Fortifying Hume's Miracle Prison (2): Miracles and Historical Testimony

In our Part Two here, we turn our attention towards the historian's task of investigating past events. The same problems confronting the believer in the realm of science emerge in the realm of history. The same basic regularities that happen today also happened in the past, so they preclude miracles. Or the regularity of past and present events makes miracles highly improbable compared to people's testimony.

The goal of this essay is modest. It is to keep Hume's prison doors open, so that professional skeptics do not close them even before investigating them. This essay, repeating and reinforcing the aim throughout, is a follow up to an earlier one: Fortifying Hume's Miracle Prison (1).

For our purposes a believer says that miracles happen, a theist is a believer, and a skeptic says that miracles do not and even cannot occur.

Aren't miracles maximally improbable?

Theistic philosopher Francis Beckwith explains what "maximally improbable" means: ". . . [T]here are no events more improbable than miracles, which is to say that miracles are the most improbable events that can be conceived" (Beckwith, "History and Miracles," p. 94, emphasis added).

Skeptic J. L. Mackie (d. 1981) says that miracles are indeed maximally improbable:

[The believer] must in effect concede to Hume that the antecedent [causally or logically prior] improbability of this event [miracle] is as high as it could be, hence that, apart from the testimony, we have the strongest possible grounds for believing that the alleged event did not occur. This event must, by the miracle advocate's own admission, be contrary to a genuine, not merely supposed, law of nature, and therefore maximally improbable. It is this maximal improbability that the weight of the testimony would have to overcome (p. 25, emphasis original).
In reply, however, theistic philosopher Keith Ward challenges this in three ways.

First, "there is something odd about trying to balance the improbability of a testimony being mistaken against the improbability of some event occurring" (p. 132). Ward explains further:

What has gone wrong here is the idea that we have two independent probabilities to balance against one another. It would not, despite Hume, be reasonable to say that the improbability of my table rising into the air and the improbability of my being mistaken in claiming to see it, just cancel each other out; so that I must remain agnostic, refusing to believe my own eyes because of some probabilistic balancing-act. On the contrary, as long as I pinch myself and look especially carefully, it would be reasonable to accept, without tentativeness, that the improbable has certainly occurred, and that I have certainly seen it. (p. 133)
That is, if we witness an actual miracle with our own eyes, do we have to deny our capacity to be accurate just because of the improbability balancing act? Why should our senses be refused? We may be extra careful in making sure that they are not. Ward adds:

I will reasonably place a greater weight on the trustworthiness of my senses than on any set of customary expectation of how the world may go. And I will reasonably accept the similar testimony of reliable witnesses in good conditions of observation, in default of some plausible explanation of how they could have fallen into error (p. 133)
Second, Mackie frames his argument in a way that believers must also, along with skeptics, assume miracles are maximally improbable. Ward replies:
But that seems quite false to the way miracles are generally portrayed in the Biblical tradition, at least. If someone believes that there exists an omnipotent God, who created the universe for a purpose; then there is an antecedent probability that he will act within the universe to accomplish his purpose. It seems antecedently improbable that God's purpose could be accomplished without any action in the world by him. (p. 134)
Ward is right about this. Once we allow the existence of God, then we thereafter cannot ultimately dictate terms. A believer steps in and says that his God works miracles.

Third and finally, to counter the believer's acceptance of God's existence, a skeptic may assert a closed system, what Ward calls "scientific objectivism"-"the theory that all events in the natural universe are non-purposive, and wholly determined by general physical laws, which form a closed, universally determining system" (p. 136). But in reply, Ward says it is unfalsifiable and hence suspect. It proves too much. He writes:
If the postulate [of scientific objectivism] is accepted, it will make highly improbable the occurrence of any falsifying instances-such as particular purposive occurrences in nature, not determined by laws of regular succession alone. But such falsifying instances may occur, and must in a sense be looked for, if the explanatory postulate is to be reasonably upheld . . . It will not be enough to say that, since they are maximally improbable on the theory, no testimony to their occurrence will be acceptable. That will be obviously cheating (p. 136)
Thus, a skeptic must keep open the prison doors shutting in miracles because if miracles are completely excluded, then once again this begs the question in favor of naturalism or scientific objectivism.

Beckwith agrees:
But we can know that miracles are maximally improbable only if we already know that miracles could or have never occurred or that they are logically impossible (that is, conceptually impossible, like a "square circle" or a "married bachelor). Opponents of miracles must therefore maintain that miracles are maximally improbable because they already know that miracles could or have never occurred . . . Consequently, opponents of miracles beg the question if they claim that miracles are maximally improbable. ("History and Miracles," p. 94)
Beckwith goes on in the same chapter to warn that we should not be gullible. The laws of nature are regular. But they are not "inviolable," unless we assume in advance that they are not. So the door to them must be left open in the face of sufficient testimony; otherwise, the wise man risks becoming "dogmatic and obscurantist" in relentlessly denying their occurrence (Beckwith, p. 95).

Finally, Ward likewise concludes this about the maximal improbability argument:
What has gone wrong is that, by speaking of a violation of a law of nature as maximally improbable, we ensure that mistaken observation can never be more improbable. If our claim is the more moderate and reasonable one that such violations are contrary to normal expectation, it is no longer always much more unlikely that events should occur, and be correctly observed, which go against the run of expectation. Indeed, it is not particularly improbable that such improbable things should occasionally be observed to happen, unless you have a general theory-such as "scientific objectivism"-which renders them impossible. (p. 138)
How much evidence does it take?

Can a genuinely miraculous event be known on historical evidence? Flew answers in the negative.

The criteria by which we must assess historical testimony, and the general presumptions which alone make it possible for us to construe the detritus [remains] of the past as historical evidence, must rule out any possibility of establishing, upon purely historical grounds, that some genuinely miraculous event has indeed occurred (God and Philosophy, p. 150)

However, Larmer replies to Flew's skepticism, which stands on a faulty definition of a miracle:
Flew's argument in support of this conclusion amounts to the following: (1) historical investigation presupposes the truth of the laws of nature; (2) the laws of nature must be defined as exceptionless; and (3) miracles must be defined in terms of inconsistency. This argument is valid, but not sound. Leaving aside discussion of Flew's first two premises, his third premise cannot be defended. Fundamental to the concept of miracle is the idea that nature must be overridden if a miracle is to occur. But Flew is mistaken in thinking that this implies that the laws of nature must be violated. As I have already shown, miracles, considered as objective events caused by God, can conceivably occur in a world which behaves, always and everywhere, completely in accordance with the laws of nature. (Water into Wine? p. 100)
Thus, Larmer says that Flew assumes that the laws of nature must be violated if a miracle occurs, but a miracle can take place in accordance with the laws of nature. When God works a miracle, he changes "the material conditions to which the laws of nature apply," but he does not violate his own laws.

What about studying history?

Flew spells out three propositions concerning the historian's philosophy going into his study of history. His philosophy boils down to the uniformity of the past and the present. Flew says:
The basic propositions are: first, that the present relics of the past cannot be interpreted as historical evidence at all, unless we presume that the same fundamental regularities obtained then as still obtain today; second, that in trying as best he may to determine what actually happened the historian must employ as criteria all his present knowledge, or presumed knowledge, of what is probable or improbable, possible or impossible; and third, that, since miracle has to be defined in terms of practical impossibility the application of these criteria inevitably precludes proof against miracle. (God and Philosophy, p. 150)
These propositions can be examined more closely on three fronts.

First, it is true that continuity and consistency between the past and present must exist for us to acquire historical knowledge. But as stated, Flew's watertight propositions beg the question, which assumes the answer at the start of the investigation, thus making the investigation hollow and unable to be defeated or falsified. Beckwith explains:
It is one thing to make the uncontroversial claim that historians must assume some continuity of regularities between the present and the past to have historical knowledge. It is quite another thing, however, to claim . . . that we must assume a nonmiraculous worldview in order to have historical knowledge. Such a position calls for the automatic rejection of new data, regardless of how well grounded evidentially, which may support the historicity of the miraculous and count against the nonmiraculous worldview. ("History and Miracles," p. 97)
Second, if we assume that a report about a miracle is well founded, then it becomes "highly artificial as well as woefully inadequate" to deny consistently and relentlessly any and all reports and evidence about miracles. We do not weigh one probability but many of them, "a convergence of independent probabilities" (Beckwith, p. 95). At the conclusion of Beckwith's chapter "History and Miracles," he writes: "Even if miracle claims need support from more evidence than ordinary claims in order to be held rationally, it can sometimes be the case that one has sufficient evidence to believe that a miracle has occurred" (p. 98). The evidence can mount up.

Third, Flew's propositions assume that miracles do not continue into the present. "Defenders of [Flew's] argument must show that there are no present miracles and not merely to assume there are not" (Beckwith, "History and Miracles," p. 97). This may be stated positively. In the next article in this series ("Do Miracles Happen Today?"), it will be shown that believers today observe and experience miracles; therefore, their testimony says that they indeed occur now. It is special pleading to ignore their testimony that counters the skeptics' ironclad position, without investigating it on a case-by-case.

Does Flew apply his skepticism consistently?

Corduan analyzes a part of Flew's challenge to believers, which I have put in three steps.

First, Flew says that Christians believe in God's love, no matter how much evidence from evil may militate against the belief in his love; they redefine the concept so that God always maintains his reputation. It seems Christians refuse to listen to contrary evidence and to change their mind.

Corduan summarizes Flew's challenge:
Whenever there is a challenge to their belief in God's love, [Christians] simply redefine the concept so as to avoid having to deal with the possibility that maybe the loving God in whom they believe does not exist. ("Miracles," p. 173)
In one of Flew's earlier works, he says that the assertion that God loves us undergoes a "death by a thousand qualifications" or counter-examples of his love ("Theology and Falsification," p. 97). He then asks: "What would have to occur or to have occurred to constitute for you [believers] a disproof of the love of, or the existence of, God?" (p. 99).

Corduan goes on to say that the problem of evil is hotly contested. It is possible to maintain a belief in God's love in the face of evil, but Corduan's goal is not to argue that point.

Second, Corduan spots an inconsistency in Flew's challenge to believers about evil and the love of God. This is, the believers' never-lose position is exactly how Flew opposes the reality of miracles: his skepticism can never lose.

According to Flew, it is contrary to the nature of science to make allowances for the supernatural, and historical reports should not be permitted to violate the nature of reality as science circumscribes it for us. According to Flew, that does not mean that science necessarily has the correct explanation for every historical event, but that whatever the ultimate explanation may be for a special event, it cannot be miraculous. (Corduan, p. 174)

Third, how does Flew's skepticism about miracles parallel the believer's (seeming) unquestioning belief in God's love in the face of evil? Corduan answers the question:

Consequently, Flew's claim that no events are miraculous is completely unfalsifiable, and, just as he accuses the Christian in his defense of the love of God, it becomes a meaningless mantra. No evidence could conceivably count against his rejection of any event as miraculous. (Corduan, p. 174)

Thus, we may pose this question to Flew and other neo-Humans, rephrasing the one he asked of believers in the love of God: "What would have to occur or to have occurred to constitute for you a disproof of antisupernaturalism?" (Geisler, "Miracles and the Modern Mind," p. 83).

What is your mindset?

It has been repeated throughout this essay that the prison doors locking up miracles must be kept open, if a fair and proper investigation into them is to be done. If the skeptic shuts them before the investigation begins, then this begs the question always in favor of naturalism. This worldview says that nature is all that exists, and it cannot be influenced by causes outside of itself; it is the Only Fact.

Most opposition to the possibility of miracles in the West comes from people controlled by the presuppositions of naturalism. Most such people refuse even to consider evidence that appears to support the actuality of miracles...because-consciously or unconsciously-their minds are closed on that subject. Miracles are judged to be impossible before the fact. (Ronald Nash, "Miracles and Conceptual Systems," pp. 130-31, emphasis original)

That is, miracles must be physically possible, not merely logically (conceptually) possible. And to determine whether they in fact occur, an open and fair investigation must be done. We must leave our offices and computers and look for evidence.

So what does all of this mean?

The Conclusion here applies also to Fortifying Hume's Miracle Prison (1).

Miracles happen, if they do, in the realm of perceptions, in the empirical realm, in the world of matters of fact. Therefore, they are investigable. However, can sufficient testimony ever outweigh the (nearly) unalterable laws of nature? If the claims of miracle are poised on one side of the scale, and the laws of nature on the other side, which side is heavier? Where should we place our confidence? Yet, it is one thing to dispute over miracles two thousand years ago, but what if miracles happen today under the watchful eye of science?

Further, what if the evidence for a miracle is multifaceted? What if it amounts to a convergence of probabilities (Beckwith, "History and Miracles," p. 95)? For example, a man has a lump under his arm. He and his wife can feel it. A doctor feels it too and concludes that it is likely a tumor. Also, a CT scan says that a tumor is there. A biopsy is done. A tumor is confirmed.

However, a new factor is introduced into human biology or the laws of health. People pray directly to God to remove the tumor. After the prayer, the lump has vanished. A second CT scan shows that it is completely gone. The oncologist did not treat it.

Here we have a convergence of factors that make the evidence of a miracle strong. The lump was visible (sight) to the patient and his wife and others who are untrained in medicine. One could feel it with one's hand (touch). Qualified doctors-modern representatives of science-examined the tumor with their own eyes. They performed a biopsy. A CT scan-modern technology-confirmed it was there. But after prayer, it disappeared. The oncologist is baffled, but gives the (former) cancer patient a clean bill of health. After further check ups, the tumor has not returned, no traces. For most people, the side of the scales tips, if only a little, in favor of a miracle.

The meaning of skepticism in ancient Greek is "to look closely" or "to examine."  Skeptics have a duty to their name to investigate.

The next part in the series on miracles considers the evidence of real-life miracles today. It is time we supplement the philosophical arguments with empirical evidence, using modern technology to examine and perhaps confirm miracles-though I claim here in advance that, according to the evidence proffered in the next article, miracles in fact occur.

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

Part two 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