of reason could somehow contradict a revealed truth; on the contrary, putative refutations are supposed themselves to be refutable. Nor is a mystery supposed to be unintelligible, in the sense that the words in which it is expressed simply cannot be understood. After all, we are asked to believe the propositions expressed by the words, not simply that the words express some true propositions or other, we know not which.

Cartwright is surely right in this, and even if he is not right with regard to every possible doctrine, surely that “faith seeking understanding” will want to adopt what Cartwright says here as her motto. And yet the emphasis on intelligibility and meaning is troublesome. Much of what Anselm has to say is difficult to swallow even for an inveterate Platonist, but for the anti-Platonist and even the mere non-Platonist, it still seems less intelligible than just abandoning the doctrine altogether. After all, this is simply not what we mean by the terms “sameness” and “identity”: we mean what Leibniz said we mean. But this, too, is troublesome, since it is the custom of the analytical philosopher to insist that meanings are either stipulated or discovered through use. If the meaning of the sameness relation is merely stipulative, then there is no reason to regard our own conception of it is more objectively reliable than Anselm’s, and if meaning is use then of course there is an entire community within which Anselm’s concept of sameness is entirely meaningful, and it perhaps says something about the rest of us if we can’t make sense out of their meaning and assume that the problem lies on their side.

Saint Anselm, then, attempts to avoid confounding the Persons, without dividing the substance, of the Trinity, and he does so by means of four fundamental principles:

(1) The Principle of Opposition
(2) The Principle of Non-ampliative Inference
(3) The principle of distribution
(4) The Principle of Platonic sameness

Taken together, these principles appear to support both the logical consistency of the doctrine of the Trinity as expressed through its relational properties and the addition of the Filioque clause to the credal statement that is the Niceno-Constantinopolitan Creed. Although the Greeks did not find Anselm’s argument persuasive, it is fair to say that his metaphysical principles were not much of an innovation beyond that which was already abundant in the tradition of Greek metaphysics more generally.