Hume's Argument on Laws of Nature and Miracles, Study Guides, Projects, Research of History

Download Hume's Argument on Laws of Nature and Miracles and more Study Guides, Projects, Research History in PDF only on Docsity! 1 ON HUME’S PHILOSOPHICAL CASE AGAINST MIRACLES …For our sake he was crucified under Pontius Pilate; He suffered death and was buried. On the third day he rose again… —The Nicene Creed 1. Introduction As I understand the Christian faith, if certain events did not occur in human history, then many of its distinctive claims are false. What I have to say below concerns one such (alleged) event, an event affirmed in the earliest Christian creeds.1 It is known as “the Resurrection of Jesus Christ”. I choose this event as my focus not because I believe it occurred but because the great 18th century Scottish philosopher, David Hume contends that it did not occur, or at least that it is beyond credibility, and hundreds of intelligent and admirable men and women like him concur.2 In one of the most stirring and impassioned pieces of prose written in the English language, Hume summarizes his contention in these words: [U]pon the whole, we may conclude, that the Christian Religion not only was at first attended with miracles, but even at this day cannot be believed by any reasonable person without one. Mere reason is insufficient to convince us of its veracity: And whoever is moved by Faith to assent to it, is conscious of a continued miracle in his own person, which subverts all the principles of his understanding, and gives him a determination to believe what is most contrary to custom and experience. (131) As it turns out, Hume’s reasons for drawing this conclusion are perfectly general; if they apply to the Resurrection, they apply to any event that a religious adherent might regard as a miracle. (I invite the reader to substitute for the Resurrection whatever event they wish, so long as it is widely regarded as a miracle by adherents of a theistic religious tradition. The parting of the Sea of Reeds would be a good example.) Now, why should we care about why Hume thought that Jesus did not rise from the dead? Well, in a word, I take it that whether Hume is right is important. If Jesus did not rise from the dead, we can pretty much ignore his claims and the claims of his associates that he was the Son of God, the Messiah, Emmanuel; and along with them, we 1 From this point on, I will stop using the parenthetical qualifier “alleged”, but I want my readers to insert it whenever it might otherwise seem that I am affirming that some miracle occurred, or that some text accurately reported one. What I have to say here is perfectly at home with denying that any miracle has ever occurred. One of the earliest Christian creeds that affirms the Resurrection was cited by Paul in his first letter to the Corinthian church (circa 55AD): “For what I have received I passed on to you as of first importance: that Christ died for our sins…, that he was buried, that he was raised on the third day,…and that he appeared to Peter, and then to the Twelve. After that, he appeared to more than five hundred of the brothers at the same time…” (15:3ff). The bishops gathered at the first ecumenical council (Nicea, 325AD) reaffirmed this event in the words quoted at the beginning of this paper. 2 “Of Miracles,” section X, An Enquiry concerning Human Understanding, in Enquiries concerning Human Understanding and concerning the Principles of Morals, ed. L.A. Selby-Bigge, 2nd ed. (Oxford: Clarendon Press, 1902). I quote from the 3rd edition (1975), ed. P.H. Nidditch. Parenthetical page references in the body of my paper are to this work. 2 can pretty much dismiss the Christian faith too, or at least its distinctive doctrines.3 However, if Jesus did rise from the dead, then we would be most foolish to dismiss these claims. If Jesus was crucified, suffered death and was buried, and on the third day rose again, then, on the face of it, a supernatural explanation of the latter event is by far the likeliest, and his explanation of it becomes a live option, one that we would be most unwise simply to ignore. Here enters Hume. I would have thought that to investigate whether Jesus rose from the dead, we would need to do some historical research: we would need to assess the reliability of the New Testament documents and related manuscripts; we would need to look into the credibility of the witnesses to his post-mortem appearances, the empty tomb, and the like. The task looks rather daunting.4 But, according to Hume, we don’t have to do anything of the kind. For no matter how strong the testimony in favor of a miracle is—indeed, even if we “suppose…that the testimony considered apart and in itself, amounts to an entire proof” (114)—we have at our disposal a “full proof…against the existence of any miracle” (115). So we can avoid all that tedious historical work. We can simply use Hume’s shortcut, a proof against the existence of any miracle, and hence a proof against the Resurrection. So Hume has a “a decisive argument,” “an everlasting check to all kinds of superstitious delusion,” superstitious delusions like the Resurrection. There is good news and bad news in Hume’s proclamation. The good news is that he really has two arguments, not just one—or, at any rate, many scholars discern in his writings two arguments. The bad news is that neither succeeds; at any rate, try as I might, I can’t see how they do. 2. On the impossibility of miracles Everyone discerns in Hume’s writings the claim that there is good reason to believe that miracles claims are not credible, reason that is at least good enough to counterbalance or outweigh the strongest testimony for a miracle. Many discern a different claim, however, namely that miracles are absolutely impossible, impossible no matter what. It is this latter claim, and the argument for it, that I will assess first. Now, on the face of it, this claim is rather astonishing. While many of us may well think that Jesus did not in fact rise from the dead, or that the total available evidence for his having done so is 50-50, so to speak, none of us—I dare say—think that he could not have arisen from the dead no matter what. After all, “no matter what” includes “no matter if there is an omnipotent and omniscient God,” and it seems quite incredible that if there is such a God, he could not have miraculously raised Jesus from the dead if he had wanted to. But that’s what Hume says. 3 We are strongly encouraged to draw these inferences by the earliest Christians, e.g. Paul, in the aforementioned letter: “if Christ has not been raised, your faith is futile; you are still in your sins…. If only for this life we have hope in Christ, we are to be pitied more than all men” (15: 17,19). 4 I do not mean to imply that historical research of the sort envisaged here is the only way in which one could justifiably believe that Jesus rose from the dead. Perhaps one could justifiably believe that Jesus rose from the dead on the basis of the testimony of some authoritative community, or of the Holy Spirit, or a little of both, etc. All I mean to imply is that you would have thought that it was about as likely that we could arrive at a negative verdict on the Resurrection by a priori philosophizing of the sort Hume engages in as you would that we could arrive at a positive verdict by the same means—which is to say it isn’t likely at all. Nevertheless, that’s the strategy what Hume holds out to us. 5 and yet (b) the copper wire in my house was unable to conduct electricity on Thanksgiving Day, 2001. Things can’t take a holiday from the laws of nature. Now, what is a miracle? As Hume put it, “A miracle is a violation of the laws of nature” (114). But what is a violation of a law of nature but a counterinstance to it? Indeed, don’t those who use miracles as evidence for their beliefs rely on this fact? For example, the Resurrection is supposed to be weighty evidence for the existence of the supernatural, in general, or for Jesus’ religious authority in particular. But how could that be unless it’s a violation of a law of nature? If it were merely unusual for dead men to rise again, the Resurrection would not pack the wallop it is supposed to. Only if it is a violation of a law of nature—say, a violation of the generalization that all dead men stay dead—can the Resurrection be evidence for the supernatural or Jesus’ authority. We can summarize the two points of the last two paragraphs like this: 1. It is a necessary truth that, if a miracle m occurs, then there is a law of nature, L, such that m is a counterinstance to L. 2. It is a necessary truth that, if L is a law of nature, then there are no counterinstances to L. With these two points in hand, the argument for the impossibility of miracles is as simple as counting to five. Suppose, for reductio,9 that 3. It is possible that a miracle m occurs. It follows from 1 and 3 that 4. It is possible for there to be a law of nature, L, such that m is a counterinstance to L. But, on reflection, if 2 is true, then so is 5. It is impossible for there to be a law of nature, L, such that m is a counterinstance to L. And the conjunction of 4 and 5 is a contradiction. Thus, by reductio, 3 is false; and it doesn’t take much ingenuity to see that it follows God could not have miraculously raised Jesus from the dead. What should we make of this argument? We certainly cannot fault its inferences; the logic is impeccable. So if there is anything wrong with it, either premise 1 or 2 is false or undeserving of our assent. Let’s examine premise 2 first. 2.3 Must a law of nature have no counterinstances? To get at the question of whether premise 2 is true or false, we need to consider two very different pictures of what a law is, Prescriptivism and Descriptivism. In this section, I’ll briefly sketch the difference between these two pictures, then I’ll fill in the Prescriptivist picture with a bit more detail, and explain how it portrays the laws of nature. We will then be in a better position to assess premise 2. Two pictures of laws. The slogan for Prescriptivism reads, “Laws prescribe, not describe”; the slogan for Descriptivism reads, “Laws describe, not prescribe”. That is, 9 Reductio ad absurdum is a form of argument where you suppose something is true in order to show that it’s false—since an absurdity, like a contradiction, follows from the supposition. Any elementary logic textbook will explain it in more detail. 6 according to Prescriptivism, a law tells someone or something how to act; it prescribes behavior. “Do this…Don’t do that”—that’s what a law looks like, that’s the form that a law takes, according to Prescriptivism. According to Descriptivism, however, a law states how someone or something acts; it describes behavior. “It did this…It didn’t do that”— that’s what a law looks like, that’s the form a law takes, according to Descriptivism.10 So, in broad strokes, the two pictures come to this: SLOGAN FORM OF A LAW PRESCRIPTIVISM Laws prescribe, not describe “Do this” “Don’t do that” DESCRIPTIVISM Laws describe, not prescribe “It does this” “It doesn’t do that” As with most big pictures, the broad strokes of these two need filling in, and may even be somewhat misleading as they stand. This is especially the case for the picture of Prescriptivism. Prescriptivism. According to the prescriptivist slogan, a law prescribes, it does not describe. Strictly speaking, that slogan does not represent Prescriptivism accurately. While it’s accurate in saying that laws do not describe, it is inaccurate in saying that laws prescribe. That’s because, strictly speaking, only persons or groups of persons can prescribe, and no law is a person. Still, a law is a prescription, in the sense that it is the result or consequence or upshot of somebody with the relevant authority prescribing something. But here we must take care to understand what the prescriptivist means by the words “prescribe” and “prescription”. On the one hand, to prescribe something is to stipulate how things are to behave; it is to impose, order, or command something. When the state says, “Pay your taxes by April 15,” it prescribes certain behavior for taxpayers, behavior indicated in the resulting prescription, and the very act of the state’s prescribing it makes it a law. A prescription— that is, a law—is what results from the proper authority prescribing something. This is what the prescriptivist means by the words “prescribe” and “prescription”. On the other hand, to prescribe something is to recommend or advise it. This may well be what, nowadays, we mean typically by the words “prescribe” and “prescription”. But it is not what the prescriptivist means. When a doctor writes on a slip of paper, “Take 50 mg of Zoloft each night for 6 months,” and hands it to her patient, she advises him— very strongly, perhaps—to do something. But her prescribing it does not make it a law. A mere recommendation—no matter how firmly or wisely given—is not a law. We might express the difference this way. When a patient does not do what his doctor prescribes, does she have the authority to punish him? No. When a citizen does not do what the state 10 I’m oversimplifying here. Many descriptivists think that laws—and the laws of nature in particular— describe not merely what things do and don’t do, although such descriptions might be a consequence of the laws. But everything I say in the text about prescriptivism and descriptivism applies equally even if we were considering these other descriptivist views. 7 (properly) prescribes, does it have the authority to punish him? Yes. That’s the difference between a prescription that is a strong recommendation and a prescription that is a law. The state, unlike the doctor, has the authority to makes its prescriptions laws. So, according to the prescriptivist, although laws, strictly speaking, do not prescribe, they are prescriptions, but not recommendations, that is, they are the result of someone with the proper authority performing a certain sort of act, namely prescribing. We might pose a question for the prescriptivist at this juncture: why suppose that there is something that is distinct from the performance of a linguistic act (a particular occurrence of a prescribing), something that comes into existence as a result of performing that act? An authority prescribes; but there aren’t, in addition, prescriptions, the results of an authority’s prescribing. It seems to me that a start at answering this question would include the following thought. Suppose my wife and I enter the kitchen only to see standing on the counter reaching for the cookie jar our two-year old boys. (They’re doing a lot of that sort of thing these days.) We both say, simultaneously, “Get off the counter now!” There are clearly two particular occurrences of prescribing here—my wife’s, on the one hand, and mine, on the other—but, equally clearly, there is a sense in which my wife and I prescribed the same thing. What is that “same thing”? It is neither of our acts of prescribing, for they are two things, not one and the same thing. The prescriptivist says that it is a prescription; something distinct from the acts of prescribing but which results from such acts. Considerations such as these—and there are others, specifically those that parallel considerations offered on behalf of the theory of propositions—lend some support at least to the prescriptivist’s claim.11 Our excursion from the slogan “Laws prescribe” to the more accurate “Laws are prescriptions,” and our clarification of the meaning of “prescribe” and “prescription,” allows us to understand a point the prescriptivist wants to make, a point that is so important that it is best to regard it as an axiom of Prescriptivism: Axiom 1. There cannot be a prescription without a prescriber. Now, if you are at all like me, this hardly sounds surprising. After all, if no state had ever stipulated “Don’t drive on the left-hand side of the road, except to pass,” it would be silly to suppose that, nevertheless, there was a law that would be properly expressed by these words.12 We are now in a position to see a second important point. The state says “Pay your taxes by April 15”. Later, for one reason or another, it says “You don’t need to pay your taxes by April 15; pay them by April 30”. The original law is repealed, and a new 11 For considerations in defense of the theory of propositions, which are analogous to considerations in response to the question raised in this paragraph, see Michael Loux, Metaphysics (New York: Routledge, 1993), chapter 4. I’ve suggested that prescriptions come into existence as a result of the proper authority’s act of prescribing. But the prescriptivist might think otherwise. She might distinguish a prescription’s existence from its being in force or its being binding, say, in the way proponents of the theory of propositions distinguish a proposition’s existence from its being true. She might then say that those prescriptions that are in force or binding are just those that are expressed on certain occasions by an authority’s act of prescribing. 12 The prescriptivist described in the second half of note 11, would put the point of this paragraph like this: “…Axiom 1. Prescriptions cannot be in force without a prescriber…. After all, if no state had ever stipulated ‘Don’t drive on the left-hand side of the road, except to pass,’ it would be silly to suppose that, nevertheless, there is a law in force that would be properly expressed by these words.” 10 behave in a regular fashion, perhaps even an exceptionlessly regular fashion. But no such regularity—and no description of such a regularity, and no description of the structure of copper, and no description of the powers copper has in virtue of having that structure— should be confused with the natural law, Copper shall be conductive, which the law-giver lays down. A more accurate sketch of Prescriptivism can be expressed in the following table: SLOGAN FORM OF A LAW NATURAL LAWS AXIOMS Imperative: “Do this,” “Don’t do that” [To the void] Be conductive (in this sort of way, here and there) PRESCRIPTIVISM Laws are prescriptions or decrees, not descriptions Decree: “Let it be the case that…” “Fs shall be Gs” Let it be the case that copper is conductive Copper shall be conductive No prescription without a prescriber Prescribers and law-givers may repeal and replace as they please Now, our first question is this: if Prescriptivism is true, what does that imply for premise 2 of the Humean argument? Premise 2, recall, is this: 2. It is a necessary truth that, if L is a law of nature, then there are no counterinstances to L. Prescriptivism and counterinstances. To answer this question, let’s suppose that Prescriptivism is true, just to see what follows. And, for the sake of illustration, suppose that it is a law of nature that Copper shall be conductive. It follows from these two suppositions and Axiom 1 that there is a law-giver—say, God—who exercised his authority to bring about such states of affairs just by prescribing or decreeing them. Moreover, it follows from Axiom 2 that God can rescind and/or replace this law as he pleases. Suppose it suits his purposes on some occasion to do so. Suppose he decrees, “Let the copper in the Howard-Snyder home stop being conductive on Thanksgiving Day, 2001.” Well, that’s his prerogative, and if God decrees something, that’s how it is. Consequently, the copper in my house takes a holiday (unlike me!). We have here, then, an exception to the law Copper shall be conductive, a counterinstance to it. Here’s another illustration. Suppose, as David Hume did, that there is a law of nature that has to do with dead men staying dead. If Prescriptivism is true, this law will not, contrary to Hume, be an exceptionless regularity or a true universal generalization like All dead men stay dead. Rather, it will be something like Dead men shall stay dead, 11 where this is understood as the upshot of an authoritative act of decreeing, not a predicting or asserting. It follows from this supposition and Axiom 1 that there is a law- giver who exercised his authority to bring about such a state of affairs. And from this and Axiom 2 it follows that he can revoke his decree. Suppose his purposes call for such a revocation. Then, I suppose, he can bring it about that some dead man is no longer dead. If he does, the raised man is an exception to (what, at any rate, once was) the law. It seems we must conclude that if Prescriptivism is true, premise 2 of the Humean argument is false. Of course, the million dollar question is this: is Prescriptivism true? I suspect that I am like most of my readers. Prescriptivism holds a certain plausibility for us; it’s not an unnatural way to think about laws in general or natural laws in particular. Even so, we must investigate it more thoroughly before we endorse it.19 In the meantime, however, provided that nothing we reasonably believe rules it out, then, wouldn’t it be unwise to affirm that laws of nature can have no counterinstances? If so, we can’t affirm the impossibility of miracles—or the Resurrection—on the basis of the Humean argument. Let us turn now to an assessment of premise 1. In the remainder of section 2, I will assume—just for the sake of argument—that Prescriptivism is false and that Descriptivism is true. 2.4 Must a miracle be a counterinstance to a law of nature? Recall that premise 1 of the Humean argument says this: 1. It is a necessary truth that, if a miracle m occurs, then there is a law of nature, L, such that m is a counterinstance to L. We might adopt one of two strategies in assessing this premise. First, we might press on what a law of nature is, and discover that a miracle is not a counterinstance to a law of 19 Here’s one objection: the prescriptive force of an imperative or command cannot have the descriptive content that we take laws of nature to in fact have. See Jan Cover, “Miracles and (Christian) Theism,” in Philosophy of Religion: The Big Questions, eds. Eleonore Stump and Michael Murray (Malden, Massachusetts: Blackwell, 1999), 343. At first glance, this seems to be little more than the question- begging assertion that laws are descriptions. On closer inspection, however, perhaps Cover means to argue (1) laws have content, but (2) prescriptions don’t, so (3) laws aren’t prescriptions. The prescriptivist may well deny (1), but that would be desperate. After all, people obey prescriptions because they understand them, and they understand them because they have content. It seems to me that (2) is false, however. Here are two (incompatible) ways to argue for (2)’s falsity; I’m not sure which is correct, but I’m sure one of them is. Way 1. Prescriptions have prescriptive content. When I command my two-year old sons, “Stay off the street!,” the prescriptive content of my command is that They shall stay off the street (which, again, is neither a prediction nor a moral assertion). Way 2. Prescriptions have the same content as descriptions. When I command my sons, “Stay off the street!,” the content of my linguistic act is that They stay off the street. When I command my sons, I prescribe a certain state of affairs to them, Their staying off the street. Suppose they obey, and a passerby notices it. She might describe my sons behavior with “They stay off the street,” the content of her linguistic act is that They stay off the street. When she describes my sons, she describes a certain state of affairs, Their staying off the street. Perhaps Cover meant a different argument, say this: (1’) laws have descriptive content, but (2’) prescriptions don’t, so (3) laws aren’t prescriptions. The proponent of Way 1 will say (1’) is nothing but the implicit question-begging conjunction laws are descriptions and their content is descriptive. The proponent of Way 2 will, first, distinguish a speech act— like, prescribing and describing—from its content, and note that the content of two distinct speech acts can be the very same, as I illustrated with my command and the passerby’s description. She’ll then say that the phrase “descriptive content” is, strictly speaking, ill-formed. If it means “content of a description’, then (2’) is the false claim that prescriptions don’t have the content of a description. If it does not mean that, then (1’) is the question-begging assertion the proponent of Way 1 noticed. 12 nature, properly understood. Second, we might press on what a miracle is, and discover that a miracle, properly understood, need not be a counterinstance to a law of nature. I shall approach each strategy in the order mentioned. 2.4.1 Laws of nature, the goals of science, and implicit restrictions According to the Humean, laws of nature are (or are expressed by, or imply) true universal generalizations of the form All Fs are Gs.20 Now, suppose our best scientists developed the wisdom and technology to discover all the laws of nature, every last one of them—and suppose they succeeded. Then, according to the Humean, we would have a list of true universal generalizations, call it The List. Of course, many of the items on The List would be full of mathematical equations and technical jargon that very few people could understand. But many of them would be more homely, like this: 1. All copper is conductive. 2. Nothing travels faster than the speed of light. 3. All water freezes at temperatures below 32 degrees Fahrenheit. Let’s imagine, then, that we have The List before us. I want to make an observation about the particular items on it. Note that items 1 and 3 of the portion in view before us are about copper and water. Other items on The List are about more exotic things: muons, gluons, fields of force (perhaps). But what is 2 about? Certainly not nothing! Well, then, what is it about? Presumably, everything. Item 2 is equivalent to the statement “Everything travels no faster than the speed of light”. What can we infer from this? We can infer that no copper penny, no molecule of water, no photon, absolutely nothing, travels faster than the speed of light. Right? Well, not exactly. Let me explain. Suppose a note is taped to my office door when I arrive back from class: “Dan, Frances wants you to come home quickly.” I race home, run in the door, and shriek, “Is everything alright?” My wife responds, “Yes, everything’s fine; sorry to alarm you. Peter just got a bloody nose. I thought it was broken.” No one would mistake her for having claimed that—strictly speaking—everything is fine. There are still wars, famines, droughts, diseases, abused women and children, and so on. Everything—strictly speaking—is not fine. Or suppose you’re the janitor at a high school, and, after a basketball game, you clean up the mess in the gym. Your boss says to you the next day, “The gym sure was a mess last night. Did you clean up everything?” You answer, “Yep. It’s all gone; nothing’s left.” Obviously, you don’t mean the bleachers and backboards and banners are gone. You don’t even mean that there aren’t several scuff marks here or there, and a bit of dust in the corners. In everyday life, we invariably use the words “everything” and “nothing” with certain goals and restrictions in mind, indicated by contextual clues. Given the goal of getting the gym ready for the next game, none of the mess from the night before is left. Everything within the scope of my wife’s immediate concern is fine. The moral here is this: we need to understand a universal generalization as it is intended in the context in which it is put forward. So let’s return to The List and our hypothetical scientists who put it together. What do they intend by those universal generalizations? Well, what is the goal of science? Presumably, one of them is to understand how things in the natural world work. In that case, when they write down item 2 on The List—“Nothing travels faster than the 20 I’ll leave this parenthetical remark tacit from here on out, but I do mean it to be understood. 15 Some people say that if premise 1 is false, then miracles cannot be strong evidence for the supernatural, or the authority of some religious figure.27 To test this contention, let us imagine that we have just arrived at the scene of a horrifying traffic accident. A body has been separated into several, large pieces, strewn about the street. We get out of our car, careful to avoid an arm here, a leg there, and we check the inside of the smashed truck to see whether there are any survivors. There’s no one there. We turn back to our car and, as we walk toward it, the body parts begin to move together so that, in a minute or so, a woman arises from the ground, rubbing her head, but otherwise looking perfectly fine. Now, suppose that premise 1 is false. That is, suppose that it’s possible for a miracle to occur even though it is not a violation of a Humean law of nature. And suppose that, unbeknownst to us, the event we just witnessed is not such a violation, say, because long ago some dead man didn’t stay dead. Still, prior to witnessing what we just saw, the proposition that All dead men stay dead was exceedingly well-established for us by a large body of inductive evidence as being a law of nature, evidence that surpassed the most rigorous criteria for inductive evidence that you please. In those circumstances and given these suppositions, might not the event we just witnessed be strong evidence for some sort of supernatural intervention? I should hope so. So we’re left with a question: why should we believe that it is a necessary truth that a miracle is a violation or a counterinstance of a law of nature? Since we have no satisfactory answer to that question, we would be most unwise to affirm premise 1 of the Humean argument. Let’s now turn to Hume’s second argument against miracles, and the Resurrection in particular. 3. On the unreasonableness of believing the Resurrection occurred The most salient portions of Hume’s “proof against proof” argument against the Resurrection occur in these three passages: In order to encrease the probability against the testimony of witnesses, let us suppose, that the fact, which they affirm, instead of being only marvellous, is really miraculous; and suppose also, that the testimony considered apart and in itself, amounts to an entire proof; in that case, there is proof against proof, of which the strongest must prevail. (114) A miracle is a violation of the laws of nature; and as a 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. (114) 27 Antony Flew seems to be such a person. See, among other things, “Neo-Humean Arguments About the Miraculous,” In Defense of Miracles, eds. Doug Geivett and Gary Habermas (Downer’s Grove, Illinois: InterVarsity Pres, 1997), 46, where he says that in order for a miracle to serve as a “demonstration” that “the Christian revelation, or any rival candidate, constitutes an authentic self-revelation of the true God,” the word miracle has to be construed (as both Hume himself and all his contemporary opponents did construe it) in a very strong sense. It must involve an overriding of a law of nature, a doing of what is known to be naturally impossible by a Power which is, by this very overriding, shown to be supernatural. Only if this is given can the occurrence of a miracle under the auspices of some particular system of belief constitute an inexpungible divine endorsement of that system. 16 And as a uniform experience amounts to a proof, there is here a direct and full proof, from the very nature of the fact, against the existence of any miracle; nor can such a proof be destroyed, or the miracle rendered credible, but by an opposite proof, which is superior. (115) To get clear on Hume’s contention here, we need first to understand what Hume regarded as a “proof”. 3.1 What’s a “proof” according to Hume? Hume tells us that “we ought to divide arguments into demonstrations, proofs, and probabilities”.28 A demonstration is a deductively valid argument with known premises. A probability is a statistical induction with known premises. And a proof is a nonstatistical induction with known premises.29 Let me explain the latter a bit, as it is Hume’s main concern. A proof, according to Hume, is an argument that has the form of a nonstatistical induction. The form of a nonstatistical induction is this: 1. All hitherto observed (i.e. examined) As have been observed are Bs. 2. So, all As are Bs. In addition, the premise of every proof is known. So, for example, this is a proof, according to Hume: 1. All hitherto observed (i.e. examined) water has been composed of H20. 2. So, all water is composed of H20. It has the form of a nonstatistical induction, and its premise is known. There are three things to keep in mind about nonstatistical induction. First, we reason in this way frequently, and such reasoning sometimes provides very strong grounds for believing the conclusion to be true. Second, the premise of a nonstatistical induction is not intended to guarantee the truth of the conclusion; it is intended only to make the conclusion likely to one degree or another. So we can’t write off some particular nonstatistical induction by saying that even if all the As we’ve observed are Bs, some A that we haven’t observed might not be a B. That’s true but irrelevant to the strength of any such induction; for even the strongest one has that feature. Third, there are several standard criteria for assessing nonstatistical inductions. The two most important are sample size and sample representation. That is, the more As we’ve observed that are Bs, the more likely it is that all As are Bs, given the truth of the premise; and the more that the As we’ve observed differ in other respects, the more likely it is that all As are Bs, given the truth of the premise. Of course, the premise should be true as well. So, with these things in mind, what, exactly, is Hume’s argument? 3.2 Hume’s “proof against proof” argument Hume has in mind two proofs, one based on “testimony” in favor of a miracle and the Resurrection in particular, and the other based on “uniform experience” (115) against any miracle and the Resurrection in particular. At the most general level, the idea is that, no matter how strong the first proof is, the second cancels it out; “there is a mutual 28 Enquiry, 56, note 1. 29 See 110-11, “A wise man…observations,” and consult David Johnson, Hume, Holism, and Miracles, 11- 16, for textual evidence. Consult a logic textbook for more information about these kinds of arguments. 17 destruction of arguments” (116). Consequently, “no testimony is sufficient to establish a miracle” (115-6), and a miracle “cannot be believed by any reasonable person” (131). So it’s not reasonable to believe the Resurrection occurred.30 But how exactly do these proofs go? And how do they cancel each other out? And how are we supposed to think of testimony as a proof, anyway? Of course, testimony isn’t the sort of thing that can be a proof, in Hume’s sense of the term—a nonstatistical induction with a known premise. Suppose you tell me today that you turned 20 last week; naturally enough, I believe you. But how can we regard your testimony—that particular act at that particular place and time, or its content, that you turned 20 last week—as a nonstatistical induction with a known premise? We can’t. In all fairness, Hume does not ask us to consider a case in which testimony is a proof; he asks us to consider a case in which testimony “amounts to an entire proof” (114, my emphasis). But if we’re really going to have “proof against proof” so that “there is a mutual destruction of arguments,” we can’t have on the one side a proof and on the other side something that just “amounts” to a proof, something perhaps that isn’t even an argument at all. If we are going to take Hume seriously, we’re going to have to fill in the blanks for him. Here’s how to fill in the blanks.31 On the one side, the side in favor of a miracle— the Resurrection, say—we have the following argument, the first two premises of which constitute a proof, in Hume’s sense of the term: 1. All hitherto observed (that is, tested for accuracy) witnesses with such- and-such credentials32 are completely accurate reporters. 2. So, all witnesses with such-and-such credentials are completely accurate reporters. (from 1, nonstatistical induction) 3. Some witness with such-and-such credentials reported that Jesus died, and on the third day rose again. 4. So, Jesus did not stay dead. (from 2 and 3, deduction) On the other side, the side against the Resurrection, we have this argument, the first two premises of which constitute a proof, in Hume’s sense of the term: 5. All hitherto observed dead men have stayed dead. 6. So, all dead men stay dead. (from 5, nonstatistical induction) 7. So, Jesus stayed dead. (from 6, deduction) Hume’s contention is that the second proof cancels out the first. Thus, it is unreasonable to believe that the Resurrection occurred. Now, we might wonder why he affirms premise 5, which, as he sees things, describes “a firm and unalterable experience”. What he offers is this: “because that has never been observed in any age or country” (115). 30 Hume says that his argument has only to do with testimony considered as an “external evidence,” something publically available. Whether in jest or in earnest, I’m not sure, he explicitly states that his argument has no bearing on the (alleged) testimony of the Holy Spirit, whereby the truth of some proposition is “brought home…by the immediate operation of the Holy Spirit” (109). For extensive treatment of this idea, see Alvin Plantinga, Warranted Christian Belief (New York: Oxford University Press, 2000). 31 Here I rely on David Johnson, Hume, Holism, and Miracles, 17-18. 32 The reader is invited to fill in the “such-and-such” so that the credentials are as impressive as experience can make them: apparent fine character, seeming sincerity, an apparently unblemished track-record, and so on. 20 7 September 2001 Dear Sirs and Madams, Although we have closed the case of Mr. Manuel Kont, and although it is not our usual practice to respond to inquiries such as yours after a case has been closed, we feel that your kind-spirited letter, of 13 August 2002, is worthy of a response. Our response is this: the proof that you have provided is cancelled out by the following proof, which is based on a uniform experience. 5. All hitherto observed (that is, tested for accuracy) king salmon that have been caught weigh less than 178-pounds. 6. So, all king salmon that are caught weigh less that 178-pounds. (from 5, nonstatistical induction) 7. So, Mr. Manuel Kont did not catch a 178-pound king salmon, on 4 July 2001. (from 6, deduction) This full proof is founded on an infallible experience, and its premise—which, no doubt, you wonder why we affirm—describes a firm and unalterable experience, namely this: a 178-pound king salmon has never been observed in any age or country, or in any sea. We hope this settles the matter for you, as it should for any reasonable man. And we also hope you appreciate—perhaps more fully now—why the Guiness Foundation is respected round the globe for its insistence on the highest principles of reason and understanding. Sincerely, and with utmost best wishes, The Guiness Foundation The Bellingham City Council, upon further consultation, drafted the following letter, and sent it by registered mail on the date indicated: September 21, 2001 Dear Administrators, Thank you for your response of September 7, 2001, regarding the case of Mr. Manuel Kont. We realize (all too well, now) that you must receive countless complaints about your decisions, and we are most grateful that you found our query worthy of such a generous response. We wish only to raise three brief questions about it. First, might something in addition to the affidavits be relevant to your judgment? If so, perhaps the enclosed photos, local press clippings, and video of the event will be of use. Second, if your proof cancels ours, then no world-record can be reasonably believed to be broken—a rather odd consequence, don’t you think? Third, the premise of your proof states that “All hitherto observed (that is, tested for accuracy) king salmon that have been caught were less than 178- pounds.” With all due respect, we are perplexed by the reason you gave for this premise, namely this: “a 178-pound king salmon has never been observed in any age or country, or in any sea.” For when you say “a 178-pound king salmon has never been observed in any age or country, or in any sea,” either you mean to include the observations of the 512-plus, or you don’t. If you do, then doesn’t 21 your reason presuppose the denial of what is precisely at issue, namely whether the 512-plus really did witness Mr. Kont’s 178-pound king salmon—in which case your argument begs the question? (Of course, if you have some independent reason to deny the report of the 512-plus, we would be most anxious to hear it; but, as it stands, you profess your proof cancels ours without considering any other independent reason.) If you do not, however, why not? After all, you claim to have given us a nonstatistical induction with a known premise. But you know the premise only if you have taken into consideration all the information available to you, including the affidavits and supporting material from us (which, we remind you, you have said “amounts to a proof”). And if you take it into consideration, wouldn’t you agree—at the very least—that it is not at all clear that “a 178-pound king salmon has never been observed in any age or country, or in any sea”? Consequently, wouldn’t you agree that you don’t know your premise, after all? If the Foundation would be so kind as to clarify these matters for us, we would be in their debt. Sincerely, The Bellingham City Council As of November 22—two months after this letter was sent—the Council has yet to receive a response, despite several queries and phone calls.33 Daniel Howard-Snyder Department of Philosophy Western Washington University http://www.ac.wwu.edu/~howardd 33 I leave as homework to the reader the application of this story to Hume’s “proof against proof” argument for the unreasonableness of believing in the Resurrection. Help can be found in Cover, “Miracles and (Christian) Theism” (see note 19). For an extensive bibliography on contemporary philosophical literature on miracles, see Michael Levine, “Miracles,” in the Stanford Encyclopedia of Philosophy, on-line at http:/plato.stanford.edu. Also, see David Johnson’s Hume, Holism, and Miracles (note 19)—upon which I have relied heavily—and John Earman’s Hume’s Abject Failure (New York: Oxford University Press, 2000). Thanks to Jan Cover, Frances Howard-Snyder, and Hud Hudson for comments on earlier drafts of this paper.