Is It Rational to Believe in Miracles?

Whether it is rational to believe in miracles depends a great deal on what one might call the ‘big picture’ we have of reality, the assumptions and ideas that form the canvas or background to the way in which we interpret and understand the world. Even the word ‘miracle’, from the Latin miraculum meaning ‘object of wonder,’ implies a contrast with a background, an event that is remarkable and different from what is experienced ordinarily. Wonder alone, however, is not enough to specify a miracle. As Thomas Aquinas observed (Summa Contra Gentiles 3.101.1), an astronomer is not astonished when he sees an eclipse of the sun since he knows the cause. A miracle, by contrast, exceeds the productive power of nature, either because something is done by God that nature could never do (such as the sun reversing its course), or that nature can do but not in the observed order (such as an animal living, after death), or because God does what is usually done by the working of nature but without the operation of nature (such as a cure by divine power from a fever that could be cured naturally). The recognition of what is generally denoted by the term ‘miracle’ presupposes therefore a robust appreciation of the regular course and powers of nature. As these examples suggest, the use of the word ‘miracle’ also implies some theological background: a supernatural cause with intellect, will and the power to intervene in nature, most often identified as a personal God.
At this point, it should be noted that this definition of a miracle is not the one made famous by David Hume in An Enquiry Concerning Human Understanding (X), namely “a violation of the laws of nature.” There is not the space here to survey the voluminous discussions generated by Hume’s definition and I refer those interested to the excellent article on ‘Miracles’ by Timothy McGrew in the Stanford Encyclopedia of Philosophy. The key problem is that it is hard to see in many cases how this redefinition adds clarity. For example, it is not always easy to identify what laws are being violated in specific reported cases, such as a resurrection, and there are many disagreements today about the nature and efficacy of natural laws. The definition was perhaps more plausible in the eighteenth century, when there was still confidence in the cosmos being wholly law-like or even machine-like, perhaps a vastly more complex version of a clockwork planetarium. Contemporary research, however, suggests a more ‘organic’ cosmos, including spontaneity and phenomena that are naturally generated but irreducible, even in principle, to states described by the application of laws to initial conditions. Such developments cannot be used unproblematically to argue for ‘gaps’ in the world where God slips in unnoticed, but they highlight how even natural phenomena are not invariably ‘law-like’. Hence the specification of a miracle as a violation of a natural law lacks sufficient precision today.
Once the confusion generated by the issue of laws is dispelled, it is somewhat easier to see that special divine action of any kind cannot be ruled out definitively, either by the methods of science or by the philosophical study of the world. Even in cases where the behaviour of a physical system is relatively simple and law-like, such as a pebble falling under gravity, that fact cannot be used to draw any conclusions about whether a person, human or divine, will choose to intercept this fall. The more difficult challenge is to obtain sufficient warrant for adjudicating such actions when the personal agent is not perceptible directly. After all, the scientifically-informed study of particular events cannot be used to investigate any purported supernatural cause directly, but only the effects and possible alternative and natural explanations. Hence the classical definition of a miracle as an event that exceeds the productive power of nature cannot be applied without some concurrent assessment of what nature is capable of producing or has produced in the case in question. Of course, some reported miracles exceed the powers of nature so dramatically that they practically compel those present to believe, such as the account of the resurrected Jesus appearing to Doubting Thomas and inviting him to put his finger into the wounds made by the nails of the crucifixion (John 20:24–29). In many cases, however, it is often possible to think of at least some possible, even if otherwise implausible, non-miraculous explanations. Any assessment therefore still depends crucially and irreducibly on the natural and theological background or ‘big picture.’ Does God exist? Is God willing and able to interact with the created world? If so, why does God intervene in some ways and at some times but not in other instances, especially against evils? Hence judgments about miracles, in general and in particular, can never simply be a matter of adding more facts, but depend crucially on how these facts are understood and organized into an ordered whole. For this reason, the most persuasive advocates of decisive views on the subject of miracles tend, upon close examination, to be masters of metaphor and effective story-tellers, even though they may at times borrow the language of scientific objectivity.
Given that miracles cannot be ruled out definitively except within the framework of certain all-encompassing worldviews, it is not irrational in an unqualified sense to believe in miracles, in the sense of believing in a square circle. This fact does not mean, however, that it is equally rational to believe in every possible explanation of some wondrous event that is compatible with what is reliably known. On this point, there is some parallel with a challenge faced in science. At the Large Hadron Collider, for instance, once all the data in the search for the Higgs boson are collected, heavily theory-laden curves are then traced through the data points to find the best fit. Other causes are not in fact ruled out, but at some point a judgment is reached about the most likely underlying cause. In recent centuries, two heuristics have proved especially fruitful in guiding such judgments: first, that the underlying causes of the cosmos are in a vague sense ‘simple’; second and perhaps more mysteriously, these causes are understandable. In the case of miracles, what are the equivalent, most fruitful heuristics for guiding rational assessment?
The minimal baseline and least controversial heuristic is that miracles are rare, practically by definition, which implies the need for caution in ascribing direct supernatural agency. For ancient Judaism and Christianity, for example, reports of miracles are generally associated with the initial phases of new developments and are even treated as morally suspect if they happen too often. For example, a curious letter of Pope Gregory I (Bk 9:36) warns Augustine and his monks, who arrived in England in 597 as missionaries, not to become too proud of their power to work great miracles. This caution echoes the gospel warning that miracle-working is no guarantee of final human flourishing (Matthew 7:22), which is what God is reported to desire from interacting with the world at all (1 Timothy 2:4). Given the cognitive and moral risks associated with interpretations of events as miraculous, acknowledged even by those who hold that miracles are possible, why consider them at all? Why is it not preferable in practice to make use of a strict ‘no miracles’ heuristic?
First, anticipating a common objection, history offers little credence to the view that consideration of miracles as a possibility promotes dispositions that are irrational or hostile to what is today called ‘science’. On the contrary, assessments of purported miracles as defined by Aquinas require a robust sense of natural powers, and it may be no accident that the development of experimental methods to isolate and identify causes can be traced back to the same period. Second, miracles are one aspect of a broader understanding of the world under the providential direction of God, who intervenes out of love for humanity. This starting point seems more promising in general for the fruitful perception of harmony in the cosmos than the belief that we are nothing but meat-machines or cosmic accidents. Third, this starting point is also more hopeful in the personal sense, since the world of natural causes appears to lack sufficient resources to make us perfectly and permanently happy. Out of enlightened self-interest and inspired by Pascal’s wager, it seems foolish therefore to treat the exclusion of special divine action as a fundamental premise. Finally, the particular ways in which supernatural personal agency is excluded from consideration often impact severely on interpretations of other important matters, including causation and natural personal agency. If, for example, the universe is wholly law-like or even machine-like, what is a human being apart from a remarkably complex cog? In my view, we should be slow to accept such cold, self-limiting conclusions as long as we have reasonable alternatives. Since we do have reasonable alternatives, I conclude therefore that a heuristic that regards miracles as possible in principle, rare in practice, but still worthy of consideration in certain cases, to be the most rational stance to adopt, one that is ultimately the most interesting and promising for intellectual and personal flourishing.
Discussion Questions 
1. Does belief in the possibility of miracles require belief in the general consistency of natural causes and effects?
2. Are miracles compatible with a strictly physicalist conception of nature? If so, how are miracles to be distinguished from instances of natural causation? If not, why not?
3. The definition of the miraculous in this essay does not include events that come about naturally but are so extraordinarily well-timed that they might suggest intervention. Should the definition of a miracle be expanded to include such events and what criteria could be applied to adjudicate their meaning?
4. The heuristic proposed in this essay in regard to miracles is, “possible in principle, rare in practice, but still worthy of consideration in certain cases.” Is a different heuristic preferable and, if so, why?

Discussion Summary

A key theme that emerges from the comments on the original essay, “Is it Rational to Believe in Miracles?” is the challenge of demarcation, not only of the meaning of the term “miracle” but also of the term “rational.” Several commentors referred to the long-standing problem of demarcating miracles from natural phenomena, against a backdrop of assumptions that range between the complete exclusion of miracles to the perception of something miraculous about every moment of creation. Although the latter may be a desirable attitude for jaded eyes, both the “no-miracle”  and “all-miracle” perspectives exclude the possibility of a meaningful distinction from natural phenomena and forestall further investigation. In practice, as noted in the essay, the perception of miracles and their philosophical study therefore require that a purported miracle be contrasted with what is ordinary, whether this contrast is by means of a sudden, amazing encounter (such as meeting a resurrected, previously crucified man) or by a careful and systematic exclusion of natural causes after the event, such as the investigation of the healing of an eye.
This need for a contrast is easy to state and has long been acknowledged in historical writings on these matters, even though many valuable texts are little known today (a problem we shall be addressing with our John Templeton Foundation -funded “Special Divine Action” project at Oxford University). For example, the observation in 1767 of William Adams, “A river must flow, before its stream can be interrupted” (An Essay in Answer to Mr. Hume’s Essay on Miracles, 3rd ed., London: B. White, 15), expresses one of the preconditions of the recognition of a miracle, namely the ordinary experience of uniformity in the course of nature. This experience of uniformity in turn requires other capacities, such as experience extended over time, memory, and the ability to identify and exclude a wide range of natural causes. For such reasons, as noted in the essay, it may be no accident that the development of experimental methods to isolate causes can be traced back to the roughly the same period of history as detailed philosophical demarcations of the miraculous from what is natural. Working therefore on the basis of the heuristic advocated in the essay (“miracles as possible in principle, rare in practice, but still worthy of consideration as possible explanations for certain events”) the first Big Question I propose is, “(1) For a range of common views of reality, by what kinds of criteria will miracles most stand out, and what cognitive capacities would be most effective for identifying them if they take place?”
A second Big Question that emerges from the essay and ensuing comments is, “(2) To what extent do views regarding personal divine action also shape views about personal human action?” Expanding on an example in the original essay may help to illustrate the connection. Imagine throwing a pebble into the air. As long as the motion of this pebble is uninterrupted, its trajectory will be described by an application of the law of gravity except for the braking effect of air resistance. If now I reach out and catch the pebble, what has happened? If my choice of action is not determined by those operations described by “laws of nature” (as that phrase is commonly understood) acting on prior states of the cosmos, then presumably we are free to alter the outcomes of such operations on at least some other beings in the cosmos, such as pebbles. If we enjoy this freedom of action, then there does not seem to be any good reason to rule God out from enjoying such freedom as well, for example by means of what are called “miracles.” If, on the other hand, the presupposition is made that nothing happens or could happen in the cosmos that is not describable, at least in principle, by laws of nature applied to previous states of the cosmos, then any special divine action that brings about an outcome that diverges from the expected outcome of an application of these laws is excluded. By the same assumption, however, any human action that brings about an outcome that diverges from the application of these laws is also excluded, to the point that human beings are in principle reducible to mere matter in motion. On this worldview, a severe restriction of what is deemed possible for divine action is not easy without also restricting severely what is deemed possible for human action, to the point that the effective abolition of God and of man appear surprisingly interconnected.
New Big Questions:
1. For a range of common views of reality, by what kinds of criteria will miracles most stand out, and what cognitive capacities would be most effective for identifying them if they take place?
2. To what extent do views regarding personal divine action also shape views about personal human action?