Hume's Miracle Prison: How They Got Out Alive

One of the great geniuses of the Enlightenment was David Hume (1711-1776). In his essay on miracles (Section X) in his book Enquiries concerning Human Understanding, he doubts that miracles have ever occurred and even can occur.

Hume's short analysis shakes believers (for our purposes those who believe in miracles) and theists (for our purposes those who believe that God exists and acts in his creation). But skeptics (those who say that miracles do not or even cannot happen) gladly accept Hume's verdict.

However, do Hume's arguments stand up under close scrutiny? Can we exclude miracles as a priori impossible (a priori means before investigation)? That is, should we reject miracles outright? But what if miracles can be investigated with modern technology? Do they happen today?

This article has a modest goal. It aims to keep the door to miracles open, even after Hume's assault on them. It uses the metaphor of prison and the legal system to illustrate their escape into the modern skeptical world. They seem to be in hiding, but sometimes they peek out and show themselves today.

What is a miracle prison?

In Part II of Hume's discussion of miracles in Section X, he is in the process of answering the question of whether miracles do or even can occur. Is there even one criterion that any reasonable person can use to affirm their occurrence? Apparently not, for two major reasons. Bulleted examples follow each reason.

(1) No witness for the defense is reliable enough.
  • No man, not even many, can be of "such unquestioned good sense, education, and learning" or can be of "such undoubted integrity" or of "such credit and reputation in the eyes of mankind" that they can convince skeptics that the witnesses for miracles are not deluded or do not have "any design to deceive others" or are not exempt from self-interest and shame of being detected in promoting miracles (pp. 116-17).
  • When the spirit of religion joins itself to the love of miracles, therefore, people of religion naturally or have a propensity to believe such things (p. 117).
  • "A religionist may be an enthusiast, and imagine he sees what has no reality" (p. 117).
  • An eloquent speaker may manipulate the masses (p. 118).
  • Barbarous peoples in an unenlightened and bygone age produce the reports of miracles. But as we enter the civilizations of more enlightened peoples in recent times, the miracles, not surprisingly, dry up (pp. 119-21).
(2) No testimony is strong enough.
  • Countless testimonies of forged miracles produce suspicion against all miracles; they also demonstrate how gullible people are (pp. 118-19; 125-27)
  • Hume recounts "one of the best attested miracles in all profane [secular] history." The Emperor Vespasian healed a blind man and a lame man, as reported by Tacitus, whom Hume praises as reliable. But after all the confirmation of the miracles, "no evidence can well be supposed stronger for so gross and palpable falsehood" (pp. 122-23).
  • Testimonies about the Cardinal de Retz also fall into the same categorical doubt. He seems to have witnessed the result of a miracle, but later thought better of it because even well-attested, strong evidence "carried falsehood upon the very face of it, and that a miracle, supported by human testimony, was more properly a subject of derision than of argument" (pp. 123-24).
  • At the tomb of the Abbé of Paris alleged miracles were produced and confirmed by "judges of unquestioned integrity and attested by witnesses of credit and distinction, in a learned age." Hume goes on to build a theoretically strong case for the miracles, but none of this is sufficient or even supportive of belief in them (pp. 124-25).
  • Hume offers a hypothetical. Let us imagine that Queen Elizabeth died on January 1, 1600. "All historians" (Hume's words) who specialize in English history agree on her time of death. She was seen by her physicians and courtiers before and after her death. She was interred (buried) a month. But then, lo! "She again appeared, resumed her throne, and governed England for three years." Would Hume believe this testimony of her death and then reappearance? Not in the slightest. (p. 128)
Why is Hume so skeptical? Witnesses for the defense and their testimonies are not good or strong enough, but compared to what? Simple. All miracles violate our firm and unalterable experience that establishes the laws of nature.
A miracle is a violation of the laws of nature; and as firm and unalterable experience has established these laws, the proof against a miracle, from the very nature of the fact, is as entire as any argument from experience can possibly be imagined. (p. 114)

A miracle may be accurately defined [as] a transgression of a law of nature by a particular volition of the Deity, or by the interposition of some invisible agent. (p. 115, note 1, emphasis original)
Thus, no person, even if he has the utmost integrity and honesty, can overturn by his testimony the laws of nature established by firm and unalterable experience.

Hume uses, as it were, a two-sided scale, like the scales of justice on the outside of the Supreme Court building. On one side he places our firm and unalterable experience with the laws of nature; on the other he places the reliable testimony for miracles. The first side is always heavier or wins the contest. This is why he could establish the witnesses and testimonies (in the bulleted lists) with such confidence, proclaiming their veracity. But firm and unalterable experience establishing the laws of nature must by the very nature of the case always outweigh the testimonial evidence for miracles. "A wise man, therefore, proportions his belief to the evidence" (p. 110).
[And it is a general maxim (principle) worthy of our attention] that "no testimony is sufficient to establish a miracle, unless the testimony be of such a kind that its falsehood would be more miraculous than the fact which it endeavors to establish". . . (pp. 115-16)

But what have we to oppose to such a cloud of witnesses [to the miracles at the tomb of the Abbé of Paris], but the absolute impossibility or miraculous nature of the events, which they relate? And this surely, in the eyes of all reasonable people, will alone be regarded as sufficient refutation. (p. 125, emphasis added)

Upon the whole, then, it appears that no testimony for any kind of miracle has ever amounted to a probability, much less a proof. (p. 127)

And therefore we may establish it as a maxim [principle] that no human testimony can have such a force as to prove a miracle, and make it a just foundation for any such system of religion. (p. 127)
As for the hypothetical death and reappearance of Queen Elizabeth in 1600 (she actually died in 1603), Hume would still not believe the testimonies, despite the events being observed by learned and trusted men who testify to her death and burial for one month:
All this might astonish me; but I would still reply that the knavery and folly of men are such common phenomena that I should rather believe the most extraordinary events to arise from their concurrence than admit of so signal a violation of the laws of nature. (p. 128)
Apparently, the (hypothetical) learned and wise historians, physicians and courtiers, who witnessed everything with their own eyes (Elizabeth's death, burial, and reappearance), turn into fools and knaves, according to Hume. What a "miraculous" reversal for such reliable and impeccable witnesses. Apparently, their shocking, quick-change falsehood is not "more miraculous than the fact which it endeavors to establish."

Thus, miracles are locked up in Hume's prison, though they are innocent. Under his ironclad presuppositions, they cannot get out even on parole for good behavior. This is unjust.

See Craig, pp. 130-32, for the idea of the scale that Hume seems to have in mind.

No way out?

However, Hume's stacked deck against miracles begs the question or goes in circles. This fallacy means that the answer to a question is found in the premises or in the front end of the investigation. We assume the answer before we inquire into it. His super-high definition of a miracle does this (it violates the laws of nature established by firm and unalterable experience). He is trying to determine whether miracles can occur, but he slams shut the prison doors on them before they can make their appeal, not to mention while they were on trial.

The words firm and especially unalterable are the crux of the fallacy. How do we know that miracles cannot occur? Because they "violate" or "transgress" the laws of nature that are established by firm and unalterable experience. But why cannot our experience with the laws of nature be "violated" on occasion? Because that would be a miracle. And they don't happen because of our firm and unalterable experience establishes the laws of nature.

Next, Hume's definition of a miracle is so stringent that no historical or empirical investigation will possibly argue the case for miracles. To repeat the circular argument, why are no multiple honest and reliable testimonies in favor of miracles acceptable? Because the laws of nature are firmly and unalterably established by experience. The testimonies are ipso facto less accurate and less probable, no matter what. Therefore, no testimonies whatsoever for the defense will open the prison doors, because they are permanently locked in advance, no key existing to open them.

C. S. Lewis describes the circularity:
Now we must agree with Hume that if there is absolutely "uniform experience" against miracles, if in other words they have never happened, why then they never have. Unfortunately we know the experience against them to be uniform only if we know that all the reports of them are false. And we can know all the reports to be false only if we know already that miracles have never occurred. In fact we are arguing in a circle. (pp. 132-33)
However, miracles are matters of perception and therefore investigable. They exist, if they do, in the realm of matters of fact. To laugh out of court all exonerating testimonies, regardless of how strong and reliable they are, is prejudicial. Miracles do not get a fair hearing; they never did get a fair trial. They got a bum rap in Hume's court, and now in his prison.

The circular reasoning keeps going round and round, in favor of skeptics.

Minimum security prison?

It is possible that scholars would suggest that I am not being fair to Hume here. It is not clear that he is begging the question. If he simply claims that since the evidence for the laws of nature is always greater than the evidence for a miracle, and since the two bodies of evidence inevitably conflict, then we are never justified in accepting reports of miracles.

Norman L. Geisler refers to Hume's "softer" argument that focuses on the (un)believability of reports about miracles, not their (im)possibility. The evidence for the rare event-or singular event-is weaker than the evidence for the regular event, so the wise person believes in the regular event ("Miracles and the Modern Mind," p. 75). We also cite the argument by Benjamin F. Armstrong to defuse the accusation of unfairness to Hume. To continue our prison motif, miracles are in a minimum security prison, but they still cannot escape.

(1) Hume speaks of "uniform experience" against miracles. Thus, he either still begs the question or he engages in special pleading. Geisler writes:
It begs the question if Hume presumes to know the whole field of experience to be uniform in advance of looking at the evidence for uniformity. For how can one know that all possible experience will confirm naturalism [which says nature is all there is], unless one has access to all possible experiences, including those in the future? (p. 76).
Then Geisler explains how Hume engages in special pleading, which is a fallacy that ignores unfavorable evidence. But Geisler counters: "If, on the other hand, Hume simply means by ‘uniform' experience the select experiences of some persons . . . then this is special pleading. For there are others who claim to have experienced miracles" (p. 76).

In addition, after explaining modern Humeans and then quoting Hume himself ("No means of detection remain save those which must be drawn from the very testimony itself of the reporters"), Armstrong says:
Hume's argument, then, appears to be either question-begging or superfluous. If the laws of nature are needed to rule out (past) miracle reports, then these laws may not be used, for nothing will have rendered the reports "non-data" with respect to the laws. If the laws may be used to rule out the reported events, then it will only be because something else has already ruled out these events. The verdict against Flew [a modern Humean who seeks to strengthen Hume's arguments] (and others) is that it is question-begging. ("Hume on Miracles," p. 327)
Armstrong continues:
...The fact that we/some use various nomologicals [law-like generalizations often of nature], whether drawn from science, conventional wisdom, or elsewhere, does not preclude an investigation of these nomologicals. Such an investigation of the nomologicals that Hume would wield against resurrections in no way challenges our ordinary practices. Rather such an investigation may simply show us the limits of what can be provided by our ordinary practices. The limits on our practices may be such that there simply is no particular stock of nomologicals that has epistemic primacy to rule out resurrections in the way attempted by the Humean argument. (p. 327)
In short, Armstrong says in these two excerpts that to investigate reliable reports on miracles fairly, which (allegedly) take place in the realm of perceptions, no one can escape the charge of begging the question, if one uses the laws to preclude those reliably reported miracles. After our investigation, it may be the case that "there is no particular stock of nomologicals" that completely enjoy "epistemic primacy" to rule out miracles or resurrections, to use Armstrong's example. (Epistemic pertains to knowledge or knowing.)

Armstrong is right, for the key is to investigate and then to formulate laws accordingly. Can we say absolutely that a miracle cannot happen? Sometimes reports, especially in the Age of Science, confirmed by CT scans and even videos, may be so reliable that to shut them out leads to prejudice and obscurantism. How much evidence would it take if a CT scan and the oncologist's own eyes detected cancer, but immediately after prayer the cancer vanished? It is absurd to rule something out that actually happened, no matter how rare.

(2) Geisler says that Hume adds up the evidence against miracles, instead of weighing it in favor of them. But Geisler disagrees: "Rational beliefs should not, however, be determined by majority vote. Hume seems to commit a kind of consensus gentium fallacy, an informal fallacy arguing that something should be believed to be true simply because it is believed by most people." (p. 79). Further, sometimes the exceedingly rare event happens. For example, a perfect bridge hand has been dealt, though the odds against it are 1,635,013,559,600 to 1.
Sometimes the "odds" against an event are high (based on past observation), but the evidence for the event is otherwise very good (based on current observation or reliable testimony). Hume's argument confuses quantity of evidence with the quality of evidence. Evidence must be weighed, not added. (p. 79, emphasis original)
(3) Hume proves too much. If a miracle really happens, then should we disbelieve it, regardless of whether the evidence is overwhelming? Geisler explains:
For [Hume's] argument does not hold that miracles have not occurred but only that we should not believe they have occurred simply because the evidence for the regular is always greater than that for the rare. But on this logic, if a miracle did occur-rare as it may be-one should still not believe it. It is patently absurd, however, to claim that an event should be disbelieved, even if it has occurred, that is, when the evidence is overwhelming that the purported miracle has occurred. (p. 80)
Geisler is right, as noted in our analysis of Armstrong's argument (see the first point in this section). The key is to investigate them without prejudice, in case the evidence shows overwhelmingly that an extremely rare event has indeed happened. Geisler goes on to relate two more arguments against Hume's softer version (or minimum security prison), which seemed at first glance to have avoided the fallacy of begging the question, but these three suffice for now. Hume does not avoid that fallacy, and he commits others.

Believers would not want to relinquish the regularity of the laws of nature. Miracles are much rarer than strictly rare natural events, even though anomalies happen. But skeptics seem to wield, even with their "softer" arguments, the laws against miracles as if the laws are the judge, prosecutor, jury, and executioner. That is unfair. So we reach the same conclusion as noted before and reaffirmed later in this article: miracles may go free after they are investigated and put on trial; their reality must be a live option. If not, then this begs the question always in favor of naturalism.

Five ways out?

Maybe Hume, surprisingly, leaves five small ways out for miracles from their false imprisonment. Maybe this allows him to escape from the accusation of begging the question or circular reasoning. Again, here are his definitions of miracles.
A miracle is a violation of the laws of nature; and as firm and unalterable experience has established these laws, the proof against a miracle, from the very nature of the fact, is as entire as any argument from experience can possibly be imagined.

A miracle may be accurately defined [as] a transgression of a law of nature by a particular volition of the Deity, or by the interposition of some invisible agent.
(1) In the first definition, Hume says that our firm and unalterable experience has established the laws of nature. It seems that Hume returns to his theory about the foundation of human knowledge concerning matters of fact (e.g. the sun rises; salt dissolves in water), as opposed to relations of ideas (e.g. proofs in geometry). The foundation of human knowledge about matters of fact is experience with cause and effect, he says (e.g. speaking or talking produces [causes] sound [effect]). And the foundation of this is the accumulation of many experiences with cause and effect. And the foundation of this is mere custom or habit (Hume, pp. 25-47).

If our experience is built on such a weak foundation as custom or habit, can our experience rule out miracles altogether? Granted, miracles may be rare, but impossible? How can any court claim in advance that they are impossible when the court is investigating whether they may occur? Therefore, to investigate miracles, their reality must be a live option and a real possibility, not a fake one. However, it seems that Hume wants things both ways. Our knowledge about matters of fact is (a little) unstable. But when it comes to miracles, which are in this same empirical realm, our experience militates against them because it is unalterable. A little unstable or unalterable. Which is it? There seems at the very least to be an inconsistency. Geisler agrees that Hume is not being consistent with his own epistemology (how we acquire and define knowledge).
[Hume] himself recognized the fallacy of this kind of reasoning [that the past always determines or even resembles the future] when he argued that based on past conformity, nothing can be known with certainty about the future. We cannot even know for sure that the sun will rise tomorrow morning. Hence, for Hume to deny future miracles based on past experience is inconsistent with his own principles . . . (p. 80)
If this analysis is true, then it leaves the prison doors open to miracles.

Also see Robert A. Larmer, Water into Wine? pp. 36-37.

(2) In the second definition Hume assumes the existence of God. If he works a miracle, then the terms "violation" or "transgression" of the laws of nature are wrong. By analogy, if the prison warden allows a concert in chapel, then he commits no violation or transgression of the rules because he permits it within the limits of his own authority. But if a lone guard does this for his own purposes and without permission, then this would be a violation or transgression because he does not act as a rightful authority. The existence of God lifts the analogy beyond the human level. More than a warden, God does not violate or transgress anything of his creation when miracles occur, because he is the final authority over it.

See Kreeft and Tacelli, pp. 111-12, who use the metaphor of a high school principal and a gym teacher.

Lewis writes wisely about how nature naturalizes the immigrant or miracle, so it is not a violator, but a welcome guest. The regularity of nature says, if A (cause), then B (effect). But a miracle introduces a new cause and effect: if A2, then B2, and the new situation conforms to all the laws.
It is therefore a mistake to define a miracle as something that breaks the laws of nature. It doesn't . . . If God annihilates or creates or deflects a unit of matter, He has created a new situation at that point. Immediately Nature domiciles this new situation, makes it at home in her realm, adapts all other events to it. It finds itself conforming to all the laws. If God creates a miraculous spermatozoon in the body of a virgin, it does not proceed to break any laws. The laws at once take over. Nature is ready. Pregnancy follows, according to all the normal laws, and nine months later a child is born . . . The moment [the newcomer, e.g. miracle] enters [Nature's] realm, it obeys her laws. Miraculous wine will intoxicate, miraculous conception will lead to pregnancy, inspired books will suffer all the ordinary processes of textual corruption, miraculous bread will be digested. The divine art of miracle is not an art of suspending the pattern to which events conform, but of feeding new events into that pattern. It does not violate the law's proviso, "If A, then B": it says, "But this time instead of A, A2" and Nature, speaking through all her laws, replies, "Then B2" and naturalizes the immigrant . . . . (pp. 80-81)
(3) Hume says that no miracle can be proved as it relates to the foundation of a religion (again begging the question, though we let that pass). But other miracles found in less important contexts may be possible.
For I own that otherwise [from the foundation of a religion] there may be the possibility of miracles, or violations of the normal course of nature, of such a kind to admit of proof from human testimony; though perhaps it will be impossible to find any such in all the records of history. (Hume, p. 127)
Despite Hume's concession, he still believes that he has a watertight case against even non-foundational miracles because they are mentioned only in recorded history. And such history is unconvincing, for the more remote it is, the less reliable it is (p. 109). Incidentally, this means that the Christian religion, founded by the miracles of Jesus, notably his Resurrection, have no reasonable foundation (p. 130-31). Finally, Hume already stated that miracles did not happen in his modern times and enlightened society (pp. 119-20). It seems, then, that he has once again shut the prison doors on non-foundational miracles, so his concession is empty. Nonetheless, we should take what we can get from Hume and his super-high, cannot-lose definition of miracles, so maybe the prison doors are left a little ajar. This is all the more true if we move forward from an investigation into past history and towards miracles today.

(4) Hume says that probability, not a full proof, may be a criterion for determining the veracity of witnesses for non-foundational miracles (p. 127). Maybe this probability (or perhaps strong possibility) is all that an open-minded person needs in order to move in the direction of belief.

(5) Hume may allow another way out of his prison. He says that "if a person claiming divine authority should command a sick person to be well . . . which immediately follow upon his command, [this] might justly be esteemed [a miracle]" (p. 115, note 1). He also lists other miraculous events, but they do not concern us here because in the linked article Do Miracles Happen Today? (see below), we limit the testimonies to recovery from physical ailments. Has anyone recovered immediately after words of prayer or even commands of healing have been spoken?

What about today?

What if today miracles happen that have been verified by the science that examines cause and effect in the human body? It is one thing to rely on an ace up your sleeve-no one can find sufficiently reliable historical records, and miracles simply don't happen in the Age of Enlightenment. But what about miracles we can see with our own eyes, in the twentieth and twenty-first centuries, in the Age of Science? Technology may render testimonies in defense of miracles probable and exonerative. Unfortunately, Hume lived before these modern times so he could not avail himself to confirmatory, high technology.

Part one in this series on miracles may be read here.

James Arlandson is a frequent contributor to American Thinker. He may be reached at
If you experience technical problems, please write to