The Attributes

Scripture does not hand over a list. It says that God is holy and jealous and merciful and everlasting, in narratives and psalms and prophecies, scattered across sixty-six books, and the work of gathering those into headings is the work this track exists to do. Every gathering imposes an order, and the order is never neutral. Before you learn any attribute you should know what the filing system is doing to it.

The scheme most common in Protestant dogmatics divides the incommunicable attributes from the communicable ones. Incommunicable: aseity, simplicity, infinity, eternity, immutability, omnipresence, unity. These have no creaturely counterpart at all. Communicable: knowledge, wisdom, goodness, love, mercy, holiness, righteousness, will. Of these there is a creaturely likeness, faint but real, since a man may be wise and no man may be eternal. The division is old and useful and it is taught in the standard Reformed textbooks, Herman Bavinck's Reformed Dogmatics and Louis Berkhof's Systematic Theology among them.

It also has a defect that Bavinck himself names while keeping the scheme. If God is simple, then his love is not a separate ingredient from his eternity, and it will not do to say that one of them is shared with creatures and the other is not, as though God came in two grades of property. Worse, the division does practical work that nobody votes on. Put love in the communicable column and immutability in the incommunicable one, and you have already suggested which of the two is the more properly divine, and therefore which one gives way when a reader thinks they collide. Scheme first, conclusions second, and the conclusions look like discoveries.

A second scheme divides absolute attributes, God as he is in himself, from relative attributes, God as he stands towards the world. This has the advantage of matching the way the Bible mostly speaks, since scripture describes God acting far more than it describes God existing. Its own defect is that eternity and omnipresence land on the relative side, as relations to time and space, which is not where most of the tradition wants them. A third approach, the medieval threefold way, reaches attributes by three routes at once: denying of God what belongs to creatures, attributing to God the cause of every perfection found in creatures, and attributing every perfection to him in an eminent degree. Aquinas structures the opening questions of the Summa Theologiae largely by the first route, asking what God is not before asking what he does.

Two other arrangements deserve mention because they are not Protestant textbook arrangements and they are held by very large parts of the church. Karl Barth refused to call them attributes and called them perfections, and paired them: every perfection of the divine loving is matched with a perfection of the divine freedom, grace with holiness, mercy with righteousness, patience with wisdom, so that neither list can be read without the other. His reason was that a list allows a reader to pick a favourite, and pairing does not. The Eastern Orthodox tradition, following the line traced in the last lesson, is reluctant to treat the attributes as a catalogue of properties inside the divine essence at all, since the essence is unknowable. What we name are the uncreated energies, God as he acts and gives himself, and the attributes are descriptions of the God who meets us rather than an inventory of what he is made of.

So the working rule for this lesson. A scheme is a tool, it is defensible, and it is not innocent. When you meet an attribute in what follows, ask what the arrangement is doing as well as what the doctrine says. And expect the schemes to disagree, because the churches that produced them disagree about more than filing.

Aseity is the doctrine that God is from himself, a se, and depends on nothing. He did not come to exist, is not sustained, is not completed, is not enriched. He has life in himself: John 5:26 says, 'For as the Father has life in Himself, so also He has granted the Son to have life in Himself.' Paul at the Areopagus says God is not 'served by human hands, as if He needed anything, because He Himself gives everyone life and breath and everything else' (Acts 17:25). Romans 11:36 closes the argument: 'For from Him and through Him and to Him are all things.' The Latin noun is aseitas, and the doctrine is not disputed between the great traditions.

Its anchor text has been Exodus 3:14, and this is where a careful student has to slow down. Moses asks for a name. God answers with a Hebrew clause, ehyeh asher ehyeh, and the BSB prints it as 'I AM WHO I AM'. Three facts about that clause need to be on the table together, and most popular treatments give one and suppress the others.

First, the Hebrew verb form is imperfect, which in Hebrew is not a tense in the way English tenses are. It can be rendered as present, 'I am who I am', or as future, 'I will be who I will be'. The BSB puts the present in its text and the future in a footnote ('Or I WILL BE WHAT I WILL BE'), which is a translators' choice that the Hebrew does not settle, and the future reading deserves full weight. The decisive piece of evidence is two verses earlier: at Exodus 3:12 the identical form ehyeh is translated 'I will surely be with you', in a promise of God's presence with Moses in the task. The same word in the same conversation is rendered as a future promise in one verse and as a statement of being in another, and the translator is making a judgement both times.

Second, the construction is the idem per idem shape, a verb repeated across a relative pronoun. Exodus 33:19 is the same shape in the Hebrew, which repeats the verb on either side of the relative; the BSB gives it as 'I will have mercy on whom I have mercy, and I will have compassion on whom I have compassion.' That construction characteristically answers a question while refusing to be pinned down by it. On this reading, God is not supplying Moses with a metaphysical definition; he is declining to be handed a name of the sort that ancient religion used to get leverage on a god, and turning the question back. Brevard Childs, in his commentary on Exodus, argues that the formula deliberately resists both a purely ontological and a purely functional reduction, and that this reserve is the point.

Third, the Greek. The Septuagint renders the second half as ego eimi ho on, 'I am the one who is', which converts the clause into a statement about being. That is a translational decision, taken in Alexandria in the third or second century before Christ, and it carries a weight the Hebrew may not carry on its own. But it is not an outsider's imposition on the Bible either, because the New Testament writes in that Greek. Revelation calls God 'Him who is and was and is to come', a phrase built out of the same participle and deliberately stretched. And the whole subsequent tradition, Greek and Latin, reads Exodus 3:14 through the Septuagint. The Vulgate has ego sum qui sum, and Aquinas, at Summa Theologiae, first part, question thirteen, article eleven, concludes that Qui est, He Who Is, is the most proper name of God, precisely because it signifies being itself rather than any particular form. Etienne Gilson called the whole line of thought the metaphysics of Exodus.

Now say what is honest. If the Hebrew is doing what Childs and many Old Testament scholars think it is doing, then the verse is about God's sovereign, self-determined presence with his people rather than about the metaphysics of necessary existence, and the whole edifice built on it has been built on the Greek. Martin Buber and Franz Rosenzweig translated it into German as a promise of presence, along the lines of 'I will be there as the one who will be there', and their rendering has been influential. Set against that, three things tell the other way. Verse 15 immediately ties the clause to the covenant name and to the God of the fathers, so the verse is not a detour into philosophy at all but an identification. The Greek reading is inside the canon by the time Revelation is written, so a Christian reader is not free to treat it as a later accretion. And the doctrine of aseity does not actually depend on this verse: Acts 17:25, John 5:26, Psalm 90:2 and Romans 11:36 would carry it if Exodus 3:14 were removed. That last point is the one to remember. A doctrine that rests on a disputed clause is fragile; this doctrine does not rest there, and it is better taught from the texts that are not disputed.

One live objection to aseity itself, since the doctrine is agreed but its implications are not. If God is in no way affected by us and gains nothing from us, critics ask, in what sense does he love us? Charles Hartshorne, developing Alfred North Whitehead, argued for a dipolar God who is unchanging in character and genuinely changed in experience by what creatures do, on the ground that a love which receives nothing is not what anybody means by love. The classical answer runs the other way and is worth hearing at its strength: it is precisely because God needs nothing that his love for you is entirely gratuitous. A needy love is interested love. Acts 17:25 makes that argument in its own setting, against a religion of feeding and housing the gods, and the pastoral cash value is considerable. God did not make you because he was lonely, and he does not keep you because you are useful.

Divine simplicity says that God is not composed. Not of physical parts, obviously, since he has no body; but also not of metaphysical parts. He is not a substance with attributes attached, as a wall is a thing that has a colour. He is not a mixture of essence and existence, so that what he is and that he is are two facts about him. His attributes are not ingredients that together make him up. Whatever is in God is God. The consequence that startles people on first hearing is that God's justice and God's mercy are not two things in him; they are one reality, which our minds, working with creaturely concepts, can only take in under several headings.

Two misunderstandings should be cleared at once, because they account for most of the popular objections. Simplicity does not mean that God is easy to understand; it means uncomposed, and the previous lesson should have made clear that uncomposed and incomprehensible go together. And it does not mean that the distinctions we draw are false. On the classical account the distinctions have a foundation in God even though there is no composition in God, which is why the doctrine does not license saying that mercy and wrath are interchangeable.

The doctrine's credentials are as strong as any in this module. Augustine states it in The City of God, book eleven, where the divine nature is called simple because it has nothing it could lose, so that being and having are the same in God. Anselm and Aquinas develop it, Aquinas devoting the third question of the Summa Theologiae to it and placing it before everything else he says about God. It is conciliar in the Roman Catholic Church, stated at the Fourth Lateran Council in 1215 and again at the First Vatican Council in 1870, which calls God altogether simple and unchangeable. It is confessional in Protestantism: the Westminster Confession, in its second chapter, describes God as without body, parts, or passions, and the same phrase stands in the first of the Thirty-Nine Articles and behind the first article of the Belgic Confession. The Eastern churches affirm divine simplicity too, though they hold it alongside the essence and energies distinction, which is exactly why that distinction is controversial in the West. A student should not be told that this is a Thomist speciality. It is the common inheritance of the church.

And it is nevertheless a live argument now, in a way it was not for most of the last fifteen centuries, and a lesson that suppressed that would be dishonest. The critics are serious Christian philosophers, not sceptics.

Alvin Plantinga's Does God Have a Nature?, published in 1980, put two objections into wide circulation. If God is identical with his properties, then God is a property; but properties are abstract objects and no abstract object is a person who knows and acts, so the doctrine seems to depersonalise God. And if each of God's properties is identical with each other one, then God has exactly one property, and the difference between God's knowing the world and God's creating it disappears. Nicholas Wolterstorff pressed related concerns. More recently Ryan Mullins has argued that simplicity entails modal collapse: if God is identical with his act, and God exists necessarily and could not have been otherwise, then his act of creating this world could not have been otherwise either, so creation is not free and this world is necessary. That conclusion is one almost no Christian will accept, so if the argument works it is a reductio of the doctrine.

The defenders answer at the same level. James Dolezal, in All That Is in God in 2017, and Steven Duby, in Divine Simplicity: A Dogmatic Account in 2016, argue that the doctrine is mainly a set of denials, a refusal to treat God as composed, rather than a positive claim that God is an entity of the kind called a property; on that construal Plantinga's first objection targets a Platonism about properties that classical theists never held. Eleonore Stump and Brian Davies argue that the single-property objection assumes that our several concepts must correspond to several things in God, which is precisely what the doctrine denies while granting that the concepts are not arbitrary. On modal collapse, Christopher Tomaszewski argued in the journal Analysis in 2019 that the standard version of the argument is formally invalid, substituting an identity into a modal context where such substitution does not go through; Mullins and Joseph Schmid have since produced refined versions, and the exchange is still running. Nobody should report a winner. What can be reported is that the debate is now technical, that it is being conducted by people who pray, and that it has not ended.

There is a second dispute, inside conservative Protestantism, and it is worth naming because students meet it in sermons rather than in journals. Dolezal's book charged a range of respected evangelical writers, John Frame and Bruce Ware among them, with what he called theistic mutualism: a doctrine of God that allows genuine give and take between God and the world and thereby, in his judgement, abandons simplicity and immutability as the confessions state them. Frame replied in a review that the classical formulation, as Dolezal states it, cannot accommodate the biblical narrative of a God who responds, and that the confessions were not intended to rule that narrative out. Both men appeal to the Westminster Confession. That is the shape of the argument, and a reader who wants to see it whole should read Dolezal's book and Frame's review together rather than trusting either summary, including this one.

Immutability says God does not change. The texts are direct. Malachi 3:6: 'Because I, the LORD, do not change, you descendants of Jacob have not been destroyed.' Notice the logic there, because it is the pastoral point of the doctrine: God's unchangeableness is the reason Israel still exists. James 1:17 says every good gift comes down from the Father of the heavenly lights, 'with whom there is no change or shifting shadow'. Numbers 23:19: 'God is not a man, that He should lie, or a son of man, that He should change His mind.' Psalm 102:25-27 sets the perishing of heaven and earth against the God whose years have no end. Hebrews 6:17-18 makes the unchanging nature of God's purpose the ground of a believer's assurance.

And then there are the other texts, which a teacher must not leave out and which the tradition has never pretended were absent. Genesis 6:6: 'And the LORD regretted that He had made man on the earth, and He was grieved in His heart.' Exodus 32:14 says the LORD 'relented from the calamity He had threatened to bring on His people'. Jonah 3:10 says the same of Nineveh. Jeremiah 18:7-10 lays it out as a stated policy: if a nation under threat turns, 'I will relent of the disaster I had planned to bring', and if a nation under promise does evil, 'I will relent of the good I had intended for it'. Hosea 11:8-9 has God say, 'My heart is turned within Me; My compassion is stirred!', and then, in the same breath, 'For I am God and not man'. And 1 Samuel 15 affirms the regret in verse 11 and again in verse 35, and denies it in verse 29 (where the BSB's English is 'change His mind'), inside one chapter.

The classical position reads the second set as accommodation. God describes himself in terms fitted to creatures who have no other terms available, and the language of regret and relenting describes a real change in the situation and in the relation between God and a people, not a change in God's eternal will or being. Calvin's image is that God speaks to us as a nurse speaks to an infant, lisping rather than lying. On this account Jeremiah 18 is the key: what looks like God changing is God's unchanging principle meeting a changed people, and a God who did not relent when Nineveh repented would be the one who had changed. Aquinas adds a technical point that many find clarifying and others find evasive: the relation between creature and God is real in the creature and a relation of reason in God, so that when I move from enmity to friendship with God, something changes, and it is I.

Open theism reads the second set straight and is the sharpest modern challenge. The essays collected as The Openness of God in 1994, by Clark Pinnock, Richard Rice, John Sanders, William Hasker and David Basinger, and Gregory Boyd's God of the Possible in 2000, argue that God genuinely responds, that the future free acts of creatures are not yet settled, and that prayer therefore makes a difference that is more than psychological. Their exegetical move on the other set of texts is worth stating properly rather than dismissing: they argue that Numbers 23:19 and 1 Samuel 15:29 are situational denials, each refusing a particular reversal that God will not make, rather than general metaphysical statements, and that reading them as general denials makes 1 Samuel 15 self-contradictory, whereas reading them as specific makes the chapter coherent. The classical reply is that both verses give a reason, that God is not a man, which is a claim about the kind of being God is and not about a single case, and that the same reading-as-specific could be run on the relenting texts with equal ease.

Positions in between are numerous and should not be ignored, since most working pastors hold one. Karl Barth preferred to speak of God's constancy rather than immutability, arguing that God is unwaveringly the one he is, including in freely becoming what he was not, so that the incarnation is not an exception to his constancy but an expression of it. Bruce Ware and others have argued for ontological immutability with real relational mutability. Kevin Vanhoozer has proposed reading the relenting texts as God's communicative action rather than as a window into divine psychology. These are attempts to keep the two sets of texts without subordinating either, and they are criticised from both directions for exactly that.

Impassibility is the neighbouring doctrine and it is more often misstated than any other in this lesson. It does not mean that God is indifferent or emotionless. It means that God is not acted upon: nothing outside God causes a state in him, and his love, compassion and anger are his own act rather than something a creature produces in him. Anselm, in the eighth chapter of the Proslogion, puts the classical position with unusual candour: God is compassionate in terms of our experience and not compassionate in terms of his own being, since he acts towards us as the compassionate act and does not undergo what we undergo.

The twentieth century saw the most serious challenge to it in the history of the doctrine, and the challenge came out of suffering rather than out of the study. Kazoh Kitamori's Theology of the Pain of God, written in Japan in 1946, and Jurgen Moltmann's The Crucified God in 1972 argued that the cross reveals God, that a God incapable of suffering cannot be the God revealed there, and that after Auschwitz a doctrine of divine apathy is not merely wrong but obscene. Dietrich Bonhoeffer's line from prison, that only a suffering God can help, has carried the argument further than any monograph. Behind all of it sits the older claim, associated with Adolf von Harnack, that impassibility entered Christian theology from Greek philosophy rather than from scripture.

The defence has been equally substantial and rests partly on history. Thomas Weinandy, in Does God Suffer? in 2000, argues that suffering is a deficiency, that a God who suffers as creatures do is a God who needs saving rather than one who saves, and that the biblical language of divine grief describes God's utterly active love rather than a passion inflicted on him. Paul Gavrilyuk, in The Suffering of the Impassible God in 2004, attacks the Hellenization story directly on the evidence, arguing that the Fathers used impassibility apophatically, to mark divine emotion off from creaturely passion, and that the same Fathers cheerfully wrote paradoxes about the impassible one suffering. That last point is the hinge, and it is christological. Cyril of Alexandria taught that the impassible Word suffered in the flesh, and the Second Council of Constantinople in 553 endorsed the formula that one of the Trinity suffered in the flesh. So the classical tradition does say that God the Son suffered, and it says he suffered in his human nature. Passibilists reply that the qualification takes back what it gives. Classical theists reply that without the qualification suffering becomes a divine attribute, and a God who suffers by nature is a God with a problem of his own. Both of those are real arguments and the church has not settled between them.

Omniscience. God knows all things: Psalm 139 traces it through every corner of a life; Hebrews 4:13 says nothing in all creation is hidden from his sight; Isaiah 46:9-10 has God stake his deity on 'I declare the end from the beginning, and from ancient times what is still to come'. The puzzle is old and it is genuine. If God infallibly knew yesterday what you will do tomorrow, then what you will do tomorrow cannot fail to happen, and it is hard to see in what sense it is up to you.

Four answers are given, and they should be stated as answers rather than as riddles. The Boethian answer denies the premise about yesterday: God is not in time, and Boethius, writing in prison around 524, defined eternity as the whole, simultaneous and perfect possession of unbounded life. On that account God does not foreknow anything; he knows everything as present, and my seeing you sit does not make you sit. The Augustinian and Reformed answer accepts that God's decree settles all things and denies that settledness destroys freedom, defining freedom as acting from one's own will rather than as the power to have done otherwise; the Westminster Confession states in its third chapter that God ordained whatsoever comes to pass, 'yet so as thereby neither is God the author of sin, nor is violence offered to the will of the creatures'. The Molinist answer, from the Jesuit Luis de Molina's Concordia in 1588, posits middle knowledge: before deciding what to create, God knows what any possible free creature would freely do in any possible circumstance, and he chooses which of those worlds to actualise, so that the outcome is both certain and freely chosen. Open theism denies that there are yet truths of the relevant kind to know: the future free acts of creatures are not there to be known, so not knowing them is no more a limit on omniscience than not making square circles is a limit on omnipotence.

Each has a standing objection its own defenders have to answer. Against Boethius: timeless knowledge is itself hard to explain, and infallibility seems to do the damaging work whether or not it is early. Against the Reformed: many readers judge the reconciliation verbal, and the Confession's own 'yet so as' is a boundary marker rather than a mechanism. Against Molina: the grounding objection, pressed by Robert Adams and William Hasker, asks what makes it true that a particular person would freely do a particular thing in circumstances that never occur, when the person does not exist and the choice is never made. Against open theism: Isaiah 41 to 48 stakes the case for the LORD against the idols precisely on announced futures, and the position is a recent minority one in a church that has held otherwise for most of its history. Two historical notes keep this from feeling abstract. Rome appointed a commission, the congregatio de auxiliis, which sat from 1598 to 1607 on the Molinist and Dominican dispute and closed without deciding, forbidding each side to call the other heretical. And the Evangelical Theological Society voted in 2003 on expelling two open theist members; the motion did not reach the required majority. Churches have found this hard, not just philosophers. One more distinction, because it is constantly muddled: classical Arminianism affirms exhaustive divine foreknowledge and is not open theism. Jacob Arminius held it, and the Remonstrant articles of 1610 assume it. Whatever an Arminian and a Calvinist are arguing about, it is not whether God knows the future.

Omnipotence. God can do all things: Jeremiah 32:17 and 27, Matthew 19:26, Luke 1:37 (the BSB prints 'no word from God will ever fail' and footnotes the alternative 'For nothing will be impossible with God', so the case for omnipotence from this verse rests on that alternative). The schoolroom version of the puzzle asks whether God can make a stone too heavy for him to lift, and Aquinas answered it in the twenty-fifth question of the Summa's first part. God's power extends to whatever is absolutely possible; what implies a contradiction is not a thing that could be done and fails to be done, it is not a thing at all, and so it marks no limit. Rene Descartes took the other road in his letters to Mersenne in 1630, holding that God's will creates even the eternal truths and that the logical necessities are necessary because God willed them, which preserves omnipotence at the price of making logic contingent. Peter Geach, in a 1973 article, argued that the strict philosophical term is indefensible and that Christians should say what the creed says, that God is almighty, since the Greek pantokrator means all-ruling rather than able to do absolutely anything. And scripture itself states limits that are not deficiencies: Titus 1:2 says God cannot lie, Hebrews 6:18 that it is impossible for God to lie, 2 Timothy 2:13 that he cannot deny himself. A power to act against one's own nature would be a defect, not an addition.

Omnipresence. God is wholly present everywhere: 'Where can I go to escape Your Spirit? Where can I flee from Your presence?' (Psalm 139:7); 'Do I not fill the heavens and the earth?' declares the LORD (Jeremiah 23:24); and Solomon at the dedication of the temple, 'The heavens, even the highest heavens, cannot contain You, much less this temple I have built' (1 Kings 8:27). The classical clarification is that God is not present by being spread out, as a gas fills a room, since he has no parts to distribute. He is present by essence, power and knowledge: wholly present at every point, not partly here and partly there. Aquinas puts it at the eighth question of the Summa's first part, and the formulation is shared across the traditions.

Two qualifications that matter pastorally. First, scripture also speaks of God being present in special ways, in the tabernacle, in the temple, in the assembly, in the believer, and it is not contradicting itself; omnipresence is about God's relation to all that exists, while covenant presence is about blessing, favour and dwelling. Hell in the classical account is not a place God is absent from but a place where his presence is not blessing. Second, the Lutheran and Reformed churches divided in the sixteenth century over whether the human nature of Christ shares the divine omnipresence, the Lutherans affirming a real communication of that attribute to the humanity and the Reformed denying it in the doctrine their opponents nicknamed the extra Calvinisticum. That argument belongs to the Christology module, but its root is here, and it shows what is at stake in a doctrine that can look like a list of adjectives.

A closing word about all of this. The attributes are not a specification sheet. Every one of them, treated apart from God's acts in Israel and in Christ, hardens into something that can be argued about without being worshipped. Exodus 34:6-7 remains the pattern: when God proclaims his own name it comes out as compassionate, gracious, slow to anger, abounding in loving devotion and faithfulness, forgiving iniquity, transgression and sin, and yet by no means leaving the guilty unpunished. Every item in this lesson has to be readable back into that sentence, and if your account of simplicity or of impassibility cannot be, the fault is more likely in your account than in Exodus 34.