The Trinity and the Indo-European Tripartite Worldview

Andrew P. Porter, Graduate Theological Union, U.S.A. and Edward C. Hobbs Wellesley College, U.S.A.

As Published in Budhi, III, nos. 2-3 (1999), pp. 1-28

Posted to the Net with permission from the editors of Budhi. Page numbers in the version published in Budhi are indicated in [brackets].
[page 1]

In the middle of this century, evidence has accumulated to delineate a three-part conceptual structure(1) characteristic of Indo-European culture: a world-view. For Indo-Europeans, all human and divine society are divided in three parts, having to do with order, action, and sustenance. If pre-Christian Indo-European polytheism has a characteristic tripartite conceptual structure, it is reasonable to ask whether monotheism, when it moved from Judaism and a semitic language to Greek and Indo-European ideas, was influenced by the Indo-European tripartite worldview. The doctrine of the Trinity is the focus of such a question.

Some years ago, Edward C. Hobbs proposed a connection between the Trinity and Indo-European tripartition.(2) The three functions of the IE(3) system can be seen in the three Persons of the Christian Trinity, albeit
[page 2]
not in the same order. In addition, it is possible that there has been a partial exchange of functions between the first two Persons of the Trinity. It is also clear that in the doctrine of the Trinity, tripartition has moved from narrative and symbols to the realm of philosophy and abstract conceptual analysis.


What are the three IE functions?

The first is concerned with order, the structure and legitimacy of the cosmic and social world: the work of priests and lawyers, and more recently, scientists and the media of communication.

The second is concerned with action: in society, the work of the army, the warriors, the executive functions of government.

The third is concerned with sustenance: food, nutrition, fertility, health, agriculture, manufacturing, commerce, trade.

The IE pantheons, in their oldest form, all show three characteristic divisions along just such lines. In the commonest form, the first function is sovereign, and the third function is of a lower status in society. But this need not be so; whatever may be the case in the ancient examples, there are exceptions in the broader IE history. There is often a conflict between the first and second functions, a struggle for power, as between the Papacy and the Empire in the Middle Ages.

When the IE migrations moved into older cultures that had enough vitality to bargain conceptually with the IE system and extract ideological concessions from the invaders, as the pre-IE Greeks did, the IE system appears in distorted form, with broken symmetry. Georges Dumézil, who discovered the IE system, much preferred the Greeks to the pure IE system; the pure system he likened to a prison, but called Greece la "belle infidèle à l'idéologie tri-fonctionnelle," because she was always willing to look at the world with new eyes, with criticism and observation, rather than forcing it into a pre-established pattern.(4) Our purpose is not to promote (or oppose) the IE conceptual system, but rather merely to recognize it where it does appear, both so that it may
[page 3]
be understood for what it can do, and so that alternatives to it may be recognized where it cannot by itself explain matters. Other conceptual systems (and non-systems) are available, and some have begun to attract attention. The Uto-Aztecan language family and cultures in the North American Southwest exhibit a characteristic quadripartite ideology, based on the four directions of the compass.(5) The Bantu language family in Africa spans many cultures and climates, and begs for investigation. It is not clear what system, if any, is characteristic of Semitic cultures, but they tend to produces seven-fold categorizations, and they clearly do not think like the Indo-Europeans.(6) Alternatives are available.

Our claim is that the IE tripartite system does in fact color much of Christian monotheism. Our project is then to explore the relations between the IE system and monotheism as it appears in its Christian form. To do this, we shall have to acquire an understanding of monotheism apart from the IE system. Exploration of the relation between IE tripartition and the doctrine of the Trinity can (here) be carried out only in a preliminary way, hence we present conjectures about the Trinity. A full exploration would require a re-investigation of the entire history of the Doctrine of the Trinity, addressing the question of tripartition to the original sources themselves. This is not such a study. It is our intention here to present merely the thesis, with only enough exposition to make clear what is put in conjecture, but not enough investigation to test or defend it. That must await another occasion. And this exposition is logical and metaphysical, and historical only in a way secondary to logical considerations. We do not undertake comparative study of symbols, narratives, or anthropology. The present conjecture is about the structure of the doctrine of the Trinity. Without a preliminary conclusion to the presence of IE tripartition in that philosophical structure, Trinitarian theology can contain no more than vestiges and remnants of a prior IE tripartite non-monotheist worldview. Such a result, while of great interest to historians and anthropologists,
[page 4]
would not be of much interest to theologians and others interested in understanding the Christian project as it has explained itself in philosophical terms.

Dumézil's work spans decades, and most of it is devoted to detailed exegesis of issues in particular ancient texts, rather than exposition of the system as a whole. C. Scott Littleton's The New Comparative Mythology has provided both a review of Dumézil's work and the first English access to much of it. Dumézil's own systematic treatment appeared in L'Idéologie Tripartie des Indo-Européens.(7) The tripartition hypothesis emerged only from the perplexities and impasses of earlier theories; Dumézil started from the late nineteenth-century work of Müller and Frazer in anthropology and comparative mythology. The search for the origins of myth turned eventually to language and social structure, under the influence of Durkheim. Dumézil eventually turned to comparative structural analysis of myths and the parallels between social structure and mythological systems, first, and in a rudimentary form, for Iranian culture, with connections to India, and later, with confirmation all over the IE world.

Our conjecture is that the tripartite ideology has influenced the doctrine of the Trinity, with modifications. In order to see what is involved in this claim, it will be necessary to proceed in six steps:

  1. To examine the tripartition hypothesis as it applies to polytheism, with minor amendments;
  2. To reach a definition of monotheism apart from tripartition;
  3. To examine monotheism with tripartition: the Trinity in first approximation;
  4. To examine the Trinitarian adjustments of the IE tripartite system;
  5. To examine the theory of appropriations as a general guide to tripartite thinking in a Christian context; and
  6. To note some non-IE ideologies that by their contrasts to the IE system give it some conceptual precision.

[page 5]

Tripartition Without Monotheism

As we have said by way of introduction, the IE system is divided into three functions, of which the first has to do with order, the second with action, and the third with sustenance.(8) Usually in practice, and invariably in Dumézil's and Littleton's accounts, the first function is elevated to a sovereign position, and the third is depressed to a subservient position, so that the three departments of society are not only concerned with separate functions, they are in a hierarchical relation to one another. Thus Littleton's denomination of the functions as sovereignty, force and nourishment.(9) We would modify his terms in order to separate the issue of sovereignty from the question of the proper business of the first function. Without such a distinction, it would not be possible to understand historical conflicts over sovereignty between the first and second functions. While history exhibits the first function as sovereign in the early, pure examples, it takes little reflection to see that this need not be the case. Comparative anthropology is necessarily concerned with the actual particulars of actual history, but systematic (and even historical) theology have to focus on the nature of the functions in the broader range of their possible relations. To put it another way, the sovereignty of the first function is a contingent historical fact, not something inherent in the logic of the functions. The Trinitarian development moves in the direction of making the three Persons more symmetric, and of avoiding any subordination of one Person to another; this is the thrust of the "Athanasian" creed. We shall speak not of sovereignty in the first function but of legitimacy and order, in the sense which Dumézil elaborates in Mitra-Varuna, where the first function is presided over by two coordinate deities, one in charge of cosmic order, and the other in charge of enforcement of contracts. In the Roman case, they are Jupiter and Dius Fidius.(10) If the principle of the first function is identified as legitimacy and order, it will be possible to understand power struggles between the first and second functions, and to understand
[page 6]
the principles of the functions in a way that does not a priori subordinate one to another.

"Function" refers not so much to social strata or to a class of divinities (though these are the clearest surface manifestations of functionality) as to the principles operating in this scheme of analysis. It is true, nonetheless, that the three functions together analyze what is an underlying unity, whether it is the life of an individual, of society, of the pantheon, or life in some other context. While no single three words capture the principles of the three functions, "cognition," "action," and "emotion" serve reasonably well at the level of individual life, and "order," "action," and "sustenance" will do at a more general level. The ideology extends from triads of classes, castes, orders, and their collective representations, to triads of calamities, colors, talismans, cures, even celestial and geographical regions. The ideology forms the core of its culture; "it provides the basic framework in terms of which phenomena are categorized and thus rendered meaningful. It is in terms of its ideology that a society structures its religious beliefs, validates its social organization, and generally conceives its relation to the phenomena around it."(11)

The tripartite social organization is manifested in a system of three social strata, of priests, warriors, and herders/cultivators. The king comes from the second stratum and remains secular; a thoroughly secular monarchy is a uniquely Indo-European institution. There are few if any parallels in pre-Columbian America, the ancient Near East, the Nile civilizations, China or India, before Indo-European migrations into those areas. It is secular, in that it does not combine sacerdotal and administrative functions. The Indo-European gods of the first function regulate magico-religious and juridical or legal legitimacy and order. These gods are sovereign in the system. Generally the numinous and legal aspects of order are supervised by different gods, as in the Roman case. The second function gods endow physical prowess, and those of the third function, sustenance, well-being, and fertility.

Ancient pantheons comprise far more than just three deities, but they were grouped in three departments. The first was subdivided into two parts, usually with two coordinate gods, one in charge of cosmic
[page 7]
order, the other in charge of juridical order;(12) in the Roman pantheon, Jupiter and Dius Fidius; in the Indian, Mitra and Varuna. The first function always enjoyed priority in the attention of researchers, and the second and third remained comparatively underdeveloped.(13)

The Indian and Iranian cases were worked out first, and provided the basis for generalization to other Indo-European mythological systems.(14) In India, there are four social classes (varnas), Brahmans, Ksatriyas, Vaisyas, and Sudras. Brahmans are the priest varna or class, Ksatriyas the warrior class, and Vaisyas the economic class. The last, the Sudras, do not belong to any of the three functions, but serve them all;(15) they were probably a conquered indigenous people, a later "addition . . . to the basement of the previously tripartite varna structure" or caste system.(16) The first three classes are Arya, a word root originally meaning simply "people." The tripartite ideology appears in the Rig-Veda, the oldest literature of the civilization. Mitra presides over rational and legal aspects of sovereignty, and is in effect contract personified. Varuna is magico-religious, awesome and terrible, presiding over the numinous. Indra presides over the second function gods, e. g., the Maruts. He fights monsters and so represents power and prowess. The third function in India as elsewhere is represented by several gods, here the Asvins and Sarasvati. Fertility, harvests, comfort, health, well-being, are all bestowed by the third-function deities. Brian K. Smith has recently pointed out a close parallel of the tripartite caste system to the tripartite character of the Vedas, with the later addition of the Atharva, the fourth Veda, becoming another "example of adding an inferior fourth to a prior triad."(17)

[page 8]

The three functions are reflected in the early Roman pantheon, where each of the functions is headed by one or two gods, and there is a sense of distinctness of the functions. In later Roman mythology, the symmetry of the three functions is distorted, broken, and obscured. At the beginning, Jupiter and Dius Fidius head the first function, but later, Dius Fidius recedes and is obscured. Together, they are the gods of mystery, or the numinous, and contract, or the honoring of promise, the two components of the first function. Mars is the god of war and physical prowess, in parallel to Indra in the Indian system. The third function is headed by Quirinus and Ops. From Ops we get the word "opulent," and the stem of Quirinus supplies also the verb curo, curare, care for, pay attention to, trouble about. Social classes in Rome follow the Indian pattern. To the Brahmans, Ksatriyas, and Vaisyas correspond the Roman Flamines, Milites, and Quirites.(18) Both Rome and India have historicized or legendary versions of the divine or mythical representatives of the three functions.(19) Both kings and gods can be found to correspond to members of the Indian pantheon. Of the early kings, Romulus and Numa correspond to Varuna and Mitra, Tullus Hostilius to Indra. The third function is not clear, though Dumézil feels that the Sabine War represents Indo-European myth in which the classes of the first and second function defeat and integrate the third function.(20) The pre-Capitoline triad of Jupiter, Mars, and Quirinus preside over the gods; Dius Fidius and Ops are added in the first and third functions.(21) The later Roman pantheon, the Capitoline triad, is composed of Jupiter, Juno, and Minerva. This represents a considerable departure from the IE scheme, and it is induced by contact with Etruscan culture.(22) This is an example of the ability of pre-existing non-IE cultures to modify the ideology of the IE invaders.

The IE system appears more clearly around the edges of the Indo-European world than in the center, where advanced civilizations existed prior to the Indo-European immigration, with myths and conceptualities sturdy enough to survive in the resulting mixtures. Here
[page 9]
problems, distortions, and broken symmetries surface, with confusion of functions. Greek culture offers an example of just such blurring of Indo-European thinking by an older culture. It was not originally Indo-European, and its later Indo-European development reflects only broken symmetry and modified features of the tripartite system. Zeus enjoys both first and second function attributes, cf. Mithra in the pre-Zoroastrian Iranian mythology. The breaking of symmetry and blurring of functions within a system that is nevertheless tripartite appears in the story of the judgment of Paris. Paris is asked to choose between regal Hera, warlike Athene, and voluptuous Aphrodite. To his sorrow, he chooses the third function when he should have chosen the first. But even in this story, Hera offers sovereignty and wealth, Athene victory, handsomeness, and wisdom, combining two or even three of the functions.(23) What is surprising, in view of the distortions in the center of the IE world, is the high degree of survival of the tripartite ideology throughout all of the IE migrations. The system occurs in more or less this form throughout the IE world, from Ireland to India, but it does not appear outside the IE world. It seems to be a conceptuality that is available to all the IE cultures. Comparatively late in IE history, it is available to the first IE speaking monotheists, Christians of the first centuries of the Common Era. It would be only natural for them to interpret their new faith in terms of the (tripartite, and only) world they knew. Indeed, it would have been odd if they had not done so. But before we turn to examine the way in which the first IE converts expressed their new monotheism in old IE categories, it is necessary to reach some sort of identification of monotheism apart from tripartition, in order to see the faith-commitments that are expressed in new form in tri-functional concepts.

Monotheism Apart from Trinity and Tripartition

There are other examples in the ancient world of people who worshiped only one god among many, and at least one non-Biblical example of a ruler who attempted to enforce worship of only one god (Pharaoh Akhenaton). What is different about the God of the Bible, and original with Judaism, is not the oneness of God, but his relationship with the believers, one that can be characterized in the term covenant. A later term for this relationship is providence, though that is an IE word, and
[page 10]
not a term that TNK(24) would use. The God of the Bible provides good in all of life, not just parts of it, however unfinished the working out of that good is. This peculiarity of Biblical religion was not originally a matter of arithmetic, the numerical oneness of God. Hebrew monotheism was amalgamated from originally diverse cults of various gods; the God of the Patriarchs was not originally One.(25) The goodness and benevolence of the emergent God was the basis for the sense that in different tribal manifestations, there might be an underlying divine unity.

The fact that Biblical religion affirms the world and life in it is not sufficiently concrete for our purposes; it affirms life as history, not just as nature. This distinction appears not so clearly in H. Richard Niebuhr as in Mircea Eliade's Cosmos and History, and it is elaborated in Merold Westphal's God, Guilt, and Death.(26) Eliade turns from the apparently safe world of nature, safe because cyclical and therefore predictable, to the terrors of history: unaccountable suffering, the contingent, above all, whatever cannot be subsumed under the regularities of nature. History is the realm of freedom and responsibility, and the realm of risk, contingency, and lack of control. To trust that history is also the locus of blessing is supremely difficult unless one comes to history with faith in a covenant as the focus of one's religious life. Our interest is not so much on the dynamics of history as it unfolds as it is on how covenantal religion handles the disappointments of life: it must embrace them as bearing blessing of some sort, eventually, if not obviously now. By contrast, in Westphal's typology, nature-focused religion affirms life as nature, but not as history, and seeks to integrate human life into nature, by ritual mimesis of nature. He calls it mimetic religion for this reason. Religion which views human life and the material world as an exile from an other-worldly state of grace, he calls exilic religion;
[page 11]
gnosticism is its most prominent example, and it is characteristically world-denying.

If God's goodness is reflected in all creation, as history, then all of life, even in its disappointments, bears some blessing. No parts are barren. It was a long struggle to see this: the prophets strove to remind their hearers that disaster was coming, and that it would bring a blessing from the Lord. It is this transformation or inversion, in which judgment and sentence are eventually transformed, ending in a restoration, which James A. Sanders takes as the crucial mark of the radical monotheism of the prophets.(27) The prophets' message was one of judgment and sentence: impending disaster, followed by restoration. The first was easy to hear (fright usually sells well), though the second made no sense.

But it was the message of hope after judgment which was so incredible and still is quite incredible if we judge from all the modern commentaries which miss the point. The people could hear the message of doom, and they did not like it. But what they just could not grasp was that in the doom was Israel's true hope.(28)
Not only the Exile, but also the earlier experience of the Exodus was one of unexpected blessing found in disappointment: how God brought Israel out of slavery, through the Testing in the wilderness, and into the promised land. During the Testing (and before, in slavery), it did not look at all as if they would see any blessing ever again. And the thesis that all of life and the created world are good, in implied contrast to the selective rejection of parts as bad, is one of the major themes of the first chapter of Genesis. This inversion, whereby disappointments are transformed into blessings, is the one utterly crucial characteristic of monotheism.

H. Richard Niebuhr's own formulation of the thesis that the goodness of God is the defining mark of monotheism, before the oneness of God, appears in Radical Monotheism and Western Culture and in Faith on Earth.(29)
[page 12]
His inquiry starts from human faith, and proceeds from the structure of that faith to its "object" the god(s) believed in. His own analysis shows tripartite categories in places, but he focuses the central definitions of Radical Monotheism in a function-neutral way in a 1943 essay, "The Nature and Existence of God: A Protestant View," reprinted under the title "Faith in gods and in God" in that book. To have a god is to live for a cause, and the cause is defined by the goods it provides, implying a separation of life into two parts, the good provided by the god, and the rest of life rejected as evil. This is a fair characterization of religion that takes nature as the ultimate locus of meaning, in the way it rejects history and all parts of life that are irreducibly contingent (the historical, in effect) as ultimately barren.(30) Covenantal religion supposedly takes all of life as good, or to be integrated into a good yet to be consummated. Niebuhr is quite repetitious in emphasizing this goodness in Radical Monotheism. Yet Merold Westphal's exposition of covenantal religion (his chapter 11, esp. section 11a) is visibly embarrassed at all the ways that Christianity as it actually is can be confused with exilic religion. We would go farther, were this the place for it, and claim that Christianity in actuality has often become intimately involved with exilic religion, not just mistaken for it. It is, of course, characteristic of gnosticism that it rejects the created world as defective, and seeks escape from it, and especially from the human body. In a strict definition of monotheism as covenantal but not exilic religion, we focus on its unqualified affirmation of the created world in order to bring some conceptual clarity to our argument. The fact that real Christian monotheism is often a confusing mixture of religious types complicates matters more than it is necessary to treat here.(31) Official Christianity has been affirmative of the created world, as Westphal has carefully shown, and this is sufficient to warrant our claims about the logic
[page 13]
of the monotheistic transformations of life, as their articulation changes with the move of monotheism from a Hebrew-speaking community into one informed by the IE tripartite ideology.

In Niebuhr's estimate, trying to separate life into the good and the evil eventually ends in failure. The God of monotheism is behind this frustration and disappointment of all human causes, and to trust in that God is to trust that the disappointments of life are transformed, one way or another, into blessings.(32) It is implied in Niebuhr's emphasis on the good of all being, as being, that monotheistic faith finds blessings in disappointments, simply as part of life. Exilic and mimetic religion, we claim, do not. And for Niebuhr, henotheism, the worship of one god appropriated to the service of the interests of a separate community, still rejects part of the world as barren, and part of humanity as unsaved and unsaveable. The fact that a henotheism worships only one god does not in the least make it monotheistic, that is, universally covenantal in its attitude toward the disappointments of life. To make concrete the transformations of disappointments into blessings that is characteristic of Biblical religion in its IE setting, it is necessary to pass to the next part of our conjecture, the instantiation of monotheism in its tripartite form. There we shall see the differences between IE tripartition expressing the transformations implied in covenantal religion and IE tripartition in its original nature-focused polytheistic home.

Monotheism with Tripartition: The Trinity in First Approximation

A people that organizes the world into order, action, and sustenance will conceive disappointment in each of the three functions in a characteristic way. When one encounters disappointment in a situation, that disappointment can come in three ways. The experience of exposure is to be shown as the person one is, inconsistent with the image he has put out of himself. When the individual encounters disappointment of action, whether from causes in nature or society, he faces limitation.
[page 14]
When he encounters another who makes demands on his sustenance, he faces the other's need. This tripartite characterization was first seen by Edward Hobbs. He describes the responses to these three sorts of encounter with disappointment:

(1) in the face of the situation which exposed or revealed the discrepancy between one's pretensions and one's actual life-as-lived, one responded with acknowledgment of the true situation and a "change of understanding" (Greek: metanoia, poorly translated "repentance"); (2) in the face of the situation which confronted one with the contingency or limitation of his existence, one responded with creative thankfulness for the new -- albeit in many cases unwanted and limited -- possibilities presented by the limiting situation itself; (3) in the face of the encounter with others in their need for help, one responded with action directed to the benefit or good of those others.(33)

Exposure is judgment, and it brings redemption from sin. The one who brings judgment was known in Judaism as the "anointed" one, Meshiach in Hebrew, Christos in Greek. In the limiting situation, one discovers one's creatureliness, and a faithful creature responds to it as possibility-offering. It is like meeting the father whose gifts "are not always what was desired or even understood," but presented opportunities nonetheless. Others in need offer relationship, community, the fellowship of the Church and the sustenance that comes with it: the work of the Holy Spirit.

What is immediately surprising is that the first function is assigned to the second Person, the Son, and the second function is assigned to the first Person, the Father. The order has been changed. The situation is slightly more complicated than this, but this is a good approximation, and an approximation quite sufficient to deal with most problems. God the Son brings redemption from sin, a first-function problem, and God the Father brings blessing in contingency, a second-function problem. Moral order is the business of the first function for monotheists and non-monotheists alike, and the monotheist trusts that disappointment in the first function is redeemable and not barren, as part of the larger monotheist commitment to the goodness of all of life. In the second
[page 15]
function, action and contingency, arises the problem of the goodness of creation: is the world, as a given, as limitation, good? This is the department of God the Father. The change of order is the clue which will unravel the extent to which Patristic reflection on the mystery of God was shaped by the IE conceptual system, and the extent to which the early Church's inheritance from Second Temple Jewish monotheism extracted modifications from the IE conceptual system. To that issue we now turn; it is the central part of this conjecture about the Trinity.

Trinitarian Adjustments of the IE Tripartite System

As Christian monotheism moves into the IE world, the development to be expected is a little different from that found in comparative IE studies of polytheistic religion. In the polytheistic cases, the evidence seems to support a genetic hypothesis: a common ancestor in proto-IE culture underlies all the later mythological developments.(34) The case of monotheism is clearly different: Jewish religion and IE conceptual structures meet, and some adjustments are to be expected between them. The interesting task is to trace out those conceptual adjustments. The IE system will not appear in monotheistic religion as repetition of aboriginal myths so much as a tendency to think in triples related to one another as order, action, and sustenance. That is, tripartition in Christianity will appear as a conceptual genre, not a rigid thesis. And this is what the history bears out. The general councils which regulated the doctrine of the Trinity in effect were promulgating genre-rules, or a "grammar" of Trinitarian discourse, more than permanently fixing the content of the doctrine. Those "grammar" rules have served not to constrict the doctrine, but, in defining a genre of theological thought, have made possible a doctrinal development that has been as rich and various as any in Christian theology.

There is, we suggest, a characteristic way in which the IE worldview approaches a problem. When a phenomenon has been discovered as interesting, at some point, enough of it has been seen to suspect that
[page 16]
it has a structure. The IE approach tends to look for tripartition as the model for all structures, and so in consequence, when a second feature of the phenomenon of interest has been seen, the IE worldview will make function assignments, whereby two of the three functions have been identified. The search for the remaining function, to complete the triad, is usually fairly short once two functions have been spotted.

The history easily bears out this general rule. The theological problem which appears immediately, and comes to a head in the fourth century, is the problem of defining the relation of the Father and the Son: the Arian controversy; whether the Son is co-equal with the Father, or is subordinated to the Father. When that question is resolved in favor of co-equality, it is only natural to integrate the Holy Spirit into a completed triad.

That describes the later stages of Trinitarian thinking. But the roots of this development of thought are present already in the New Testament. The common greeting is "Grace and peace to you from God our Father and the Lord Jesus Christ." A greeting of this form appears in the beginning of 1 Thessalonians (the earliest written document of the New Testament), and, interestingly, the Holy Spirit is mentioned a few verses later. Complete Trinitarian doxologies do appear in the NT, in both the IE order and the permuted Trinitarian order: "The grace of our Lord Jesus Christ, the love of God, and the fellowship of the Holy Spirit be with you all" (2 Cor. 13.14); and "Go therefore and make disciples of all nations, baptizing them in the name of the Father and of the Son and of the Holy Spirit" (Matt. 28.19).

The effort of patristic theology in the next four centuries on these issues is directed to formulating the genre-restrictions necessary to achieve an account of God that is both faithful to the experience of the believers (this is explicit) and that is expressed in the only conceptual system they had (this is implicit). Subordination is the focus of most of the controversy. Interestingly, it is not the usual IE pattern of subordination, of the second function to the first; rather the subordinationists would make the Son, a first-function figure, subordinate to the Father, a second-function figure. In the Arian form of this move, the incidental effect was to make the Son less than really divine, and therefore to compromise the redemption of disappointment in the first function. The result would have been to make redemption from sin distinctly secondary to the creation of the world. In effect, Arian Christianity would perhaps have been monotheist in the second function, but not truly so
[page 17]
in the first (if such an inconsistency is even sustainable, which we doubt). In contrast to the Arians, Gnosticism does follow the characteristic IE pattern of subordination: the second function is subordinated to the first. God the Father, the God of the Old Testament, is made secondary to a higher god, whose representative Jesus is. And God the Father is regarded as malevolent. Clearly this (mis)understanding of the Old Testament has survived throughout Christian history, and has returned with special virulence in the nineteenth and twentieth centuries. This form of subordinationism also compromised the goodness of God, and so compromises the monotheistic transformation of disappointments into blessings.

The outcome of the Trinitarian controversies, as may be seen for example in the so-called Athanasian creed, was to insist that the Persons are co-equal, none subordinate to another, and to insist on a proper balance between the threeness and the oneness in God. Within these limits there has been enormous room for creativity in the post-patristic period, as even a casual consultation of histories of the doctrine will show.(35) Departures from the Trinitarian genre have usually resulted in compromising not only the doctrine of the Trinity but also the monotheist commitment to the goodness of all of life, not excluding any part of it. H. Richard Niebuhr has dealt with the problem, though in an argument that is historical rather than addressed to comparative IE issues.(36) In his analysis, to doubt the divinity of the Son is to doubt the power of goodness; to doubt the divinity of the Father is to doubt the goodness of power. The issue in the third Person is whether among spirits in human communities, any is ultimately trustworthy.

Parenthetically, it is interesting at this point to speculate on the possibility of a monotheism which is tripartite but does not embody a doctrine of the Incarnation as we have it in Christianity. Such would confirm our conjecture that tripartition in monotheism (and therefore
[page 18]
Trinity) can be distinguished from the Incarnation. The history of controversy over the Incarnation by which the Trinitarian settlement was reached would then appear to be a contingent path to a tripartite conclusion. The place to look for non-Christian but monotheist and tripartite thinking would be in a figure such as Philo, a Jew writing in Alexandria, in an IE language, in conversation with Greek philosophy. If our conjecture is correct that IE culture supplies the conceptual structures in which its members think, Philo ought to exhibit tripartite structures at least occasionally, and in ways that Rabbinic Judaism does not naturally do. Indeed, Philo foreshadows many of the moves the patristic writers will later make in transposing Jewish monotheism to a hellenistic and philosophical (and incidentally IE) world, with the obvious exceptions about Jesus. (For example, philosophy is to be harmonized with Scripture where possible, corrected if necessary.) The most casual search of Wolfson's article "Philo" in the Encyclopedia of Philosophy(37) turns up one clear example of just exactly what one would expect: a Hebrew term that is not assigned to one of the IE functions in its original usage requires translation by at least three different terms in Greek, and the Greek terms are function-specific, because IE speakers characteristically looked for function-specific terms. The term is tsedaqah, usually translated into English by "righteousness," and in the Old Greek ("Septuagint") by both dikaiosune (justice) and eleemosune (mercy, alms). (Philo here includes philanthropia (humanity), giving help to those in need of it.) At this point, we have the second and third functions, lacking only the first. For Philo it is necessary to find three virtues in Greek to represent the original Hebrew concept; two of them appear already in the Septuagint. Because concepts tend to be forced into one or another of the functions, justice is appropriated to right action, a second-function concept. Philo completes the triad of virtues with metanoia, "repentance," the missing first-function translation of the concept of tsedaqah. And repentance is just exactly the response to exposure which we have posited above. Interestingly also, metanoia was considered a vice in Greek philosophy, but a virtue for Philo Judaeus. Christianity follows Philo's Jewish monotheistic instinct in regard to metanoia. Thus we have here both a full tripartite analysis of the virtue
[page 19]
of tsedaqah and also the peculiarly monotheistic insistence on embracing the disappointments that come in each function, because they are not barren, but bear blessings. Joseph Soloveitchik, by contrast in Halakhic Man,(38) focuses on repentance and limitation in ways that do not at all support a clear distinction between the first and second functions.

Appropriations of the Trinity

In his treatise on the Trinity in the Summa Theologica, Aquinas raises a question representative of a general issue in tripartite thinking: whether attributes essential to God as a unity should be appropriated to individual Persons (I, Q 39, Art. 7). The risk (Objection 1) is that essential attributes could be seen to belong only to the Person to whom they are appropriated, rather than to all three equally. The objections hint that the Persons could thereby be not only distinguished, but separated, opening the way to tri-theism. Aquinas's response is to show how and why appropriation works: it makes the Persons manifest, and does so in terms available from the created world of the believer. In effect, this is the way the IE mind works: a phenomenon which is a unity manifests itself in the three functions.

The three Persons of the Trinity have evidently acquired enough conceptual life of their own, that they can all three be said each to manifest all three functions. They originated diversely, the Father as the transcendent God of Jewish experience; the Son as the God of Judaism acting in the historical person of Jesus, with a trans-historical significance; and the Holy Spirit (with roots in Second Temple Judaism) seems to stand forth in a natural triad. When they are assigned tri-functional attributes, the question of appropriation eventually arises as Aquinas formulates it. Thus tripartition is applied recursively: the aspect of a phenomenon that has shown itself in each function can itself be analyzed tri-functionally. In effect, the partition structure that results is a matrix of conceptual possibilities:
[page 20]

Son Father Spirit
first function the Son in 1st f. the Father in 1st f. the Spirit in 1st f.
second function the Son in 2nd f. the Father in 2nd f. the Spirit in 2nd f.
third function the Son in 3rd f. the Father in 3rd f. the Spirit in 3rd f.

But in practice, in the common appropriations, the matrix is usually diagonalized:

Son Father Spirit
first function redemption
second function creation
third function sanctification.

In place of the appropriation triad of redemption, creation, sanctification, one could have taken God in history, God in transcendence, and
[page 21]
God immanent in human life. Or one could have used that triad as itself one axis of an appropriation matrix:

History Transcendence Immanance
first function Son
second function Father
third function Holy Spirit.

Claude Welch and David Brown take such an approach.(39)

The appropriation matrix need not be diagonalized. The function of order may be assigned in different respects to Father and Son: moral law, statute, is defined by the Father's will. As will, is a form of contingency or limitation. Moral action, as exposure, is performed by the Son, in the Passion and redemption. Correspondingly, cosmic action, creation, in its contingency, is an act of the Father. The intelligibility of the cosmos is assigned to the Son, the Logos, a first-function aspect of God, and early taken to be present at the beginning of creation. In effect, deity has been analyzed into three functions twice, in different ways. An appropriation matrix to describe this could be devised in several ways, each growing from a different selection of "axes" on which to plot the appropriations:

Son Father Holy Spirit
history, first function juridical action (redemption) juridical order (the law)
transcendence, second function cosmic order (logos) cosmic action (creation)
immanence, third function spiritual sustenance (sanctification)

[page 22]
or, if different reference-concepts are chosen, as

History Transcendence Immanance
first function Son, juridical action Son, cosmic order
second function Father, juridical order Father, cosmic action
third function God the Holy Spirit.

It is a fact of tripartite analysis of a unitary phenomenon that features manifested in it can be appropriated on different occasions to different functions. This is perhaps an essential speciation on the way from the generic (and traditional) IE polytheist tripartition to tripartition as we have it in the Trinity. Thus cosmic and juridical order can each be taken in the first or second function. Juridical order, appropriated to the Father, as contingent, is assigned to the second function. As order (and as arising in history, which provides moral order in the present), it is assigned to the first function. Cosmic order, appropriated to the Son, as order, is assigned to the first function, but as contingent, it is assigned to transcendence and the second function. The relation of the intellect and the will (the first and second functions in a tripartite anthropology) will be an intimate one: they can be distinguished, but not separated. And order for Trinitarian Christianity will always be a contingent order.(40) We suspect that this transformation has its origin in the Jewish roots of Christian monotheism, where the first two functions are not distinguished in a systematic way at all. The attempt to express
[page 23]
that Jewish (and un-IE) commitment in tripartite and IE terms inevitably calls for a theory of appropriations. It is by means of concepts such as appropriation that a creative theologian has endless possibilities in unfolding the doctrine of the Trinity. Indeed, it is by means of appropriations, which may be elaborated as necessary to describe a phenomenon of interest, that Christian theologians, like the Greeks before them, could look at the world with new eyes, with criticism and observation, rather than forcing it into an inherited pattern -- yet still work with a pre-established set of concepts.

It is possible for an appropriation to become so scrambled that it is no longer recognizably either monotheistic or tripartite, in spite of purporting to be Trinitarian. The Trinity is sometimes explained in terms of omnipotence (the Father), goodness (the Son), and sanctification (the Spirit). Only the third function has survived intact. The trouble is that the monotheistic transformation in the second function, the work of God the Father, has been lost, and omnipotence, incorrectly ascribed to the first function, no longer is marked with the function of transforming limitation into blessing. To compensate, goodness is assigned to the Son, but as this sort of once-monotheistic thinking evolves in practice, it would appear to be goodness which exempts from limitation, rather than redeems it. As such, the Son has been put to work in the second function. Redemption from sin, the transformation of exposure, is utterly forgotten, and with it the first function.

To put it another way, this Trinity parallels a common misconception of Christmas. In that conception, Santa brings nice gifts only to the good, and abandons or punishes bad children. Here, the Father is a first function figure, but he is judge for other people, and Jesus is savior for us. There is no judge for us and our friends, no exposure that would save at cost of embarrassment, and those other people are just out of luck. God the Father is the first function judge who also imposes limitation on the bad they-people as punishment, and Jesus is the second function savior who exempts the good we-people from limitation. The monotheistic transformation of disappointments has been completely subverted. In Christmas as it really is, of course, Santa brings gifts to all children, regardless of whether they have been good or bad. The children of the rich get more than the poor, but that has nothing to do with being good, and rich children don't seem to enjoy their gifts any more than the poor do.
[page 24]

Littleton claims that Christ sometimes appears as a warrior,(41) assignable to the second function in a way that is quite unambiguous compared to the appropriation matrices constructed above. This development would be legitimate within the limits of Trinitarian discourse, provided it were complemented with a balancing assignment to the first function. Nevertheless, it is quite rare, and does not appear in early sources. The roots of the Trinity make the Son either unambiguously a first function figure, or else place him in a dialectical first and second function relation to the Father, in which the two functions can be distinguished but not separated.

Contrasts in Non-IE Conceptual Systems

The original IE system has thus, in Christian hands, become not a fixed narrative of the gods, but a genre of explanation, with considerable flexibility in application. One can tell a limerick from a sonnet. And one can easily tell a poem which is almost a limerick from one which is almost a sonnet. In the last section, we would like to return briefly to an Amerindian system mentioned at the outset, and then to examine a familiar passage in TNK which is monotheist but not tripartite in the IE sense.

The Uto-Aztecan family of Amerindian languages has a common religious mythology and it appears to partition the universe into four parts, for the four directions of the compass, or sometimes four directions and the center.(42) It is obvious that Christian theology does not think in such a four or five-membered partition-system. It is not yet obvious from the anthropological literature what are the animating principles of the four or five functions in the Uto-Aztecan system. We are able to find analyses at the level of comparative symbol sets, and comparative narrative, but not of abstract thinking. (If indeed abstract thinking is even a fair question to bring to this problem.) The question could be phrased in this way: given the five functions of the points of the compass and the center, how would the God of monotheism, the God who brings good even in disappointments of life, manifest itself in these five functions? What would it be to experience the transformation
[page 25]
of disappointment into blessing, at each of the cardinal points of the compass, and at the center? To answer that question, some sort of principles of the Uto-Aztecan functions would have to be discerned. In such principles, we might not see any interest in the contrast of transcendent and immanent, nor the focus on temporality that is so prominent in IE thinking. Such examples, by their contrasts, would tell us what the Trinity is not.

Much more familiar is a passage in the Old Testament, the Shema:

Hear, O Israel: The LORD our God is one LORD; and you shall love the LORD your God with all your heart [levav], with all your soul [nephesh], and with all your might [meod]. (Deuteronomy 6.4-5, RSV)
Hear, O Israel: The LORD is our God, the LORD alone; and you shall love the LORD your God with all your heart [levav], with all your soul [nephesh], and with all your might [meod]. (Deuteronomy 6.4-5, NewRSV, also Tanakh)

Both the RSV and the NRSV, as well as Tanakh, the New JPS translation, translate levav, nephesh, and meod as heart, soul, and might, not yet exactly an IE triad, but a translation which follows the Septuagint, which has kardia, psuche, and dunamis. The Gospels quote the passage, but they have changed it. Kardia remains, but psuche has been supplemented by dianoia (mind). In Mark and Luke, dunamis has become ischus, which also means "strength." In Matthew it has been deleted. The result is to force the Shema toward IE categories, whereby the first and second functions are displayed unambiguously. The original sense of the series levav - nephesh - meod might have been preserved in translation, but as we have it, it was not. Meod has been treated the worst in the NT translation; lost in Matthew, it appears as "strength" in Mark and Luke, but strength is not really its central meaning. Examination of the usage of these words in the Common Documents translated as Old Testament and of the comments in a lexicon amply confirms the suspicion that in translation they are often forced to become IE function-specific.

Translations into one or another IE language may differ, but the closest example (English) will exhibit the problems of translation well
[page 26]
enough. The Brown-Driver-Briggs lexicon and concordances illustrate the function-neutrality of these words.(43)

Lev, the root of levav, the first term in the series in the Shema, would seem to defy its translations, for it is translated in so many ways. Although in the overwhelming majority of cases it appears as "heart," it also is translated as "mind," "understanding," "wisdom," and significantly, "midst," and even once, "consent." It is the inside as opposed to the externals, occasionally even of things (the midst of the sea). It is seat of will and emotions: passions, appetites, trouble, courage, sorrow and joy, not just the intellectual faculties. And it is the seat of moral character. Clearly, it cannot be assigned to any one of the three IE functions, and the translation as "heart" is accordingly a reasonable one.

Nephesh, the second in the series, is usually translated as "soul," and Brown-Driver-Briggs even remark that it is often assimilated to lev. But its other meanings are instructive, for its second meaning is "life," and by extension, "body," "person," "beast," "man," "mind," "will," "desire," "self." It verges on being a substitute for pronouns -- the essential of a person stands for the person himself. If lev is the inner man, nephesh is the inner part that stands for the whole. Like lev, nephesh does not fit into an IE function, and "soul" preserves this neutrality, but the richness of its less frequent translations could easily be forgotten in its most common translation.

Meod is the most surprising of the three. It appears as "strength" or "might" in the translations of the Shema, and this might seem to fit into the IE second function. But far more common is its use as an adverb: "exceedingly," "greatly," and related ideas; the most frequent translation of all, outnumbering all others, is simply "very". The root can mean "to add," hence what one gets after addition -- more, muchness. Some translations show it as "money" or "property";(44) so it manifests neutral and possibly third function meanings, but not really a second function meaning of force, contrary to what appears in the translations. Even a meaning of "money" or "property" is more as an example of its sense indicating the person and everything remotely connected to him; since
[page 27]
these translations all occur from a Semitic language, any claim of a specifically third-function intent is dubious.

The three words lev, nephesh, and meod, taken in a series, constitute a progression from the inner man to his outermost self. This is utterly different from the IE way of analysing phenomena, and it should not be forced into or even toward IE categories, as it is in the Gospels, where mind, strength, and heart would appear to represent the three IE functions.

It is as if one were to plot phenomena in a space of several dimensions; three, for example. Then to use IE categories is to take the three IE functions as three Cartesian axes on which to plot the phenomena of interest. But the terms of the Shema might be more like thinking in terms of spherical coordinates: radius, latitude, and longitude. Only radius is named, the angular coordinates being asumed; lev, nephesh, and meod are not different coordinates, but a progression along one, the radial. One could of course complain at this point, when the Shema leaves out the "angular coordinates," because they give the functional differentiation so interesting to the IE mind. But one could just as well complain that the IE scheme de-emphasizes the "radial" coordinate, progression from the heart to the person in community.

Now it is possible to understand some of the misunderstandings and disputes in the theological history. If Hebrew and IE conceptualities are as different as one can suspect at this point, it is not surprising that Rabbinic Judaism does not lean heavily on Philo of Alexandria; even less could it make much sense of the Trinity. But within the IE tradition, disputes begin to make sense: if an analysis of a phenomenon in an IE language, in terms of the conceptual coordinates that are inevitable in an IE language, fails to preserve the symmetry and equivalence of the three functions, it will eventually fall short of a true account of the thing. If the issue is providence at its hardest, the bringing of blessings in disappointments, and if the three functions are of equal importance, and if they are each dependent upon the other two, then rejection of any one of exposure, limitation, and need will eventually result in rejection of the others.

It is also possible to understand devices such as appropriation, by which a phenomenon can be made to display more than just three features, and three features of a phenomenon can be "rotated" among the functions. Though it is viewed in IE triads, tripartite thinking, permitting
[page 28]
appropriation, can make some account for the fact that the world is not ultimately tripartite, but tripartition is only a way of apprehending the world. A thinker calls to the phenomenon to show itself, three features at a time. But concepts fit into the three functions only roughly, and some concepts can even be assigned to different functions as occasion dictates.


By way of conclusion and restatement, we may say that the challenge for a theology of monotheism is always to express the goodness of God in the problematic parts of life, transforming disappointments into blessings. In non-IE cultures one then needs to listen carefully to how the local conceptual system works, and make sure that the challenge of monotheism is met in all its parts, leaving no part untransformed and thereby able to subvert monotheism apparent in the rest of the parts.

We have seen how this process was worked out in the case of monotheism's move into the IE world. In the first approximation, the Son appears in the first function, redemption through exposure, and the Father appears in the second function, redemption through limitation, and the Holy Spirit appears in the third function, redemption through others' need. Later Trinitarian thought tidies up this correspondence, and then, in the theory of appropriations, gives it a flexibility sufficient to handle complex phenomena. At the same time, there is a move from the tripartite narrative and symbolic ideas conveyed genetically among the original IE cultures to a three-part system of abstract concepts suitable for use in philosophical theology.


We would like to acknowledge helpful conversations with C. Scott Littleton, Jonathan Weitsman, Gregory Rocca, O. P., David Aaron, and Louise Marlow, all of whom have read drafts of this paper and whose criticisms have led to amendments to the argument.


1. A guide to the considerable work of Georges Dumézil can be found in C. Scott Littleton, The New Comparative Mythology: An Anthropological Assessment of the Theories of Georges Dumézil, 3rd ed. (Berkeley and Los Angeles: University of California Press, 1963, 1973, 1982). A brief summary can be found in Mircea Eliade, A History of Religious Ideas (Chicago: University of Chicago Press, 1978), vol. 1, pp. 192-195.

2. "An Alternate Model From a Theological Perspective," in H. A. Otto, ed., The Family in Search of a Future (New York: Appleton-Century-Crofts, 1970), esp. pp. 32-33. The thesis appears in more detail in "A Theology of Visual Art," unpublished lecture for the "Space As Support" exhibition (1979) at the University Art Museum, University of California, Berkeley, and as the Dillenberger Lecture on the Arts for 1981, Berkeley, CA, Graduate Theological Union. In revised three-lecture form, presented at the Museum of Contemporary Art, Los Angeles (January 1994). Privately circulated.

3. Hereafter, we abbreviate "Indo-European" to simply "IE."

4. Le Nouvel Observateur, No. 949 (1983/01/14-20), pp. 50-54; cf. p. 53. See also Georges Dumézil, "Les festins secrets de Georges Dumézil," Entretiens avec Maurice Olender (Paris: Editions Gallimard, 1987).

5. Cf. Littleton, The New Comparative Mythology, p. 273, and Henry B. Nicholson, "Religion in Pre-Hispanic Central Mexico," in George F. Ekholm and Ignacio Bernal, eds., Handbook of Middle-American Indians, v. 10, pp. 395-446 (Austin: University of Texas Press, 1971).

6. Cf. Littleton, The New Comparative Mythology, p. 273.

7. Georges Dumézil, L'Idéologie Tripartie des Indo-Européens (Brussels: Latomus Collection, Vol. 31, 1958).

8. A description of the functions may be found in the Introduction to Littleton's The New Comparative Mythology.

9. Littleton, The New Comparative Mythology, p. 5.

10. Georges Dumézil, Mitra-Varuna: An Essay on Two Indo-European Representations of Sovereignty, 1940, 1947; trans. Derek Coltman (New York: Zone Books, 1988). See esp. pp. 74-80. Dius Fidius is guarantor of oaths, p. 78.

11. Cf. Littleton, p. 227.

12. Cf. Littleton, The New Comparative Mythology, pp. 64-65, and Dumézil, L'Idéologie Tripartie, pp. 62-67. Mitra-Varuna is focussed specifically on this aspect of IE religion.

13. Dumézil, L'Idéologie Tripartie, p. 62.

14. Cf. Littleton, The New Comparative Mythology, pp. 49-53.

15. Cf. Brian K. Smith, "Canonical Authority and Social Classification: Veda and Varna in Ancient Indian Texts," History of Religions 32:2 (1992) 103-125, esp. p. 105.

16. The outcastes are function-neutral, and so neither compromise the symmetry of the three-function system, nor constitute a fourth function based on some principle different from the given three.

17. Cf. "Canonical Authority and Social Classification: Veda and Varna in Ancient Indian Texts," History of Religions 32:2 (1992) 103-125.

18. Cf. Littleton, pp. 70-71.

19. Cf. Littleton, pp. 70; 108-122.

20. Cf. Littleton, p. 71.

21. Cf. Littleton, pp. 68-69.

22. Cf. Littleton, pp. 69; 141 ff.

23. Cf. Robert Graves, The Greek Myths (Baltimore: Penguin Books, 1955), vol. 2, p. 268.

24. TNK = Torah, Nevi'im, Ketubim; i. e., the Law, the Prophets, the Writings. Also TaNaK or Tanakh; the Christian Old Testament.

25. Edward C. Hobbs, "Pluralism in the Biblical Context," in Wilhelm Wuellner and Marvin Brown, eds., Hermeneutics and Pluralism (Berkeley: Center for Hermeneutical Studies, 1983); also Mark S. Smith, The Early History of God: Yahweh and the Other Deities in Ancient Israel (San Francisco: Harper and Row, 1990).

26. Mircea Eliade, Cosmos and History (Princeton University Press, 1954, 1965); and Merold Westphal, God, Guilt, and Death (Bloomington: Indiana University Press, 1984). See especially chapters 10 and 11 of Westphal.

27. James A. Sanders, Torah and Canon (Philadelphia: Fortress Press, 1972), pp. 73-90.

28. Sanders, p. 86.

29. Radical Monotheism was published in 1960 (New York: Harper and Row), and though Faith on Earth: An Inquiry Into the Structure of Human Faith (New Haven: Yale) was published only in 1989 it was written at the same time as Radical Monotheism. Niebuhr's monotheism is examined at greater length in Andrew P. Porter, "H. Richard Niebuhr's Doctrine of Providence in the Light of Martin Heidegger's Philosophy," Ph.D. dissertation (Berkeley, California: Graduate Theological Union, 1991).

30. Cf. Westphal, p. 200, quoting Eliade, speaking of "archaic man's refusal to accept himself as a historical being."

31. James C. Russell, in The Germanization of Early Medieval Christianity (Oxford University Press, 1994), by contrast to our problem, does have to take into account the exilic character of medieval Christianity as it met German mimetic religion. His concern is with actual historical particulars, where ours is an investigation into the logic of covenantal religion.

32. Niebuhr's characterization of Biblical religion at this moment of its faith can be found in Radical Monotheism, p. 122: "Though it slay us, yet will we trust it" (Job 13:15, AV). This is hardly a uniquely Christian commitment. It is noteworthy that the Talmud Berakhot 60b (9.5 in the Mishnah) emphasizes that the believer is to give blessings for disappointments just as for the gratifications of life.

33. "An Alternate Model From a Theological Perspective," cited above, pp. 32-33.

34. Cf. C. Scott Littleton, "Georges Dumézil and the Rebirth of the Genetic Model: An Anthropological Appreciation," in Myth in Indo-European Antiquity, ed. Gerald James Larson, C. Scott Littleton, and Jaan Puhvel (Berkeley and Los Angeles: University of California Press, 1974).

35. Cf. Bertrand de Margerie, S.J., The Christian Trinity in History (orig. 1975; trans. pub. Still River, MA: St. Bede's Publications, 1982), and William J. Hill, O.P., The Three-Personed God: The Trinity as a Mystery of Salvation (Washington, DC: Catholic University Press, 1982). See also J. N. D. Kelly, Early Christian Doctrine (New York: Harper and Row, 1960), chapters 4, 5 and 10.

36. Cf. "The Doctrine of the Trinity and the Unity of the Church," Theology Today 3 (1946) 371.

37. Harry A. Wolfson, (New York: Macmillan, 1967), vol. 6, pp. 151-155, esp. p. 154.

38. Philadelphia: Jewish Publication Society, 1983.

39. Claude Welch, In This Name: The Doctrine of the Trinity in Contemporary Thought (New York: Scribners, 1952), and David Brown, The Divine Trinity (Chicago: Open Court, 1985).

40. We have in mind Thomas F. Torrance's Divine and Contingent Order (New York: Oxford University Press, 1981) at this point, though other works could doubtless be cited as well.

41. The New Comparative Mythology, p. 231.

42. Henry B. Nicholson, "Religion in Pre-Hispanic Central Mexico," cited above.

43. Francis Brown, S. R. Driver, Charles A. Briggs, A Hebrew and English Lexicon of the Old Testament (Oxford: At the Clarendon Press, 1952).

44. See the Anchor Bible Deuteronomy.