Classical Theism Against Theistic Personalism

Two philosophers of religion can each spend a career on arguments for the existence of God and never once notice that they are arguing for different conclusions. One of them means a being without parts, outside time, whose essence is his existence, of whom nothing can be said in the way that things are said of creatures. The other means a person without a body who is everlasting, free, and knows everything, and who acts in the world by deciding to. Both call the result God. Both take themselves to be defending the same tradition. The gap between them has become the organising dispute of contemporary philosophical theology, and it does not fall along the line between theists and atheists.

Two portraits

Set the two accounts beside one another and the difference is not a matter of emphasis.

On the classical account, God is not a member of any kind. There is no genus God falls under and no other possible instance of it. God does not have properties, since having a property would make him composite. God does not undergo anything, since undergoing takes time and God has no temporal location. God does not respond to prayer in the sense of forming an intention after hearing it, because forming an intention after anything is a change of state. God's relation to the world is not symmetrical: creatures are really related to God, and God is not, in the same sense, really related to creatures. Every one of these claims is a denial, and the tradition insists that the denials leave us without a positive conception. Aquinas is explicit that we do not know what God is.

On the personalist account, God is a person without a body. He has beliefs, and his beliefs are all true. He has intentions, and he acts on them. He is perfectly good, which means that he chooses well among the options open to him, which in turn requires that there be options open to him. He knew before Tuesday what would happen on Tuesday, and after Tuesday he knows it as past. He is affected by what creatures do in the sense that he responds to it. The vocabulary is continuous with the vocabulary we use for each other, and that continuity is the point: it is what allows the claims to be assessed by ordinary philosophical means.

The label needs handling with care, and this is not a pedantic point. It was coined by an opponent, and almost nobody described by it uses it of themselves. Richard Swinburne does not call his position theistic personalism; he calls it theism. So there is a live worry that the category is a polemical construction that groups together philosophers who share little except a target on someone else's map. Anyone assessing the dispute should keep open the possibility that the disagreement is partly an artefact of how it has been framed.

What the personalists were actually doing

The unflattering summary of theistic personalism is that analytic philosophers, wanting God to be tractable, cut him down to something they could reason about. That summary misdescribes the project.

Swinburne's The Coherence of Theism (1977) sets out the problem the analytic tradition took itself to be solving. The mid-century objection to theism was not that the arguments for God's existence fail. It was that the sentence "God exists" fails to express a proposition at all, because the predicates in the theistic vocabulary cannot be given conditions of application. If the verificationist challenge is taken seriously, then a theology that refuses to say anything positive about God has not survived the challenge; it has conceded it. Swinburne's response was to build a statement of theism that could be assessed for consistency, which required stating it in terms that carry ordinary meaning.

The dialectical situation mattered as much as the arguments. Philosophy of religion had been a marginal subject in English-language departments for a generation, on the assumption that its central sentences failed to say anything. The revival that began with Alvin Plantinga's God and Other Minds in 1967 and continued through Swinburne's trilogy in the following decade proceeded by treating theism as a hypothesis with content, statable in the same vocabulary as any other and defensible by the same standards. That strategy would have been unavailable to anyone who began by announcing that we cannot know what God is. Whatever its metaphysical costs, it succeeded institutionally: the subject came back.

Seen this way, the personalist move is a response to pressure that the classical tradition never faced. Aquinas was not writing for an audience that doubted whether talk about a first cause was meaningful. Once that doubt is on the table, the apophatic strategy looks less like humility and more like an evasion, and the incentive to say something checkable becomes very strong.

There is a second motivation, and it is religious rather than methodological. The texts that the three monotheisms actually read describe a God who makes promises, hears complaints, changes his mind, and is grieved. A doctrine that classifies all of this as figurative has to explain what it is figurative for, and the classical answer, that these are ways of describing effects in creatures rather than states in God, strikes many readers as a way of keeping the language while emptying it. Nicholas Wolterstorff's argument from the previous rung is an instance of the pattern: start from something the tradition says God does, and ask what has to be true of God for it to be true.

Notice that perfect being theology is common ground. Both camps say God is whatever is maximally perfect. They divide on what perfection turns out to require, and the division is traceable to a single prior question: whether being unaffected is a perfection or a defect. The classical answer is that dependence on anything external is a defect and therefore that impassibility is required. The personalist answer is that a being incapable of responding to what happens is worse, not better, than one capable of it. Neither side is refusing to reason from perfection. They are running the same method on different intuitions about the ranking.

The charge of making God a being

The classical complaint is not primarily that the personalist God is too small. It is that the personalist God is the wrong category of thing.

The phrase has become the standard shorthand for the classical objection. Paul Tillich's Systematic Theology, the first volume of which appeared in 1951, made the formula current in twentieth-century theology by insisting that God is not a being but the ground of being, and that calling God the highest being is a category mistake rather than a compliment. David Bentley Hart's The Experience of God (Yale, 2013) presses the same charge against contemporary analytic theism and against its atheist critics jointly: his claim is that both sides have been arguing about a very powerful thing within the universe of things, and that the God of the classical tradition, on whom the existence of everything else depends, has not entered the discussion.

Behind this sits Heidegger's charge of ontotheology, developed in Identity and Difference (1957). The complaint is that Western metaphysics has consistently answered the question about being by producing a supreme being that grounds the rest, and that this manoeuvre turns God into an object of explanation. Heidegger's conclusion is that the god arrived at by this route is one before whom the religious attitudes have no purchase. The argument is not a defence of classical theism, and Aquinas is among Heidegger's targets, but the classical theist can use half of it: whatever else is true, a God who functions as the last term in a causal series is a God who has been fitted into a framework rather than one who accounts for the framework.

Critics of the modern conception often attach a genealogy to the complaint. The story, associated with the Radical Orthodoxy movement and repeated in various forms by classical theists since, locates the decisive turn in Duns Scotus, who held that "being" applies to God and to creatures in the same sense rather than by analogy. Once being is univocal, the thought runs, God takes his place on a single scale with everything else and differs only by having infinitely much of what creatures have a little of, and from there the road to a God who is the biggest thing in existence is short. Historians of medieval philosophy have contested the story vigorously, both on what Scotus actually held and on whether a semantic thesis about a term could carry that much weight. It is worth knowing that the genealogy exists, and worth treating with the caution due to any account that traces a large modern failure to one thirteenth-century decision.

The personalist reply to all of this is short and not obviously wrong. If God is not a being, then the sentence "God exists" is either false or meaningless, and the classical theist has to explain what is being asserted when the tradition says God is real. The reply from the classical side is that "exists" is not univocal, and that God's existing is related to a creature's existing by analogy rather than by identity of sense. Which returns the dispute to a question about language that rung 9 takes up in its own right, and which has a familiar shape: the analogy doctrine is either a real semantic proposal or a promissory note, and it is not always clear which.

The anthropomorphism objection

Hume gives the objection its sharpest classical statement, though it is worth noticing that he puts it in the mouth of a character who is defending orthodoxy. In the Dialogues Concerning Natural Religion, published in 1779 after his death, the pious Demea attacks Cleanthes for anthropomorphism, on the ground that reasoning from the world to a designer resembling a human mind reduces God to something finite and comprehensible. Hume's own use of this is corrosive to both parties: if the argument from design gives us a designer, it gives us one whose attributes must be read off the effects, and the effects do not license infinity, unity, or moral perfection. The classical theist and the sceptic can therefore agree on the criticism while disagreeing entirely about what follows from it.

The objection does not run only in one direction. The personalist counter-charge is that classical theism is anthropomorphic too, just about different things. Its central notions, causation, explanation, dependence, act and potency, are drawn from the analysis of ordinary things and then applied to something declared to be in no way ordinary. If applying "loves" to God is illegitimate anthropomorphism, it is unclear why applying "causes" is not. The classical reply is that causal and existential vocabulary is transcendental rather than categorial: it does not carve up a domain but applies wherever anything exists at all. Whether this distinction can be drawn without begging the question is one of the places the argument actually stalls, and honest defenders of both positions acknowledge it.

There is a further complication that neither camp likes to dwell on. Debunking explanations of religious belief, of the kind the intermediate tier covered under projection theory, are a great deal more comfortable with the personalist God than with the classical one. A cognitive story about agency detection predicts belief in powerful invisible agents; it does not obviously predict belief in subsistent existence itself. Classical theists have occasionally used this as an argument, on the ground that their God is not the one the debunking story explains. The personalist can reply that a conception no cognitive process would ever generate is a conception that requires explaining how anyone came to hold it.

What the split does to the standard problems

Abstract as the disagreement sounds, it changes the shape of nearly every familiar problem in the subject, and this is the best evidence that something substantive is at stake.

Omnipotence is the clearest case. The paradox of the stone assumes a God who performs tasks, so that the question is whether the list of tasks he can perform includes one he could not then perform. On the classical account there is no list, because power is not a capacity God has for producing effects on request; it is identical with God, and the question of whether a simple act could produce a self-defeating result does not arise in the form the paradox needs. Whether this is a solution or a refusal to engage is disputed, but the paradox plainly does not transfer intact.

Petitionary prayer divides the two camps at the level of what the practice is for. On the personalist account, asking makes a difference because God takes the request into account in deciding, which is intelligible and which requires that God's decisions come after the request. On the classical account, God's single eternal act already includes whatever it includes, and the asking is part of the created order that act establishes rather than an input to it. The classical theist can say that the prayer is a genuine cause of the outcome without being a cause of God's willing it, which preserves the practice but not the picture most people have of it.

Divine freedom divides them again, and here the classical side is under more pressure. A being with no unrealised potency does not choose between options in the ordinary sense, which is why the modal collapse literature exists. The personalist has an easy answer to how creation could have been otherwise and a hard time explaining what fixes God's own nature. Each camp's easy question is the other's difficult one, which is a sign that the difficulty is being relocated rather than removed.

The third position

The dispute is usually presented as a pair, which conceals a developed alternative that has been available since the middle of the last century and that accepts neither side's package.

Hartshorne set the position out in The Divine Relativity: A Social Conception of God, delivered as the Terry Lectures in 1947 and published by Yale in 1948. The name is chosen deliberately: he takes himself to be continuing the classical project rather than abandoning it, using classical arguments and classical reasoning about perfection while correcting what he regards as a mistake in the classical conclusion. That mistake, on his diagnosis, is the assumption that perfection excludes relativity. Hartshorne's claim is that a being unaffected by anything is not maximally perfect but maximally isolated, and that the truly perfect being is the one that surpasses all others while remaining able to surpass itself.

The view has real advantages. It gives an account of divine responsiveness that does not require pretending the biblical language is figurative, and it dissolves the impassibility dispute rather than settling it, since God is impassible in one aspect and passible in the other. It also inherits Whitehead's metaphysics from Process and Reality (1929), which gives it a worked-out framework rather than a set of adjustments.

The costs are equally real, and they are the costs the classical theist predicted. A God whose concrete states depend on the world is a God who depends on the world, which is precisely what the argument from independence in the previous rung ruled out. A God who is enriched by creation did not have to create out of pure generosity, since he had something to gain. And a God who is one pole of a relation to a world that he does not bring into existence from nothing is not the metaphysical ultimate; something has to explain why there is a God and a world in relation at all. Hartshorne's defenders answer each of these, but the pattern is worth noting: neo-classical theism buys its solution to the responsiveness problem with the currency of aseity, which is exactly the trade the classical tradition refused to make.

Rival specifications or different concepts?

The question worth asking at the end is what kind of disagreement this is. Two diagnoses are on offer and they lead to different conclusions about how the argument should be conducted.

On the first diagnosis, the two camps are giving rival specifications of a single concept. There is one thing both are trying to describe, the ultimate reality on which everything depends, and they disagree about what is true of it. If this is right, the dispute is substantive and the ordinary methods apply: work out which account is coherent, which fits the evidence, which does the explanatory job, and the better account wins. The arguments of the first three rungs are then arguments for a conclusion both camps can share, with the character of the conclusion to be filled in afterwards.

On the second diagnosis, they are not talking about the same thing. The classical theist's God and the personalist's God have different existence conditions, different individuation conditions, and different relations to the world, and the shared word is doing less work than it looks. If this is right, then a great deal of philosophy of religion has been conducted at cross purposes. The cosmological argument concludes to something classical; the argument from religious experience concludes to something personal; the problem of evil is formulated against a personalist God and lands very differently against a classical one, since a God who does not deliberate among options is not straightforwardly failing to choose the better one.

A third diagnosis is available and deserves more attention than it gets. Suppose reference is fixed not by a description but by a practice, in the way the causal accounts of reference from the philosophy of language propose. Then both camps refer to whatever the worshipping tradition has been tracking all along, and their competing descriptions are competing theories about one thing rather than definitions of two. This has the attraction of explaining why the disagreement feels substantive to the participants: they take themselves to be correcting each other, not changing the subject, and on this account they are right. It also predicts the actual sociology, in which philosophers move between the camps and regard the move as having learned something rather than as having converted. The cost is that reference-fixing by practice works cleanly for natural kinds, where samples are available to be wrong about, and it is not obvious what plays the role of the sample here. A defender can answer that religious experience plays it, which hands the question to the argument the intermediate tier treated and inherits its difficulties.

The equivocation argument is not decisive; the third premise can be resisted by anyone who thinks the classical denial of deliberation is compatible with a good enough analogue of it. But it explains something about the state of the field that is otherwise puzzling, which is why the two camps so rarely engage each other's best material. Advanced rungs 14 through 17 will formulate the evidential problem of evil with a precision that presupposes a God who chooses among options, and it is worth carrying forward the question of what those arguments show against a God who does not.

Next

One notion has been doing quiet work throughout this rung and has not yet been examined on its own. Independence from anything external, the property that generated simplicity, timelessness, and impassibility, and that neo-classical theism was willing to trade away, is the doctrine of aseity. The next rung takes it up directly and runs into a problem the classical tradition did not anticipate: if God accounts for everything other than himself, what accounts for numbers, properties, and propositions, and can God create the things his own nature seems to presuppose?