Miracles and Reason: Is the Supernatural Intellectually Respectable?

The word "miracle" carries cultural baggage that makes philosophical discussion difficult. For many secular people, it conjures images of credulous crowds, faith healers, and the wholesale rejection of critical thinking. For many religious people, miracles are articles of faith not subject to rational scrutiny. Both reactions tend to terminate the conversation before it begins.

But there is a serious philosophical question here that deserves serious treatment: Is a miracle — understood as an event that cannot be explained by natural causes alone — the kind of thing a rational person can believe in? Or does such belief necessarily involve suspending reason in favor of faith?

Hume's Argument, Stated Fairly

David Hume's argument against miracles, presented in An Enquiry Concerning Human Understanding (1748), is the most influential in the Western philosophical tradition. It is often misrepresented as simply "miracles violate natural law, therefore they can't happen." That is not quite Hume's argument.

Hume's argument is about evidence and probability. It runs roughly as follows:

Our belief in any matter of fact is justified by experience. The laws of nature are established by the uniform and repeated experience of all people throughout all time — they represent the most reliable empirical evidence available to us. A miracle, by definition, is a violation of these laws — an event that contradicts the entire weight of human experience. Therefore, when someone claims a miracle has occurred, we face a competition between two bodies of evidence: the testimony of the witnesses, and the entire accumulated experience of humanity regarding natural regularities.

Hume's conclusion: the evidence for natural law will always outweigh the evidence for any particular miracle claim. A wise person proportions belief to evidence, and the evidence always favors the natural explanation.

Three Problems with Hume's Argument

The argument has rhetorical force, but it has serious philosophical difficulties.

First problem: the argument is circular. Hume's argument establishes that miracles are improbable by defining them as violations of laws established by uniform experience. But this assumes that the scope of "natural" causes is already fully defined — that there are no causally efficacious entities outside the natural system. If theism is true, and there is an entity whose causal power is not part of the natural order, then invoking "the uniformity of natural law" as evidence against that entity's actions presupposes the falsity of theism. The argument assumes what it is trying to prove.

In other words: for a person who already accepts that the natural world is causally closed, Hume's argument works. But for a person genuinely asking whether a causal agent outside nature might exist and act, Hume's argument is unavailable without first settling the theism question — which is the harder question.

Second problem: Hume's framing mischaracterizes miracles. Hume frames miracles as "violations" of natural law, but this is not the most sophisticated theistic account of what a miracle is. C.S. Lewis, responding to Hume, argued that natural law describes the regular behavior of nature when nothing external to the natural system intervenes. A miracle is not a violation of natural law; it is an event caused by something outside the natural system. Natural laws still operate; they are simply not the only things operating.

An analogy: the laws of physics are not "violated" when a human being picks up a stone and throws it. The stone does not fall under gravity because an external agent redirected its motion. The laws of physics still apply — they describe the trajectory once the stone is thrown. Similarly, if God acts in the world, natural laws do not break down; they simply are not the complete description of what caused the event.

Third problem: Hume's standard is question-beggingly high. Hume says a "wise person" proportions belief to evidence. But how much evidence is required for a wise person to accept a miracle claim? If the standard is that no testimony can ever be sufficient — that a miracle claim will always be outweighed by prior probability — then the standard is not proportioning to evidence; it is ruling out the conclusion in advance, regardless of evidence. This is not rationality; it is dogmatism in empiricist clothing.

What Kind of Evidence Would Suffice?

If we reject Hume's dismissal, we need a more careful framework. What would constitute good evidence for a miraculous event?

The relevant considerations are familiar from ordinary historical and forensic reasoning. The quality and independence of the witnesses. Whether those witnesses had anything to gain from the claim. Whether the testimony is consistent across multiple sources. Whether the event was recorded close to the time of occurrence. Whether those who were there and had strong motivation to disprove it failed to do so.

Applied to Islamic tradition, the Quran itself presents a specific evidential challenge: that those who doubted its divine origin could produce something comparable in literary and persuasive terms, and that this challenge remained unmet by those who had every incentive to meet it. This is a different kind of miracle claim — not a suspension of physical law but a claim about communicative power and its inexplicable source.

Applied to historical miracle claims more broadly, the question is not "can we rule out in advance that anything unusual happened?" but "given the specific evidence available, what is the most probable explanation?"

Miracles and Their Function

There is also a question about what miracles are supposed to accomplish, which is separate from the question of whether they occur.

In the Islamic tradition, miracles (mu'jizat) are associated with prophets as signs confirming their prophetic status. They are not demonstrations of power for its own sake. They are, in the terminology of contemporary philosophy of religion, "calibration events" — events designed to allow a rational agent to update their probability estimate regarding the truth of a particular prophetic claim.

This functional account suggests that miracles, if they occur, are not arbitrary violations of natural order. They are communicative acts — and like all communicative acts, they can be evaluated for their purpose, their context, and the coherence of the message they carry.

The question of whether the calibration worked — whether the signs accompanying a prophetic claim are sufficient to make the claim credible — is an empirical question, not a question to be settled by philosophical argument alone. Different people, examining the same evidence, will reach different conclusions, just as they do with other historical claims. What philosophical analysis can do is clear away the bad arguments — including the argument that miracle claims are ruled out before the evidence is examined.

Where This Leaves Us

The default skeptical position on miracles — "we know these things don't happen, so reports of them must be explained otherwise" — is not the rationalist position it presents itself as. It smuggles in a strong metaphysical assumption (natural causal closure) and presents it as a conclusion of empirical reasoning.

The intellectually honest position is more open-ended: we have good evidence that natural regularities are reliable guides to what usually happens. We do not have good evidence that nothing outside the natural system exists. Claims of unusual events that might indicate the operation of such a causal agent need to be evaluated on their specific merits — the quality of testimony, the coherence of the claim, the plausibility of the alternatives.

That evaluation may not result in accepting any particular miracle claim. But it cannot legitimately result in dismissing all such claims before the evidence is examined.

---

Questions worth sitting with:

• Is there a principled difference between "Hume's standard of evidence is high but fair" and "Hume's standard rules out miracle claims regardless of evidence"? Where would you draw that line?
• If you would accept a miracle claim given sufficient evidence, what would "sufficient evidence" look like in practice?
• Does the function of miracles — as communicative acts addressed to rational agents — change how you think about evaluating them, compared to treating them as anomalous physical events?