Philosophy of Religion Beyond Monotheism

Almost every argument in this subject has been aimed at one target: a single, personal, transcendent creator, distinct from the world, who knows everything, can do anything, and is perfectly good. That target is specific, and its specificity has been doing unnoticed work throughout. Several arguments succeed against it and fail against nearby positions; several fail against it and succeed elsewhere; and a few of the hardest problems in the subject are artefacts of the target rather than of religion.

This rung changes the target and watches what happens.

The world as the divine

The standing objection is that this is atheism with better manners. If God just is everything, then calling everything God adds no information, and the position is a relabelling.

Michael Levine's Pantheism of 1994 is the most sustained modern defence, and the reply it develops is that the position asserts something substantive that a naturalist need not accept: that the totality is a genuine unity rather than an aggregate, and that this unity has a value which grounds an appropriate response. Whether that is enough to distinguish the two remains contested, but the charge that pantheism is empty is too quick.

Watch what happens to the arguments.

The cosmological argument does not so much fail as change sides. Its conclusion is that there is a necessary being on which everything contingent depends, and the pantheist can accept the conclusion while identifying the necessary being with the totality itself. What blocks this for the theist is the assumption that the universe is contingent, and Spinoza denied exactly that. The price he paid is severe and he paid it openly: if the necessary being is the world, then everything in the world follows necessarily, and nothing could have been otherwise.

The design argument fails outright, since there is no designer standing outside the arrangement to have arranged it.

The problem of evil undergoes the largest change of any argument in the subject. It requires an agent who could have prevented the evil and did not, and pantheism supplies no such agent. There is nothing to hold responsible and therefore no theodicy to construct. That is the position's clearest advantage, and it is a considerable one given how much of this tier the problem consumed.

What pantheism loses is everything the problem of evil was protecting. There is no providence, no reason to think the arrangement is good, and nothing that could answer a prayer. The relief comes from having removed the being whose conduct was at issue.

A further question is what religious practice the position supports. Prayer of petition makes no sense, since there is no one to petition and nothing that could be persuaded. What remains is a set of attitudes: acceptance of what is necessary, a form of contemplation directed at the whole rather than at a part of it, and a reordering of what one takes to matter. Spinoza's own account of the good life ran along those lines and treated the reordering as the point of the system rather than an appendix to it.

Whether attitudes of that kind are appropriately called religious is partly a verbal question and partly not. What is not verbal is whether reverence can be owed to something that is not an agent, since reverence in its ordinary use is a response to a person. The pantheist has to hold either that the ordinary use is too narrow or that the totality has features that make a personal response fitting without its being a person. Both moves are available and neither is comfortable.

The world in the divine

The volume edited by Philip Clayton and Arthur Peacocke, In Whom We Live and Move and Have Our Being, published in 2004, collects the modern discussion and shows what motivates it. Two motivations dominate.

The first is divine action. On classical theism, an immaterial, timeless being acts on a causally closed physical world, and how it does so has never been explained. If the world is within God, the acting is no longer action across a gap.

The second is impassibility. The classical view that God is unaffected by what happens was examined earlier in this tier and found difficult to reconcile with the idea of a being who loves. Panentheism drops it by construction: what happens to the world happens to God.

Three costs are standard.

The containment relation is hard to state. Being spatially inside something is not what is meant, and being logically inside is not either. The claim tends to be presented through images and resists a literal formulation, which makes it correspondingly hard to evaluate.

The problem of evil returns and is arguably worse. A God who is intimately involved in every event, and who is affected by all of them, is not distanced from what happens by the classical apparatus, and the sceptical replies of rung 15 lose most of their force when the being in question is not radically other.

And the classical attributes go. A God who changes is not immutable, a God affected by the world is not impassible, and a God who does not unilaterally determine outcomes is not omnipotent in the traditional sense. Process theists regard this as the point rather than the price, but it means the arguments of the earlier tiers, which established their conclusions about a very different being, no longer transfer.

More than one

The design argument is the clearest case. From the observation that the world exhibits organisation of a kind produced elsewhere by agents, the inference is to an agent or agents. Nothing in the reasoning fixes the number at one. Hume's Philo pressed exactly this in the Dialogues Concerning Natural Religion of 1779, pointing out that large artefacts in our experience are made by many hands, so that the analogy on which the argument depends favours a committee.

The attribute paradoxes examined at rung 9 also loosen. The tensions there arose from loading every perfection onto one being, and they do not arise if power, knowledge, and goodness are distributed. A being that is very powerful but not omnipotent generates no paradox of the stone.

Two costs are usually decisive for philosophers and are worth stating precisely rather than assuming.

Parsimony tells against multiplying entities, though the principle is weaker here than it is usually made to sound: it counts against unnecessary posits, and if the phenomena genuinely support several agents, the several agents are not unnecessary.

More seriously, polytheism has no natural stopping point on the ultimacy question. If the gods are themselves beings within a larger reality, the question of what explains that reality is untouched, and the argument that motivated the enquiry has not been answered. Monotheism's real advantage is not simplicity for its own sake but that it terminates the regress the cosmological argument opens.

A creator who does not intervene

Deism matters for a reason that is easy to miss. It is roughly what the natural theology of the intermediate tier would establish if every one of those arguments succeeded.

A cosmological argument that works delivers a necessary being on which the world depends. A design argument that works delivers a designer with the power and knowledge to arrange things. A fine-tuning argument that works delivers a being interested in the existence of life. None of them delivers a being who intervenes in history, cares about individuals, hears prayer, or has revealed anything. The gap between what the arguments reach and what any tradition affirms is exactly the size of the difference between deism and theism, and it is a very large gap that the natural theology literature tends to leave for later without ever coming back to it.

Two problems press hard on deism specifically.

There is no explanation of the reticence. A creator who established the arrangement and then did nothing further is a peculiar being, and the position offers no account of why such a being would act once and not again.

And the hiddenness argument of rung 17 lands with unusual force. The deist has conceded that a creator exists and denied every mechanism by which anyone could come to know it. If the argument from nonresistant nonbelief works at all, it works best here.

Traditions without a creator

The existence of these traditions bears on the subject in two ways.

It shows that a great many of the questions occupying this ladder are local. The problem of evil, the attribute paradoxes, the compatibility of foreknowledge with freedom, and the hiddenness argument all require a creator with a will, and none of them arises for a tradition that has none. A field that presents these as the problems of the philosophy of religion has mistaken the problems of one family of traditions for the problems of the subject.

It also shows that other questions have been neglected. Whether there is a self of the kind that could persist, whether suffering has a diagnosis and a cure, what the relation is between an ultimate that is not an agent and the world of ordinary experience, and how a tradition can be authoritative without a revealer are all serious philosophical questions taken up in detail elsewhere and barely at all in the anglophone literature.

The treatment of suffering is the sharpest illustration. Where the traditions of this ladder ask why suffering is permitted, and spend centuries on the answer, a tradition without a creator asks instead what suffering is, what produces it, and what ends it. The change is from a question about justification to a question about diagnosis, and it eliminates the entire problem the last five rungs were occupied with while raising different difficulties of its own about whether the diagnosis is correct and the prescribed cure effective. The anti-theodicists of rung 16 were reaching for something in this vicinity when they argued that the demand for justification is itself the error, though they were arguing inside a tradition committed to a creator and so could not follow the move all the way.

What the positions have in common

Setting them beside one another suggests a better organising question than the one the field usually asks.

Every position surveyed here, including atheism in most of its forms, holds that something is ultimate: something on which everything else depends, or beyond which explanation does not go. They differ on two matters, and the two are independent.

The first is whether the ultimate is an agent. Classical theism, deism, and polytheism say yes; pantheism, most nontheistic traditions, and naturalism say no; panentheism and process theism occupy a contested middle where the ultimate has some agent-like features and not others.

The second is how the ultimate stands to the world. It may be identical with it, contain it, be separate from it, or be a feature of it.

Nearly every distinctively difficult problem in this subject arises from one particular pair of answers: an ultimate that is an agent and is separate from the world. Being an agent is what makes evil a problem, hiddenness a problem, and the coherence of the attributes a problem. Being separate is what makes divine action, divine knowledge of contingents, and creation problems. Positions that answer either question differently trade this set of difficulties for another set, and the trade is usually a good deal less favourable than it first appears.

What transfers and what does not

Collecting the results makes the pattern visible.

The cosmological argument is the most portable. Its conclusion is compatible with pantheism and with deism, and it is the only argument that gives monotheism a genuine structural advantage, since it alone requires something that terminates a regress.

The design and fine-tuning arguments are the least selective. They support deism as readily as theism and polytheism as readily as either.

The moral argument depends most heavily on the target, since it requires a being with a will who can stand in the right relation to obligation, and it fails for pantheism and for every nontheistic option.

The problem of evil is the most target-dependent of all. It is devastating against a being with the classical attributes, harder against a process deity, and simply inapplicable where there is no agent.

And the arguments from religious experience are the most awkward, since the experiences reported across traditions support different targets, which is where this rung and the last one meet.

Next

That completes the theism ladder. The beginner tier fixed the vocabulary, the intermediate tier set out the arguments on both sides, and this tier took each of them apart: the coherence of the attributes, the epistemology of belief without argument, the strongest forms of the problem of evil, the value of the whole enterprise, and finally the question of what the arguments were arguing about.

Two things are worth carrying away. The first is that almost nothing here was settled, and that this is a feature of the material rather than a failure of the treatment; the arguments that survive contact are the ones worth knowing. The second is that most of what made these arguments hard belongs to other parts of philosophy. Whether a causal series can be infinite is a question in metaphysics. Whether a properly functioning faculty confers warrant is a question in epistemology. Whether obligation requires a will is a question in ethics. Those ladders are where the tools used here were made, and they are the natural place to go next.