Mackie's The Miracle of Theism: The Case Against God

Imagine two people looking at the same night sky. One sees evidence of a creator in its order. The other sees a physical universe whose regularities still need explaining. Neither observation settles their disagreement. They need to ask what each explanation adds, which assumptions it requires, and what would count against it.

That is the task J. L. Mackie undertakes in The Miracle of Theism. His conclusion is atheist: he judges the overall case against the existence of God substantially stronger than the case for it. Yet the route to that conclusion matters as much as the verdict. A failed proof need not be worthless evidence, several modest arguments might reinforce one another, and a possible explanation need not be a probable one. Losing these distinctions makes both belief and unbelief too easy.

The title has a Humean irony. In his introduction, Mackie distinguishes a supernatural intervention from the older sense of a miracle as something surprising. His target is the continuing conviction of reasonable believers, despite what he regards as inadequate support. He also thinks religious belief can have a natural history. Its persistence therefore need not itself involve supernatural intervention. The title expresses his judgment about the arguments, not an extra proof against God. Mackie, Introduction, pp. 11–12.

Mackie begins with what he calls traditional theism: a personal creator and sustainer of the universe, unlimited in power and knowledge, and wholly good. This is a more definite claim than saying that life contains mystery. Someone might accept an impersonal ultimate reality while rejecting this personal God. An objection to one conception does not automatically defeat every religious possibility.

The book appeared in 1982. Its preface identifies Richard Swinburne's The Existence of God and Hans Küng's Does God Exist? as important opposing cases. Mackie also discusses older arguments and Plantinga's contemporary use of possible worlds. This combination explains the book's scope: it tests both familiar proofs and attempts to reconstruct them. Preface and contents.

Mackie rejects a shortcut that would declare God-talk meaningless because it cannot be directly verified. We can understand a claim even when establishing its truth is difficult. His introduction argues that a strict verificationist account also creates problems for ordinary statements about the past and other people's minds. Rejecting that account lets him engage theism as a truth claim. The believer receives a hearing; the claim still owes us reasons.

For the reader, that distinction is useful before any disagreement begins. Ask someone what they mean by God, then ask what supports that conception. If one speaker means a loving agent and another means the existence of an intelligible universe, an apparent agreement may hide different claims. Mackie's method requires us to keep the conclusion in view throughout the argument.

A deductive proof aims to make its conclusion unavoidable if its premises are true. A nondeductive argument supports a conclusion without guaranteeing it. Finding a wet pavement may support the hypothesis of rain even though a sprinkler could explain it. That alternative prevents a proof; it does not make the observation irrelevant.

Mackie expressly allows that individually modest arguments can have a cumulative effect. His introduction uses the comparison with a legal case: several considerations may together justify a conclusion that none establishes alone. Chapter 14 returns to their combined force. It is therefore inaccurate to say that he rejects theism simply because no isolated argument delivers certainty. His question is whether the whole body of evidence favors the divine explanation over its alternatives. Introduction, pp. 4–7; chapter 14(b).

Consider an ordinary example. A colleague repeatedly misses meetings. Illness could explain every absence. So could a calendar error. Before choosing, check what each hypothesis predicts: perhaps the colleague is answering messages from bed, or perhaps every missed meeting used an outdated invitation. Merely adding an explanation that can fit any outcome gives little help in deciding between them.

The analogy has limits. A creator is not a colleague, and a universe is not a meeting schedule. It nevertheless makes the reasoning problem visible: explaining why an observation would be unsurprising if a hypothesis were true does not, by itself, establish that hypothesis. Mackie's disagreement with theists often concerns that second step.

Ontological arguments start from a conception of God rather than from observations of the world. Mackie's chapter 3 treats several versions, so its objections should not be reduced to a single slogan about existence. A definition can tell us what would qualify as a certain kind of being; the further claim that such a being exists needs defense.

Plantinga's modal version sharpens the issue. In the framework Mackie considers, a maximally great being would possess the relevant excellence in every possible world. These worlds represent ways things could have been, not planets elsewhere in space. Given that framework, the possibility of such a being carries a strong consequence: its existence cannot merely be contingent.

Mackie grants the inference for the purpose of his main objection. He asks why someone not already persuaded of theism should accept the premise that maximal greatness is possibly instantiated. Saying that we have not noticed a contradiction does not settle every question about metaphysical possibility. His criticism concerns the premise's independent justification, rather than dismissing modal reasoning as meaningless. Chapter 3(c), especially pp. 57–61.

This is a useful reading habit for any elaborate argument. Formal precision can show exactly what follows from an assumption. It cannot by itself supply a reason to accept that assumption. When an impressive conclusion depends on one controversial premise, spend your attention there.

Cosmological arguments begin with the existence or general features of the world. Mackie's chapter 5 distinguishes arguments from contingency and sufficient reason, causal regress, a finite past, and Swinburne's inductive approach. Treating them all as the claim that everything has a cause would miss their differences. A defender may instead say that contingent things require explanation while a necessary being does not.

The dispute then concerns what necessity means and whether it supplies the explanation promised. Calling something necessary is not enough. The argument must establish the relevant necessity, connect it to the world, and show why it belongs to a divine being. Mackie also questions whether the demand for a sufficient reason for everything is justified. He allows the naturalist to face an unexplained world, while asking whether the theist has improved matters by postulating God. Chapter 5(a); chapter 14(b).

Design arguments add another question: why is the world structured in ways that permit organized life and inquiry? A biological explanation does not automatically answer why there are laws of nature at all. Mackie's chapter 8 therefore considers Swinburne's argument from temporal order, the regular succession of events, as well as familiar arguments from organized things within the world.

Swinburne regards a divine personal explanation as simple. Mackie disputes this. A mind that brings about a universe directly by intending it includes a substantial explanatory commitment. The familiar cases of intention to which Mackie appeals involve embodied agents and causal means. Mackie questions the move from those cases to an unlimited disembodied will. The disagreement concerns the explanatory credentials of the proposal, not a simple competition to count objects. Chapter 8(c), pp. 146–149.

There is a serious response available to the theist. A single agent with unlimited powers may be simpler in a relevant sense than many independent unexplained facts. Swinburne develops this account in his own work on simplicity and personal explanation. Mackie thinks the added kind of agency costs more than it explains. The reader must assess that disagreement; repeating either the word simple or the question of who created God will not resolve it. Swinburne, “God as the Simplest Explanation of the Universe”.

Mackie's moral discussion requires particular care. He does not defend objective, intrinsically prescriptive moral values and then try to preserve them without God. In chapter 6(e), he conditionally grants that values of that unusual kind, if they existed, could give inductive support to a divine explanation. He then disputes the premise that we have such values to explain. This is different from denying that people care about cruelty, make moral judgments, or cooperate.

He also distinguishes creating obligation by command from creating objective value. A command cannot be the ultimate source of every obligation if obeying it already presupposes an obligation to obey its issuer. His conditional concession about value therefore does not simply revive a divine-command theory. Chapter 6(e), pp. 114–118.

Imagine a manager trying to justify a rule. “Because I ordered it” identifies an authority's decision. It leaves open why that authority should be obeyed, what limits it has, and whether the rule is good. This contemporary example does not settle the theology. It helps separate a command's existence from its justification, which is the distinction the argument needs.

Chapter 10 similarly separates religious experiences from the conclusions drawn from them. Someone can undergo a vivid experience while misidentifying its source. Recognizing the experience's reality for the person does not establish that a divine being caused it. Conversely, proposing a psychological explanation does not establish that every religious interpretation is false. The dispute concerns the evidence for moving from a reported state of consciousness to an external supernatural reality.

Mackie's natural histories of religion serve a related but limited purpose. A belief's natural origin does not logically entail its falsity. An account of why people would hold a belief even without its truth can, however, weaken an appeal to its popularity. That distinction protects the discussion from an easy genetic fallacy: dismissing a claim solely because of how someone acquired it. Introduction, p. 7; chapter 10.

The problem of evil asks how a wholly good, all-powerful God can coexist with evil. Mackie says in chapter 9(a) that there is no explicit contradiction in those two claims alone. A contradiction requires connecting principles about what goodness would eliminate and what omnipotence makes possible. Those principles need examination too.

One answer appeals to goods that logically involve some evil. Courage in the face of danger, for example, cannot occur without danger. Mackie calls an evil absorbed when it is incorporated into a larger good that justifies it in the proposed way. He concedes that this possibility prevents the bare existence of evil from establishing an immediate inconsistency. He then asks whether all actual evils are absorbed. The existence of kindness in response to one injury does not establish that every injury is justified.

Plantinga distinguishes logically possible worlds from worlds God could actualize while leaving creatures significantly free. The possibility that every creaturely essence suffers from transworld depravity is meant to block the assumption that God could ensure morally significant freedom without any moral wrongdoing. This is a compatibility defense, not a claim to have discovered God's actual explanation for suffering. Plantinga, The Nature of Necessity, chapter IX.

Mackie accepts part of the restriction on what worlds God could actualize, but challenges the further limitation on the possible people God could create. He maintains that people who always freely choose well are possible and asks why an omnipotent creator must face an independently restricted selection. That is his objection. It remains open to a defender of Plantinga to question whether Mackie has established the required inference from possibility to divine actualizability.

At the end of chapter 9, Mackie allows that some adjustment of traditional doctrines might avoid contradiction, while judging that no satisfactory account has been presented. His conclusion is therefore more careful than a simple declaration of logical impossibility. It is also more critical of free-will defenses than the claim that he accepts Plantinga wholesale. Chapter 9(a), pp. 150–155; chapter 9(f), pp. 172–176.

The difference between logical and evidential questions now becomes clear. A logical argument claims that the propositions cannot all be true. A compatibility defense needs a coherent possibility that defeats that claim. An evidential argument asks instead how well a hypothesis fits what we encounter. The answer cannot be settled merely by naming a possibility. Equally, describing suffering as unexplained does not by itself establish that no morally sufficient explanation exists. Each side has a further argumentative task.

This distinction also sets a limit on practical applications. A person may gain patience or compassion after a terrible loss. That fact would not, on its own, show that the loss was necessary for that person's development, or that anyone was justified in allowing it. We should be careful when turning another person's suffering into an example of a greater good. The philosophical question asks whether a proposed justification works; offering consolation requires attention to the person who suffered. These are different activities, and a familiar theodicy need not be an appropriate response to grief.

The strongest response to Mackie can accept his demand for reasons while contesting his assessment of them. A theist can argue that the world's existence, regularity, consciousness, and religious experience jointly support a personal explanation. Mackie explicitly considers this cumulative strategy. He concludes that the divine hypothesis introduces greater difficulties than its naturalistic competitor. The disagreement is over comparative explanatory weight, not whether accumulation is allowed.

Another response questions our ability to judge which evils a good God could permit. A limited observer might fail to understand a morally sufficient reason. That possibility prevents us from treating an unexplained case as an automatic disproof. Yet it creates a further question: how much uncertainty about divine purposes is compatible with confidently interpreting favorable features of the world as evidence of divine goodness? This is a question for assessing the arguments together, rather than a quotation from Mackie.

For personal inquiry, separate the claims that an argument fails, that belief is unsupported, and that its negation is supported. They require different work. Mackie's overall atheist conclusion depends on his positive comparison of explanations and the additional pressure of evil. It does not follow from finding one flaw in one proof.

The same discipline helps in work. Suppose a team proposes an opaque system because it gives a plausible account of past outcomes. Ask which alternatives were compared, which assumptions were added, and what evidence would change the decision. This is an application of explanatory comparison, not a claim that Mackie anticipated contemporary AI. His book rewards the reader willing to reconstruct an opponent's best argument and then identify the exact point where support runs out.

• J. L. Mackie, The Miracle of Theism: Arguments for and against the Existence of God: read the introduction with chapter 14(b), then compare the two discussions of evil in chapter 9(a) and 9(f).
• Alvin Plantinga, The Nature of Necessity, chapter IX: the modal free-will defense Mackie disputes.
• Richard Swinburne, “God as the Simplest Explanation of the Universe”: a theistic account of the simplicity and explanatory role at issue.
• J. L. Mackie, Ethics: Inventing Right and Wrong, chapter 1, sections 1 and 9: background to his rejection of intrinsically prescriptive objective values, a position that should be distinguished from indifference to human welfare.