Robert Larmer and Christine Overall have a longstanding debate over whether the occurrence of a miracle would be evidence for or against the existence of God. Overall argues that if a miracle occurred, it would constitute evidence against the existence of God; Larmer, on the other hand, argues that the occurrence of a miracle would be evidence for the existence of God. Presently, the debate between them is several rounds deep, yet both remain persuaded of the main tenets of their respective positions. Larmer, for his part, has commented, in his most recent riposte, that the debate has largely run its course.Footnote 1 If by this comment Larmer means to say the prospects of one of them convincing the other through clarification and elaboration of what has already been said are dim, then I agree. But if Larmer means to announce that the issue has been settled or that all of the important moves have been made, then his announcement is, I suggest, premature. In this paper, I will argue that, where the debate stands now, neither Larmer nor Overall has got what the occurrence of a miracle would mean for the existence of God quite right. Where to go from here, I suggest, is a matter of identifying and synthesizing what is correct in their respective accounts.
I begin with Overall’s most recent version of her positive argument. Overall’s philosophical investigation of miracles begins with a questioning of what the occurrence of a miracle would imply about God.Footnote 2 In order to properly understand Overall’s answer to this question, it is necessary to review the way she defines the question’s key terms. First, by “God” she means the traditional Christian understanding of God as “the omnipotent, omniscient, and all-loving creator of the universe”.Footnote 3 Second, by “miracle” she means “a break in the space-time causal sequence”.Footnote 4 Now, if a miracle, so understood, were to occur, then it would constitute an instance of three different kinds of evil. First, it would be an ontic evil insofar as “it would be a flaw in the order, regularity, and predictability of the universe, a divine afterthought, an ad hoc intervention”.Footnote 5 Second, it would be a cognitive evil since “it would interfere with the attempt to understand the structure and regularities of reality”.Footnote 6 Third, it would be a moral evil, since it would reflect triviality, capriciousness, and bias.Footnote 7 Any event which is evil in these three ways is an event the occurrence of which increases the amount of evil in the world.Footnote 8 But increases in the amount of evil in the world are incompatible with the existence of a perfectly good, perfectly informed, and perfectly talented God. If a miracle actually did occur, then, it would be evidence that God does not exist.
Larmer’s criticisms of this argument can be collected with little remainder into two broad categories: complaints about Overall’s definition of “miracle” and complaints about whether miracles constitute ontic, cognitive, and moral evils. As for the first category of complaint, Larmer argues that Overall gets the definition of “miracle” wrong by not paying attention to the actual use of the term. When we use a term like “miracle,” he says, we are committed to a judgment of the cause of the event in question, specifically the judgment that the event is the result of divine activity.Footnote 9 For his part, Larmer builds divine causality into his own definition of a miracle as “an unusual and religiously significant event beyond the power of nature to produce and caused by an agent who transcends nature”.Footnote 10 Unlike this definition of “miracle”, Overall’s doesn’t include divine causality, leading her argument astray.Footnote 11
Larmer’s definitional objection fails because use of the term “miracle” is not homogeneous in the way he contends. Consider R.F. Holland’s famous distinction between the contingency concept and the violation concept of a miracle. Responding to the unduly restrictive condition of requiring an event to be a violation of a law of nature in order for it to qualify as a miracle, Holland identifies a concept of the miraculous according to which no law of nature is broken but a contingent and unexpected but impressive and significant event occurs.Footnote 12 This he terms “the contingency concept” of the miraculous.Footnote 13 On corresponding uses of the term “miracle,” there is clearly no commitment to the cause of the event in question and no law of nature needs to be violated; what makes an event a miracle, on the contingency concept, is not who or what is responsible for a violation of a law of nature but the impressive significance of the event’s happening when it didn’t have to. Holland’s distinction creates a problem for Larmer’s criticism of Overall’s definition of “miracle” because it offers an alternative concept of miracles—the contingency concept—that does not involve supernatural causation but might be invoked in our use of “miracle,” calling into question the idea that “an essential part of the concept of a miracle is … supernatural causation”.Footnote 14
But even if Larmer could demonstrate that the use of the term includes divine or supernatural causation, objecting to Overall’s argument on definitional grounds is still problematic. Over the course of their debate, Overall has repeatedly emphasized that the evidentiary connection between miracles and God needs to be critically assessed, it cannot be settled by definitional considerations.Footnote 15 To continue to argue that Overall has got the definition of “miracle” wrong and so errs in her argument is thus to, in an important way, miss Overall’s point. Overall demonstrates this with her tongue-in-cheek surrender of the word “miracle” to Larmer to define as he wishes, in turn defining the word “dunkle” as a break in the space-time causal sequence.Footnote 16 The issue then becomes what the occurrence of a dunkle would mean for the existence of God. As an instance of ontic, cognitive, and moral evil, a dunkle, if one were to occur, would be evidence against the existence of God.
As for his second category of complaint, Larmer charges that if they actually occurred, biblical miracles (which Overall and Larmer agree are paradigm descriptionsFootnote 17 of miraculous events) did not constitute the sorts of evils that Overall alleges. On the contrary, argues Larmer, “the vast majority of events identified as miracles by Judaeo-Christian believers constitute a lessening of evil in the world.”Footnote 18 Any apparent bias and triviality of miracles can be accounted for in a morally satisfactory way by appeal to the specifics of each case and the broader context in which biblical miracles are said to have occurred, which Overall has ignored or underappreciated.Footnote 19
This complaint also fails, for even if the examples of biblical miracles upon which Overall’s and Larmer’s debate has focused so far (the conversion of water into wine or the virgin birth, for instance), are not persuasive examples of the three kinds of evil there are many more events described in the Bible which meet both Larmer’s as well as Overall’s definition of a miracle and would be more clearly instances of evil, if they indeed occurred. Take, for instance, God’s response to the complaints of Korah, Dathan, and Abiram and their 250 supporters as recorded in the book of Numbers. “…[T]he ground under Dathan and Abiram split open and swallowed them and their families, together with all of Korah’s followers and their possessions. … Then the Lord sent a fire that blazed out and burnt up the 250 men who had presented the incense.”Footnote 20 This is clearly an unjust reaction, and it is precisely such injustice that worries Overall about more standard examples of biblical miracles.Footnote 21 One may be able to contextualize turning water into wine for the sake of a good party, or turning loaves to fishes to feed 10,000 lest millions die of hunger, but it is a good deal harder to contextualize spontaneous mass murder of religious adherents.
All of this suggests that Larmer’s criticisms to date are unsuccessful. I will now present a thought experiment I think gives us a new reason to suppose that miracles do not constitute the sorts of evils Overall alleges. Suppose two people, Lettie and Mac, have spent their entire lives within a hollow sphere of a large and fixed diameter. Although they have tried, Lettie and Mac have been unable to flex the boundary of the sphere, nor have they found any force within the sphere sufficient to do so. The sphere is translucent, letting in enough light and allowing them sufficient resources to live recognizably human lives. In fact, Lettie and Mac’s lives are very much like our primitive ancestors’ lives but for the boundary of the sphere.
Part of their daily routine is to explore the sphere systematically to learn about their surroundings. They do this by way of a cycle of hypothesizing, experimenting, and observing. They follow this methodology meticulously in their investigations. Using it, Lettie and Mac accumulate a vast amount of data on the behaviour of the objects within the sphere over the course of the finest years of their lives. They even go so far as to formalize their knowledge of the sphere into rules, which they call ‘the laws of the sphere.’ One of the laws of the sphere is that the objects within the sphere exhibit a certain attraction to one another that varies in a mathematically precise way according to the mass of the objects involved.
One day, while Lettie is alone and testing this particular law for the umpteenth time, something strange happens. As she unfetters her test object, a rock, the shell of the section of the sphere nearest the rock caves, as if it were pushed inward from the outside by a giant finger. The section of sphere envelops the rock, immobilizing it and cancelling its attraction to the other objects in the sphere. Upset by the failure of the law of the sphere, which held that the rock should have been attracted to the other objects of the sphere, Lettie runs to tell Mac what has happened. Subsequent attempts to repeat the anomalous event in similar circumstances fail and leave Mac and Lettie in a state of bewilderment.
I take it that the event Lettie witnessed was a break in the space-time causal sequence insofar as some force outside the sphere (whose identity is unknown) interrupted the space-time causal sequence which would’ve otherwise occurred within the sphere.Footnote 22 I also take it that it is a very confusing event for Lettie and Mac. But an event’s being confusing is neither necessary nor sufficient for its being evil; convincing magic tricks come to mind as examples of events that baffle but aren’t plausibly evil, and plenty of evils aren’t confusing. So to argue it is evil we will have to go beyond the fact that it confuses Lettie and Mac. That said, does the event in question constitute an instance of the three kinds of evil Overall alleges such events constitute?
It certainly seems to have been a cognitive evil for Lettie and Mac since it presents a counter-instance to a well-confirmed law which they took to be correct, leading them to question the integrity of that law and hence the aspect of their reality which it captures. However, notwithstanding their justified confusion, I do not think a cognitive evil has occurred because their understanding of their world isn’t supposed to capture what happens when their world is subject to external influence. This idea is consistent with a well-known view in the literature of the laws of nature. Steve Clarke puts it thus: “Laws of nature are of nature; they should not be expected to account for the behavior of things which have been supernaturally interfered with”.Footnote 23 David and Randall Basinger similarly write, “Direct acts of God cannot be violations because an event can only be a violation of a natural law, strictly speaking, if there are no nonnatural causal factors involved”.Footnote 24 Finally, William Rowe writes, “the laws of nature tell us what must happen only if what happens is due entirely to natural forces”.Footnote 25 Put in terms of the example of Mac and Lettie, if the immobilization of the rock was a result of extrasphere causal influence, Mac and Lettie’s laws of the sphere are still perfectly good laws of the sphere—they have not admitted of a counter-instance or violation of which their laws must take account. So their attempt to understand their reality, consisting of the formulation of good laws of the sphere, is intact and unchanged, i.e., not frustrated by the event Lettie witnessed.
On similar grounds the immobilization of the rock does not count as a flaw in the order, predictability, and regularity of the universe, i.e., as an ontic evil. If the rock had not exhibited the same attractions it always had and the boundary of the sphere was not manipulated from outside, then we could say that the sphere had become a less orderly, predictable, and regular place. But when the sphere is interfered with by something external to it, the order, predictability and regularity of the sphere itself (analogous to what we would ordinarily call “nature”) are not called into question. Only when the behaviour of the objects in the sphere does not accord with the laws of the sphere and there is no external influence should we say the order, predictability, and regularity of the sphere is impugned.
Finally, the immobilization of the rock doesn’t seem to be a moral evil. It might be argued that the event exhibited bias in favour of Lettie since it occurred only in her presence. This is right, but the bias is not a vicious one insofar as Lettie does not stand to gain anything of value which Mac must go without, as Overall finds so objectionable in, for instance, the biblical story of the conversion of loaves to fishes. It might alternatively be argued that the event exhibited capriciousness and triviality since whatever was responsible for the largely insignificant event Lettie witnessed could have done something much more significant instead. Yet this event was indeed, in a sense, the most non-trivial event two people passionate in their search for knowledge of the boundaries of their world (specifically, its boundaries and beyond) could hope to observe. I conclude that the event Lettie witnessed was not a moral evil, at least in the sense in which Overall is interested.Footnote 26 Accordingly, if successful, this thought experiment shows that if a miracle, on Overall’s definition, occurred, then in spite of being a miracle, it would not be of the right sort to wrinkle the universe, frustrate our attempts to understand it, or necessarily make the world more morally evil.
We now turn to Larmer’s argument for the claim that if a miracle occurred, it would constitute evidence for the existence of God. As mentioned above, Larmer defines a “miracle” as “an unusual and religiously significant event beyond the power of nature to produce and caused by an agent who transcends nature”.Footnote 27 If confronted by an event which appears to meet this definition, Larmer urges us to ask two questions. The first is whether we have good reason to think that the putative miracle actually occurred. In the tradition of Richard Swinburne’s writing about miracles, Larmer thinks that this question should be answered by the principled interpretation of three kinds of evidence: personal observation; relevant physical traces; and the testimony of others.Footnote 28 Like Swinburne, Larmer finds it possible for evidence to be sufficient to justify belief in a miracle.Footnote 29 His paradigm example is that of personally witnessing and filming the instantaneous regrowth of the withered arm of a religious man of exemplary character who had been praying for just this event.Footnote 30 In such an instance, the evidence from personal observation and physical traces is of sufficient volume, quality, and coincidence, to, in Larmer’s opinion at least, warrant the conclusion that an event of arm regrowth has actually occurred.
Once we have good reason to think an event occurred, we must, according to Larmer, ask whether it meets the conditions of a miracle on his definition, most notably the condition that the event was the work of a supernatural agent. This amounts, for Larmer, to deciding between naturalistic and supernaturalistic explanations of some event. “[I]f an event can be satisfactorily explained by theism as being a miracle, but physicalism can offer no satisfactory explanation of it, then we are justified in seeing it as independent evidence for the superiority of theism over physicalism”.Footnote 31 In other words, in virtue of the theistic hypothesis being the best explanation of an event, that event functions as independent evidence for the truth of theism. Returning to the example of the man with the restored arm, since (1) the man was religious and of exemplary character and had been praying to be healed; and (2) prior and subsequent attempts to regenerate similarly withered arms by way of medical science had presumably failed; and (3) it is well-confirmed that the man is not somehow biologically unique on this front, we have—with the case of withered arm regrowth—one instance where the theistic interpretation is to be preferred. Since we have good reason to think the event occurred and was the work of God, then we have—in the arm regrowth—independent evidence that God exists.
Overall has extensively criticized this argument as circular. “Larmer’s definition of ‘miracle,’ which incorporates the idea of supernatural causation,” Overall writes, “incorporates the very point at issue: that a supernatural being causes miracles”.Footnote 32 This is problematic because the very issue is what the occurrence of a miracle would mean for our evidence of the existence of God. “Larmer’s definition of ‘miracle’ short-circuits the whole debate by begging the question”.Footnote 33
On Larmer’s behalf, let me respond by pointing out that defining a category in terms of certain concepts does not presuppose that there are instantiations of those concepts. This, I think, is the point Larmer is trying to make when he says that his definition of a miracle only requires us to entertain, as a hypothesis, that God exists.Footnote 34 To define a miracle is, of course, logically distinct from asserting that a miracle exists, and also from asserting that the sorts of things implied by the occurrence of a miracle exist.
Nevertheless, there are circular arguments making use of Larmer’s definition of a miracle (or any definition that involves supernatural causation) lurking in the vicinity. Here is an example:
1. Whatever else it is, a miracle is the work of a supernatural agent.
2. A miracle has occurred.
3. There is at least one supernatural agent.
This is quite clearly a circular argument. But it cannot be Larmer’s argument, since, as we have seen, Larmer takes pains to show that the issue of what miracles tell us about the existence of God hangs on evidence rather than the definition of “miracle”. Even if we assume that a miracle has occurred—as both Overall and Larmer have agreed to do in the debate at hand—we do not know that a miracle has occurred without building that case by appeal to the evidence available. And only when we know that an event is a miracle do we have evidence for the existence of a supernatural agent or God. It is the evidence we can muster to the conclusions that the event occurred and was supernaturally caused that does the work in Larmer’s argument, not his definition of a miracle.
Overall also criticizes Larmer’s method as circular. As a self-identified theist, the argument goes, Larmer already accepts that events have occurred which are best explained as miracles prior to building his philosophical system. For instance, by virtue of his Christianity Larmer’s accepted description of the virgin birth must already include that the birth of Jesus was supernaturally caused. Overall’s worry is that, by virtue of his Christianity, Larmer has foreclosed on the issue of which hypothesis—physicalist or theistic—is best able to explain an event we have good reason to believe actually occurred, namely the event of Mary, a virgin and the wife of Joseph, giving birth to a boy named Jesus.
Larmer has certainly made no secret in his philosophical writing about miracles that he accepts the Christian description of the virgin birth and other paradigm miracles as works of God. But the important logical question is whether the cogency of Larmer’s argument depends on his having such beliefs. I do not think so. To see why, consider an example that has come up in the course of Overall’s and Larmer’s dialogue.Footnote 35 Suppose one comes upon a dead human body. Larmer’s suggestion is that one may entertain the hypothesis that the corpse has been murdered, and consider whether this hypothesis, or some other hypothesis, best explains the corpse. If the corpse and its surrounding conditions are such that they support the murder hypothesis, then we should say that the body has become a piece of evidence in favour of the existence of a murderer. Overall accepts the legitimacy of asking the question of whether the corpse is that of a murder victim. She fires back, however, that if one approaches the corpse knowing it’s the corpse of a murder victim, “then by virtue of [applying] the description, ‘homicide victim,’ [he is] (already) saying that there is a murderer, that a murderer exists”.Footnote 36 On the analogy, Larmer approaches the question of whether some events described in the Bible are best interpreted as miracles already knowing (or at least being convinced), by virtue of his Christianity, that they are. But this is not a problem for Larmer’s argument, only for Larmer’s own assessment of the merits of the theistic account of such events. Whether it is true that Larmer has, in his own reflection, decided that biblical stories are best interpreted as instances of religiously significant supernatural interventions that actually did happen, and whether he has done so by fairly considering all evidence in light of the assortment of contrary hypotheses, is irrelevant to the question of what taking the miracle hypothesis as the best hypothesis to account for a particular event would mean for our evidence for the existence of God if alternative naturalistic hypotheses were given a fair shake.
Larmer is not, however, totally correct about what the occurrence of a miracle means for the existence of God, notwithstanding the failure of Overall’s complaints of circularity. For even if an event occurs which is best explained as a miracle, this does not provide evidence that God, specifically the God of Christianity, exists. Larmer, at times, seems to think the actions of any transcendent agent meet the definition of a miracle, that the supernatural causal influence on the world need not solely be restricted to the Christian God.Footnote 37 At other times, however, as Overall points outFootnote 38, he moves easily from the idea of a supernatural agent exerting causal influence over the world to the idea of God exerting causal influence over the world.Footnote 39 In the places where God stands in for supernatural agent without argument, Larmer is reflecting his Western bias of equating supernatural agents with the Christian God. But this is indefensible given the definition of “miracle” with which he has seen fit to operate.
With this last piece of critical comment on the debate as it stands, let us pause to review. We have considered Overall’s argument for the claim that miracles constitute evidence against the existence of God. With the example of Mac and Lettie, I argued Overall’s argument fails because a miracle could occur which was not an instance of any of Overall’s three evils. I next briefly summarized Larmer’s argument that the occurrence of a miracle would constitute evidence for the existence of God. I argued, contra Overall, that Larmer does not beg the question with his argument that miracles can be evidence for the existence of God, but agreed with Overall’s criticism that the argument is not entitled to its conclusion that the occurrence of a miracle constitutes evidence for the existence of the Christian God. In making this last argument, I implicitly agreed with Overall that the relationship between any supernatural agent and the occurrences it causes must be critically assessed, rather than assumed.
Now that we have seen where the debate has been, we are in a position to see where it should go. Let us return to the example of Mac and Lettie. Mac and Lettie, good proto-scientists that they are, don’t just assume that the event Lettie thinks she has witnessed actually occurred. Surely there was something that happened, but perhaps Lettie hallucinated or was somehow otherwise misled. Let’s suppose that they investigate the evidence that the event only seemed to happen but didn’t, and come up empty-handed.
Satisfied that the event actually occurred, they seek an explanation for what has happened. What they are not entitled to do, of course, is straightaway conclude that there has been a miracle. The evidence, at this point, doesn’t warrant that conclusion.
Is there any situation that would warrant that conclusion? The proper course of action for our two natural scientists is, it would seem, to search for a cause within the sphere for the violation event. Suppose that they do this in earnest, but the investigation is fruitless.
At this point, I think they are faced with two explanatory options. The first of these is to keep the faith that the event was caused by a force within the sphere. On this option, the thing to do is shelve the anomaly as unexplained pending the results of future investigation and the resulting improved understanding of their world. This option is a measure of humility, a recognition that the end of science has not yet come. But it is coupled with a hope that the progress necessary to understand what has happened will occur soon. Perhaps it will even be necessary, once more is learned, to revise the laws of the sphere.
This first option seems to be the one most often taken when we come to scientifically anomalous events. As a case study, take the seemingly virgin birth of a hammerhead shark in a zoo in Omaha, Nebraska in 2001.Footnote 40 The most recent exposure of the three female hammerheads in the tank to male hammerheads was three years prior, when they were captured as immature animals less than one year old. Demian Chapman et. al. mention two possible explanations for the seemingly virgin birth drawn from existing knowledge of shark reproduction: sperm stored from a pre-capture sexual encounter with a male shark and intersexuality. The former explanation is judged unlikely given that the duration of sperm storage in hammerheads in the wild is five months and the animals were two years from sexual maturity at capture.Footnote 41 The latter explanation is eliminated by the lack of male sex organs on any of the adult hammerheads in the tank.Footnote 42
Chapman et. al. report that genetic analysis of the baby shark, tragically killed a few hours after its birth by a stingray in the same tank, “strongly supports the absence of paternal genetic contribution,” supporting the view, of course, that the baby shark was the product of asexual reproduction and not immaculate conception.Footnote 43 Prior to the present case, asexual reproduction had been documented in “all jawed vertebrate lineages (bony fishes, amphibians, reptiles and birds) except mammals and cartilaginous fishes”.Footnote 44 So in between the event in 2001 and Chapman et. al.’s genetic analysis published in 2007, no one really knew what had caused the birth of the pup. But references to virgin birth were tongue-in-cheek; no one really believed that the shark had been impregnated supernaturally. Chapman et. al.’s investigation is evidence that scientists did not give up the search for a naturalistic explanation. In other words, the anomalous birth was shelved as unexplained but not miraculous until the genetic evidence came to light in 2007 and a naturalistic explanation was supplied.
Is this the way every anomalous event should be handled? Imagine that the genetic test of the baby shark showed that there had been male parental genetic contribution. In such a case, what are we faced with in trying to explain what has happened? It seems as though the situation has become grimmer for a scientific explanation of the shark virgin birth. Perhaps we might wonder whether the female hammerheads truly were isolated. So we might try to find a genetic match amongst the males that have been known to be near their tank. We might investigate the possibility of secret artificial insemination. We might look still elsewhere in nature for an explanation, desperately searching for some natural hook on which to hang the event. Or we might begin to take seriously the idea of supernatural causation.
To return to the case of Lettie and Mac, they would be correct to do what we ordinarily do in anomalous situations as a first response. They should neither immediately conclude their laws of the sphere need revision (remember: these laws are well-confirmed) nor tag the event a miracle and move on, but instead investigate the world available to them more thoroughly. The more evidence accumulates that they were right about the laws of the sphere in the first place (including the law in conflict with the occurrence of the event in question), however, the better the case for opting for the second explanatory route, and chalking the event up to a non-sphere cause.Footnote 45
Some might doubt that this second route is even an explanation at all. We might think that to hang an event on God or any other supernatural entity is to hang it on one of Dan Dennett’s skyhooks, a convenient way of avoiding an explanation rather than offering one. This is the case only if the question is begged against supernatural agents in the first place, if it is settled and final that there are no supernatural agents before trying to explain some real event or another. But this is just as illegitimate as the far more commonly cited error on the part of the committed theist predisposed to leap to supernatural explanations.
So, in sum, Lettie and Mac, having failed to find evidence for the existence of a natural cause of their anomalous event, are faced with two options: shelve it for future naturalistic explanation or explain it in terms of supernatural intervention.Footnote 46 Since I think the question should not be begged for or against the existence of supernatural causes in the assessment of the evidence available, I agree with Larmer that it is possible to genuinely explain the event in terms of the actions of a supernatural agent. In other words, Lettie and Mac have it available to them to explain the event the former witnessed in terms of a supernatural agent. And if they do, then the evidence for the event becomes evidence in favour of the view that there is a supernatural being, viz. the one responsible for the event. In other words, if a miracle were to occur, and the evidence was such that those who learned of this miracle recognized it as such, then insofar as there is evidence for the explanation of the event as a miracle there is evidence for the existence of a supernatural agent.
But if Lettie and Mac opt for such an explanation they must ask, with Overall, which supernatural agent has worked this miracle? Is it the omniscient, omnipotent, and benevolent God of Christianity? Could an event like this spur one to belief in such a being? I do not think it is likely; the narrative according to which this event would be in accordance with such a divine plan is difficult to fathom.Footnote 47 A general problem with the inference from supernatural agent to Christian God is that the event, as described, admits of multiple interpretations, none with a clear leg up on the others. We have no better reason to think it’s the work of the God of Christianity than we do to think it’s the work of a less capable or less good god.
Contextual considerations and future happenings can help in the critical assessment of which of the range of available interpretations of the miracle-worker’s identity, not to mention moral character, beliefs, desires, and intentions, is correct. The arguments of this critical assessment will probably be of the weak inductive variety, especially if there is only one event which they think is best explained by the theistic hypothesis. But if there are more, Mac and Lettie may wish to make it the province of theologians to weave the miracles of their world into as coherent a whole as possible. The important point to notice is that a miracle, if it occurred, and were known to have occurred as such, would not just be evidence of the existence of at least one supernatural miracle-worker, but also evidence of the identity, moral character, beliefs, desires, and intentions of the miracle worker(s). And it is particularly not obvious that Lettie and Mac’s miracle is a work of the God of Christianity.
In this paper, I have tried to forge a middle path between Overall’s and Larmer’s respective accounts of the issue of whether the occurrence of a miracle would be evidence for or against the existence of God by identifying and synthesizing what is correct in their respective positive and negative accounts. If a miracle occurs, we don’t necessarily know it has occurred, and in the process of figuring out whether one has in fact occurred we can, with Larmer, gather evidence for the existence of a supernatural agent. But, if a miracle occurs, we can also ask, with Overall, what is the identity and character a supernatural agent would have to have to do such a thing.
In closing, it is important to make one final point about the debate between Overall and Larmer. I do not pretend to know whether Larmer’s and Overall’s respective views about the existence or non-existence of a supernatural agent or agents are defeasible for them. But I hope they are, for such beliefs are indeed the stakes of the debate over whether any event is best explained as a miracle. As such, approaching the evidence for or against the explanation of any event with an indefeasible allegiance to one or the other position on the existence of supernatural agents is, at best, inappropriate and, at worst, disingenuous. I’m not saying that everyone needs to agree about whether the evidence warrants the belief that any particular description of an event in the Bible, for instance, actually occurred and is best explained as a miracle, only that our debate over the quality of that evidence needs to be sincere and, in principle, capable of being resolved in either direction. To fail on this dimension, to approach the evidence for thinking that a miracle actually occurred with this or that position on the existence of supernatural agents indefeasibly entrenched, is to miss a large part of the wisdom of Hume’s essay on miracles. When confronted with a report of a miracle, wrote Hume, “I immediately consider with myself, whether it be more probable, that this person should either deceive or be deceived, or that the fact, which he relates, should really have happened. I weigh the one miracle against the other; and according to the superiority, which I discover, I pronounce my decision, and always reject the greater miracle”.Footnote 48
