Hume’s philosophy of religion is today regarded by many naturalists as providing a decisive refutation of the natural or philosophical theology found in the medieval philosophers as well as the modern philosophers who are Hume’s predecessors. There is no point at which Hume’s views are more at variance with those of Locke, Berkeley, Descartes, and Leibniz than in his account of religious belief. All of those philosophers claimed that one could demonstrate God’s existence with certainty. Even Spinoza, who clearly rejects the God of theism, offers a rational theology in which the reality of God is of the first importance.
Hume’s most influential writings on the philosophy of religion are found in his Dialogues Concerning Natural Religion, which was published posthumously. As the title implies, the book is in dialogue form, with three characters: Demea, Cleanthes, and Philo. Demea is a pious figure who relies on what he calls a priori arguments for God’s reality. However, the arguments give us no clear conception of what God is like, a conclusion Demea welcomes, since he affirms that God is beyond human understanding. Demea gets relatively little space in the book, and this is understandable, since his rationalistic approach to theology has little appeal to an empiricist such as Hume. Each of the other characters appeals to Demea at times, but he seems mostly intended to exhibit the weakness of nonempirical arguments for God.
Most of the book thus consists of an extended argument between Philo and Cleanthes. Cleanthes bases rational belief in God on a kind of argument from design, in which God must exist as the designer of the natural universe, which shows clear evidence of purposive intelligence. His argument at least has empirical foundations and thus engages Hume’s attention. Cleanthes puts his design argument in the form of an argument from analogy. We observe many things, such as machines of various types, which show the purposive adaptation of means to ends. We know that such things are the result of intelligence. If the universe is analogous to these designed objects, then its cause must be analogous to the causes of those designed objects. Thus the cause of the universe must be a powerful intelligent being. It is important that Cleanthes’s version of the argument rests on an analogy between the universe as a whole and human-made artificial objects. Other versions of the argument from design, such as Aquinas’s Fifth Way, reason directly from the existence of particular objects in nature that show evidence of purposiveness.27
Although Philo claims to be a truly religious person, he is skeptical about Cleanthes’s argument for a number of reasons. The roots of his skepticism are clearly grounded in something like Hume’s own view of causation. Furthermore, Philo seems to get the better of the argument throughout most of the book. Many commentators have therefore viewed Philo as representing Hume’s own views, at least throughout much of the book. Although not everyone agrees with this identification of Hume and Philo, when people speak of “Hume’s” criticisms of the theistic arguments, they usually have in mind the arguments of Philo.
We can see why a priori arguments or proofs would be suspicious on Hume’s view, since such arguments can only be rightly given for “relations of ideas.” However, the existence of God, like any existence claim, will be a “matter of fact” for Hume and thus must be known by experience. A priori arguments will thus be of little value, which explains why much of the book focuses on Cleanthes’s more empirical argument.
Many of the criticisms of this argument rest on Hume’s view of causality, which affirms that causal knowledge must be grounded in experience. Since we have no experience of universe-making or creating, it is clear that claims about the cause of the existence of the universe will be shaky. Cleanthes, as noted above, tries to avoid this problem by appeal to an analogy between the universe and human-made objects, but Philo sees this move as problematic as well.
First of all, Philo claims that the analogy is weak. The universe is vastly larger and more complex than a watch or a pump, and it is difficult to know how similar the universe is to such objects.28 Second, there are many alternative analogies that would imply a different kind of cause.29 If we think of the universe as similar to a pump, then the cause might be intelligent, but if we think of the universe as similar to a vegetable or animal, then it might come from a seed or hatch from an egg.30 (Philo recognizes how absurd the latter seems, but he argues that we have no real way of ruling out even views that seem absurd.) Third, Philo claims that for all we know, the material universe itself might contain some kind of causal principles sufficient to produce the universe we see.31 Hume is writing before Darwin, but he here seems to hope that eventually we would find a natural explanation for the apparent design we see in the world.
Finally, even if we concede that the cause of the universe is probably intelligent, Philo claims the conclusion is religiously inadequate. We have no reason to believe the intelligent cause would be the God that Christians and Jews believe in. We would have no reason to believe the designer was infinite in either power or goodness.32 Perhaps the universe was made by a junior deity still learning how to make good universes, or by an older deity whose faculties are in decline. In fact, there would be no reason to believe in just one being as the creator. The empirical facts are consistent with a committee of gods being the designer. Philo, by the way, as we will see below, himself does not see these objections as completely eliminating the force of the argument from design.
Besides attacking Cleanthes’s version of the argument from design, Philo goes on the offensive by raising the problem of evil, affirming that “Epicurus’s old questions are yet unanswered. Is he [God] willing to prevent evil, but not able? Then he is impotent. Is he able, but not willing? Then he is malevolent. Is he both able and willing? Whence then is evil?”33 Philo is scrupulous in how he raises the problem, however. He does not argue that evil provides a proof of God’s nonexistence. It is possible that a good God might have reasons to allow evil, even if we do not know what they are. However, the existence of evil at least poses a severe problem for natural theology, since it shows that the conception of God we would form if we based that conception solely on what we observe in the world would not be a conception of a God who is both completely good and all-powerful.
The theist who is a Christian might well concede Philo’s point here. If all we knew about God was formed from observing the natural world, we might lack grounds for affirming with any confidence that God is completely good. For example, if we look at primal or tribal religions, many of them have posited gods who were morally ambiguous, beings that humans are rightly afraid of and seek to placate. Even if it were reasonable to hold that there must be a transcendent being who is the cause of the universe, Christians might insist that we can only know the nature of that God through God’s own self-revelation.
Interestingly, in the Dialogue this seems close to the position that Philo himself ultimately takes. In the end Philo concedes to Cleanthes that there is at least some weak probability that the world was created by intelligence, but he denies that this conclusion has much if any religious significance. Philo concedes that the most plausible view of the world is “that the cause or causes of order in the universe probably bear some remote analogy to human intelligence.”34 That is a conclusion that seems religiously inadequate of course. Philo says that we will likely feel some “astonishment” at the greatness of the object, but also some “melancholy” from its obscurity, along with “contempt” for human reason’s inadequacy to tell us more. However, Philo says that “the most natural sentiment, which a well-disposed mind will feel . . . is a longing desire and expectation, that Heaven would be pleased to dissipate, at least alleviate, this profound ignorance, by affording some more particular revelation to mankind.”35 Some commentators have argued that this conclusion is not likely to be sincere but reflects Hume’s fear of the consequences of his supposed atheism. However, whether Philo (or Hume) believes this or not, it seems plausible from a Christian point of view. Perhaps natural theology cannot give us adequate knowledge of God, even if it can give us reasons to believe there is a God.36
Hume’s Argument Against Belief in Miracles
Besides the discussion of natural theology in his Dialogues, Hume is also famous for an argument contained in the Enquiry Concerning Human Understanding against belief in miracles. Hume actually gives two different types of arguments in this essay. First, he gives a general argument that even the best evidence we could imagine would not be sufficient to make belief in miracles reasonable. In the last part of the essay he tries to show that the actual evidence we have for miracles is weak and comes nowhere near the standard that would be necessary. The arguments he gives in the second part are more historical than philosophical. For example, he claims that the testimony comes from “ignorant and barbarous tribes” and always from far-off times and places. These arguments are really historical and not philosophical, and they are hardly decisive. They not only assume an arrogant and condescending view of ancient peoples; they also fail to recognize that there are plenty of people in the contemporary world who testify that miracles have occurred.37
In both cases it should be obvious that Hume’s arguments are epistemological, not metaphysical. Given his skepticism about metaphysics, it is easy to see why he takes this tack, since if we lack metaphysical knowledge it is hard to see how one could show that miracles cannot happen.
Hume generally assumes that belief in miracles is based on testimonial evidence, and also that testimonial evidence itself requires support. Both of these assumptions can be challenged. Many people claim to have directly observed a miracle or the outcome of a miracle. Furthermore, as we shall see in the next chapter, Reid argues that testimony is a source of basic evidence and does not always require further evidence in order for it to justify belief.
Hume says that when we receive testimonial evidence, one of the ways we evaluate that evidence is by looking at the probability of the content of the evidence. If one of my students tells me that there are cars driving on the streets of Waco, Texas, I will have little reason to doubt this, since the story is highly probable. However, if one of my students tells me that Martians are walking around the Baylor University campus, the story is so improbable that I am unlikely to believe it without strong supporting evidence. So how probable are the accounts given by witnesses who claim that a miracle has occurred?
Hume says that the probability of a type of event is a function of the frequency with which events of that type have occurred in our past experience. He then defines a miracle as a “violation of a law of nature” that is caused by God or some other invisible agent.38 Since the laws of nature simply describe what we have always observed to happen, if a miracle occurred it would be, by definition, an extremely rare event, an exception to our general experience. It follows from this that the probability of a miracle occurring is maximally low, and so it looks as if it would always be more probable that the witness who affirms a miracle has occurred is either mistaken or lying than that the miracle has actually occurred. To make belief in a miracle reasonable, the falsity of the testimony in favor of the miracle would itself have to be miraculous.39 Obviously, this is a very high bar to meet.
Some critics of Hume have fastened on Hume’s definition of a miracle as a “violation” of a law of nature. Laws of nature describe what normally happens, but the laws have the form of “If x, then y.” If something different from x occurs, then there is no violation of the law if y does not occur. Since God’s actions in bringing about a miracle represent an additional factor not normally present, it can be plausibly argued that the law of nature is not violated. The law of nature simply states what would normally happen in the course of nature if God did not act in an unusual way. However, I do not think this objection is a serious problem for Hume. He could easily restate his argument by dropping the language of “violation” and simply talking about an exception to what we normally observe in nature. There is surely something exceptional about miracles in the sense that Hume means to discuss them. People do not normally rise from the dead, and humans cannot normally walk on water. Recognizing the exceptional character of miracles does not mean that some kind of deistic view of God is being assumed, either. The person who affirms that a miracle has occurred does not have to say that God is active in nature when he normally is not active, but simply that when God acts miraculously in nature, he is acting in a special way, different from his activity in holding nature in being.
Other critics have claimed that Hume begs the question by saying that the experience that supports laws of nature is “inviolable.” They say that Hume assumes that miracles never happen (or cannot ever happen) and then concludes that miracles should not be believed. However, though Hume states his argument in a mistaken way at this point, it again looks like the error is one that is easily remedied. Instead of saying that the laws of nature are “inviolable” and describe what always happens, he should have said that the laws of nature describe what generally happens except when miracles occur, and miracles are rare events at best. His argument could still go through on this basis if it were sound otherwise.
However, I do not believe the argument is sound. As I have already mentioned, it does not seem correct that belief in miracles is based solely on testimony. There is also physical evidence and direct experience for some people to support such beliefs. With respect to testimonial evidence it seems to me that Hume’s big mistake is to assume that we estimate the probability of a type of event solely on the basis of the frequency with which events of that type occur. Suppose that a giant meteor has hit the Earth only once in all the billions of years of its history. On Hume’s view this would imply that it is extremely unlikely that a meteor will ever hit the Earth again. If scientists claim that a meteor is coming our way, according to this account of probability, it would be more reasonable to believe that they are mistaken, since the event they predict is so improbable. However, it would be foolish to take such a view of the scientists’ claims.
What has gone wrong in the meteor case? The mistake is to estimate probability solely on the basis of the frequency with which a type of event occurs. In reality we base our judgments on how probable a type of event is on everything we know that might bear on the question. In the case of the meteor we would look at how many meteors there are, and more particularly on the observations that scientists have made about the one they believe will hit the Earth, including its position and direction and velocity.
In a similar way, the probability of a miracle should not be estimated solely on the basis of the frequency with which miracles occur. (In fact, Hume may be wrong in claiming that miracles are extremely rare; many people claim to have personal experience of miracles.) How likely it is that a miracle may occur will depend on many things: Is there a God? If there is a God, would God be likely to have reasons to perform miracles? The answers to these questions might well show that, even if a miracle at any given moment may be improbable, the likelihood of a miracle’s occurring sometime might be much higher. Furthermore, in certain circumstances, in which it seems that a miracle would further God’s purposes in some essential way, a miracle might not be so improbable that a credible witness could not reasonably produce belief.
Hume’s Ethics
The longest section of the Treatise deals with ethical questions, and it is likely Hume thought that his contributions to ethics were among his most important achievements; he affirms that “morality is a subject that interests us above all others.”40 Historically, Hume’s ethical thought deeply influenced utilitarianism, through Jeremy Bentham. In the contemporary period Hume is recognized as the ancestor of “antirealist” metaethical views.
A realist account of ethics is one that sees ethical judgments as objectively true or false, independently of what humans think or feel. “Murder is morally wrong” is a true proposition, one that states a moral fact, one that would be true even if humans became so morally depraved as to value murder and view it as good. Antirealism in metaethics rejects the idea that moral judgments are true or false independently of human beliefs, attitude, and emotions. Although the term antirealism is a contemporary one that Hume does not use, he clearly holds such a view. Besides Hume’s influence on contemporary antirealism, he also has inspired a contemporary revival of what is sometimes called “sentimentalism,” which can be understood in a more realist manner.
Even if Hume is an antirealist, he is far from being a moral nihilist. He does not want to deny that we can justifiably make moral judgments, and he even affirms that some of them are true and some of them are false. However, he wants to claim that these judgments are not stating matters of fact that are true independently of us. Rather, ethical or moral truths are expressions of our attitudes or feelings. In this respect they resemble beauty. When geometers describe the qualities of a circle they do not include beauty among them. The reason is that “beauty is not a quality of the circle” but “is only the effect which that figure produces upon the mind.”41 Moral judgments then reflect the emotional responses we have to facts; they do not express facts that hold independently of those responses.
Morality then, according to Hume, does not consist of “matters of fact” that can be discovered by reason or the understanding. One argument he gives in favor of this view is that if moral judgments were simply objective judgments made by reason, then there would be no inherent connection between a moral judgment and motivation. In Hume’s psychology, a person is never moved by reason alone to do anything. Rather, “reason is and ought to be the slave of the passions.”42 Therefore, if moral judgments are linked to our behavior, so that a moral judgment is linked to some tendency to behave in a certain way, this shows that these judgments are not about “matters of fact” discovered by reason, but rather are grounded in the passions. If there is a “matter of fact” that grounds moral judgments, it lies in the facts about our responses to actions. Some actions (and states of character) we approve of and call moral. Others we disapprove of and call immoral. But actions do not possess the character of being moral or immoral independently of ourselves.
Despite this subjective basis for ethics, Hume believes that the substantive ethical judgments that are part of common sense are justified and have a kind of objectivity. The basis for this objectivity lies in human nature. The passions and sentiments of humans are similar enough that we can reach a wide measure of agreement about what is good and bad, what is a virtue and what is a vice. The moral sentiments are basically common to humans, and thus we admire moral qualities, such as generosity and bravery, even when they are found in our enemies, and we disapprove of immoral qualities, such as greediness and ingratitude, even when found among our friends and family. Morality is not founded simply in self-interest, but in sympathetic feelings we have toward others, feelings that induce us to approve of actions that are “helpful or useful” to the person acting or to others.
Many of our moral judgments are then not really claims about moral facts as they appear to be but expressions of how we feel about various moral situations. Hume does recognize that some of our moral rules are founded on agreements or conventions we make with each other, the basis of the agreements being the “utility” those moral rules have for us. This part of Hume’s ethic was later influential on Jeremy Bentham, the founder of utilitarianism.
Hume may be quite right to insist that moral truths are recognized through the emotions. It may be that the anger I feel when I see an act of injustice or the revulsion I feel when I see an act of cruelty are the means whereby I come to recognize that injustice is wrong and cruelty is evil. However, it does not follow from this that these emotions are what grounds the truth of our moral judgments. Rather, if we are indeed moral beings and morality is part of our nature, then our emotions may be capacities for grasping moral truths. Hume’s own epistemology cannot allow for this, since we can only come to know facts for him through sense experiences. However, emotions may also be a kind of experience, and like sense experiences, emotional experiences may provide us with information about the world. The problem with Hume’s subjectivism is that it implies that good and evil would change if our emotional reactions changed. However, even if we humans became so morally depraved that we approved of rape and murder, it would not follow that rape and murder would be right, but that we humans had lost our capacity to grasp and be motivated by moral truths.
Hume is clearly trying to develop an account of morality that is independent of God and any kind of transcendent order. However, many philosophers worry that human emotions are too frail a foundation for ethical truths, even if they are one of the means whereby we come to grasp those truths.

Question

Hume’s philosophy of religion is today regarded by many naturalists as providing a decisive refutation of the natural or philosophical theology found in the medieval philosophers as well as the modern philosophers who are Hume’s predecessors. There is no point at which Hume’s views are more at variance with those of Locke, Berkeley, Descartes, and Leibniz than in his account of religious belief. All of those philosophers claimed that one could demonstrate God’s existence with certainty. Even Spinoza, who clearly rejects the God of theism, offers a rational theology in which the reality of God is of the first importance.
			Hume’s most influential writings on the philosophy of religion are found in his Dialogues Concerning Natural Religion, which was published posthumously. As the title implies, the book is in dialogue form, with three characters: Demea, Cleanthes, and Philo. Demea is a pious figure who relies on what he calls a priori arguments for God’s reality. However, the arguments give us no clear conception of what God is like, a conclusion Demea welcomes, since he affirms that God is beyond human understanding. Demea gets relatively little space in the book, and this is understandable, since his rationalistic approach to theology has little appeal to an empiricist such as Hume. Each of the other characters appeals to Demea at times, but he seems mostly intended to exhibit the weakness of nonempirical arguments for God.
			Most of the book thus consists of an extended argument between Philo and Cleanthes. Cleanthes bases rational belief in God on a kind of argument from design, in which God must exist as the designer of the natural universe, which shows clear evidence of purposive intelligence. His argument at least has empirical foundations and thus engages Hume’s attention. Cleanthes puts his design argument in the form of an argument from analogy. We observe many things, such as machines of various types, which show the purposive adaptation of means to ends. We know that such things are the result of intelligence. If the universe is analogous to these designed objects, then its cause must be analogous to the causes of those designed objects. Thus the cause of the universe must be a powerful intelligent being. It is important that Cleanthes’s version of the argument rests on an analogy between the universe as a whole and human-made artificial objects. Other versions of the argument from design, such as Aquinas’s Fifth Way, reason directly from the existence of particular objects in nature that show evidence of purposiveness.27
			Although Philo claims to be a truly religious person, he is skeptical about Cleanthes’s argument for a number of reasons. The roots of his skepticism are clearly grounded in something like Hume’s own view of causation. Furthermore, Philo seems to get the better of the argument throughout most of the book. Many commentators have therefore viewed Philo as representing Hume’s own views, at least throughout much of the book. Although not everyone agrees with this identification of Hume and Philo, when people speak of “Hume’s” criticisms of the theistic arguments, they usually have in mind the arguments of Philo.
			We can see why a priori arguments or proofs would be suspicious on Hume’s view, since such arguments can only be rightly given for “relations of ideas.” However, the existence of God, like any existence claim, will be a “matter of fact” for Hume and thus must be known by experience. A priori arguments will thus be of little value, which explains why much of the book focuses on Cleanthes’s more empirical argument.
			Many of the criticisms of this argument rest on Hume’s view of causality, which affirms that causal knowledge must be grounded in experience. Since we have no experience of universe-making or creating, it is clear that claims about the cause of the existence of the universe will be shaky. Cleanthes, as noted above, tries to avoid this problem by appeal to an analogy between the universe and human-made objects, but Philo sees this move as problematic as well.
			First of all, Philo claims that the analogy is weak. The universe is vastly larger and more complex than a watch or a pump, and it is difficult to know how similar the universe is to such objects.28 Second, there are many alternative analogies that would imply a different kind of cause.29 If we think of the universe as similar to a pump, then the cause might be intelligent, but if we think of the universe as similar to a vegetable or animal, then it might come from a seed or hatch from an egg.30 (Philo recognizes how absurd the latter seems, but he argues that we have no real way of ruling out even views that seem absurd.) Third, Philo claims that for all we know, the material universe itself might contain some kind of causal principles sufficient to produce the universe we see.31 Hume is writing before Darwin, but he here seems to hope that eventually we would find a natural explanation for the apparent design we see in the world.
			Finally, even if we concede that the cause of the universe is probably intelligent, Philo claims the conclusion is religiously inadequate. We have no reason to believe the intelligent cause would be the God that Christians and Jews believe in. We would have no reason to believe the designer was infinite in either power or goodness.32 Perhaps the universe was made by a junior deity still learning how to make good universes, or by an older deity whose faculties are in decline. In fact, there would be no reason to believe in just one being as the creator. The empirical facts are consistent with a committee of gods being the designer. Philo, by the way, as we will see below, himself does not see these objections as completely eliminating the force of the argument from design.
			Besides attacking Cleanthes’s version of the argument from design, Philo goes on the offensive by raising the problem of evil, affirming that “Epicurus’s old questions are yet unanswered. Is he [God] willing to prevent evil, but not able? Then he is impotent. Is he able, but not willing? Then he is malevolent. Is he both able and willing? Whence then is evil?”33 Philo is scrupulous in how he raises the problem, however. He does not argue that evil provides a proof of God’s nonexistence. It is possible that a good God might have reasons to allow evil, even if we do not know what they are. However, the existence of evil at least poses a severe problem for natural theology, since it shows that the conception of God we would form if we based that conception solely on what we observe in the world would not be a conception of a God who is both completely good and all-powerful.
			The theist who is a Christian might well concede Philo’s point here. If all we knew about God was formed from observing the natural world, we might lack grounds for affirming with any confidence that God is completely good. For example, if we look at primal or tribal religions, many of them have posited gods who were morally ambiguous, beings that humans are rightly afraid of and seek to placate. Even if it were reasonable to hold that there must be a transcendent being who is the cause of the universe, Christians might insist that we can only know the nature of that God through God’s own self-revelation.
			Interestingly, in the Dialogue this seems close to the position that Philo himself ultimately takes. In the end Philo concedes to Cleanthes that there is at least some weak probability that the world was created by intelligence, but he denies that this conclusion has much if any religious significance. Philo concedes that the most plausible view of the world is “that the cause or causes of order in the universe probably bear some remote analogy to human intelligence.”34 That is a conclusion that seems religiously inadequate of course. Philo says that we will likely feel some “astonishment” at the greatness of the object, but also some “melancholy” from its obscurity, along with “contempt” for human reason’s inadequacy to tell us more. However, Philo says that “the most natural sentiment, which a well-disposed mind will feel . . . is a longing desire and expectation, that Heaven would be pleased to dissipate, at least alleviate, this profound ignorance, by affording some more particular revelation to mankind.”35 Some commentators have argued that this conclusion is not likely to be sincere but reflects Hume’s fear of the consequences of his supposed atheism. However, whether Philo (or Hume) believes this or not, it seems plausible from a Christian point of view. Perhaps natural theology cannot give us adequate knowledge of God, even if it can give us reasons to believe there is a God.36
			Hume’s Argument Against Belief in Miracles
			Besides the discussion of natural theology in his Dialogues, Hume is also famous for an argument contained in the Enquiry Concerning Human Understanding against belief in miracles. Hume actually gives two different types of arguments in this essay. First, he gives a general argument that even the best evidence we could imagine would not be sufficient to make belief in miracles reasonable. In the last part of the essay he tries to show that the actual evidence we have for miracles is weak and comes nowhere near the standard that would be necessary. The arguments he gives in the second part are more historical than philosophical. For example, he claims that the testimony comes from “ignorant and barbarous tribes” and always from far-off times and places. These arguments are really historical and not philosophical, and they are hardly decisive. They not only assume an arrogant and condescending view of ancient peoples; they also fail to recognize that there are plenty of people in the contemporary world who testify that miracles have occurred.37
			In both cases it should be obvious that Hume’s arguments are epistemological, not metaphysical. Given his skepticism about metaphysics, it is easy to see why he takes this tack, since if we lack metaphysical knowledge it is hard to see how one could show that miracles cannot happen.
			Hume generally assumes that belief in miracles is based on testimonial evidence, and also that testimonial evidence itself requires support. Both of these assumptions can be challenged. Many people claim to have directly observed a miracle or the outcome of a miracle. Furthermore, as we shall see in the next chapter, Reid argues that testimony is a source of basic evidence and does not always require further evidence in order for it to justify belief.
			Hume says that when we receive testimonial evidence, one of the ways we evaluate that evidence is by looking at the probability of the content of the evidence. If one of my students tells me that there are cars driving on the streets of Waco, Texas, I will have little reason to doubt this, since the story is highly probable. However, if one of my students tells me that Martians are walking around the Baylor University campus, the story is so improbable that I am unlikely to believe it without strong supporting evidence. So how probable are the accounts given by witnesses who claim that a miracle has occurred?
			Hume says that the probability of a type of event is a function of the frequency with which events of that type have occurred in our past experience. He then defines a miracle as a “violation of a law of nature” that is caused by God or some other invisible agent.38 Since the laws of nature simply describe what we have always observed to happen, if a miracle occurred it would be, by definition, an extremely rare event, an exception to our general experience. It follows from this that the probability of a miracle occurring is maximally low, and so it looks as if it would always be more probable that the witness who affirms a miracle has occurred is either mistaken or lying than that the miracle has actually occurred. To make belief in a miracle reasonable, the falsity of the testimony in favor of the miracle would itself have to be miraculous.39 Obviously, this is a very high bar to meet.
			Some critics of Hume have fastened on Hume’s definition of a miracle as a “violation” of a law of nature. Laws of nature describe what normally happens, but the laws have the form of “If x, then y.” If something different from x occurs, then there is no violation of the law if y does not occur. Since God’s actions in bringing about a miracle represent an additional factor not normally present, it can be plausibly argued that the law of nature is not violated. The law of nature simply states what would normally happen in the course of nature if God did not act in an unusual way. However, I do not think this objection is a serious problem for Hume. He could easily restate his argument by dropping the language of “violation” and simply talking about an exception to what we normally observe in nature. There is surely something exceptional about miracles in the sense that Hume means to discuss them. People do not normally rise from the dead, and humans cannot normally walk on water. Recognizing the exceptional character of miracles does not mean that some kind of deistic view of God is being assumed, either. The person who affirms that a miracle has occurred does not have to say that God is active in nature when he normally is not active, but simply that when God acts miraculously in nature, he is acting in a special way, different from his activity in holding nature in being.
			Other critics have claimed that Hume begs the question by saying that the experience that supports laws of nature is “inviolable.” They say that Hume assumes that miracles never happen (or cannot ever happen) and then concludes that miracles should not be believed. However, though Hume states his argument in a mistaken way at this point, it again looks like the error is one that is easily remedied. Instead of saying that the laws of nature are “inviolable” and describe what always happens, he should have said that the laws of nature describe what generally happens except when miracles occur, and miracles are rare events at best. His argument could still go through on this basis if it were sound otherwise.
			However, I do not believe the argument is sound. As I have already mentioned, it does not seem correct that belief in miracles is based solely on testimony. There is also physical evidence and direct experience for some people to support such beliefs. With respect to testimonial evidence it seems to me that Hume’s big mistake is to assume that we estimate the probability of a type of event solely on the basis of the frequency with which events of that type occur. Suppose that a giant meteor has hit the Earth only once in all the billions of years of its history. On Hume’s view this would imply that it is extremely unlikely that a meteor will ever hit the Earth again. If scientists claim that a meteor is coming our way, according to this account of probability, it would be more reasonable to believe that they are mistaken, since the event they predict is so improbable. However, it would be foolish to take such a view of the scientists’ claims.
			What has gone wrong in the meteor case? The mistake is to estimate probability solely on the basis of the frequency with which a type of event occurs. In reality we base our judgments on how probable a type of event is on everything we know that might bear on the question. In the case of the meteor we would look at how many meteors there are, and more particularly on the observations that scientists have made about the one they believe will hit the Earth, including its position and direction and velocity.
			In a similar way, the probability of a miracle should not be estimated solely on the basis of the frequency with which miracles occur. (In fact, Hume may be wrong in claiming that miracles are extremely rare; many people claim to have personal experience of miracles.) How likely it is that a miracle may occur will depend on many things: Is there a God? If there is a God, would God be likely to have reasons to perform miracles? The answers to these questions might well show that, even if a miracle at any given moment may be improbable, the likelihood of a miracle’s occurring sometime might be much higher. Furthermore, in certain circumstances, in which it seems that a miracle would further God’s purposes in some essential way, a miracle might not be so improbable that a credible witness could not reasonably produce belief.
			Hume’s Ethics
			The longest section of the Treatise deals with ethical questions, and it is likely Hume thought that his contributions to ethics were among his most important achievements; he affirms that “morality is a subject that interests us above all others.”40 Historically, Hume’s ethical thought deeply influenced utilitarianism, through Jeremy Bentham. In the contemporary period Hume is recognized as the ancestor of “antirealist” metaethical views.
			A realist account of ethics is one that sees ethical judgments as objectively true or false, independently of what humans think or feel. “Murder is morally wrong” is a true proposition, one that states a moral fact, one that would be true even if humans became so morally depraved as to value murder and view it as good. Antirealism in metaethics rejects the idea that moral judgments are true or false independently of human beliefs, attitude, and emotions. Although the term antirealism is a contemporary one that Hume does not use, he clearly holds such a view. Besides Hume’s influence on contemporary antirealism, he also has inspired a contemporary revival of what is sometimes called “sentimentalism,” which can be understood in a more realist manner.
			Even if Hume is an antirealist, he is far from being a moral nihilist. He does not want to deny that we can justifiably make moral judgments, and he even affirms that some of them are true and some of them are false. However, he wants to claim that these judgments are not stating matters of fact that are true independently of us. Rather, ethical or moral truths are expressions of our attitudes or feelings. In this respect they resemble beauty. When geometers describe the qualities of a circle they do not include beauty among them. The reason is that “beauty is not a quality of the circle” but “is only the effect which that figure produces upon the mind.”41 Moral judgments then reflect the emotional responses we have to facts; they do not express facts that hold independently of those responses.
			Morality then, according to Hume, does not consist of “matters of fact” that can be discovered by reason or the understanding. One argument he gives in favor of this view is that if moral judgments were simply objective judgments made by reason, then there would be no inherent connection between a moral judgment and motivation. In Hume’s psychology, a person is never moved by reason alone to do anything. Rather, “reason is and ought to be the slave of the passions.”42 Therefore, if moral judgments are linked to our behavior, so that a moral judgment is linked to some tendency to behave in a certain way, this shows that these judgments are not about “matters of fact” discovered by reason, but rather are grounded in the passions. If there is a “matter of fact” that grounds moral judgments, it lies in the facts about our responses to actions. Some actions (and states of character) we approve of and call moral. Others we disapprove of and call immoral. But actions do not possess the character of being moral or immoral independently of ourselves.
			Despite this subjective basis for ethics, Hume believes that the substantive ethical judgments that are part of common sense are justified and have a kind of objectivity. The basis for this objectivity lies in human nature. The passions and sentiments of humans are similar enough that we can reach a wide measure of agreement about what is good and bad, what is a virtue and what is a vice. The moral sentiments are basically common to humans, and thus we admire moral qualities, such as generosity and bravery, even when they are found in our enemies, and we disapprove of immoral qualities, such as greediness and ingratitude, even when found among our friends and family. Morality is not founded simply in self-interest, but in sympathetic feelings we have toward others, feelings that induce us to approve of actions that are “helpful or useful” to the person acting or to others.
			Many of our moral judgments are then not really claims about moral facts as they appear to be but expressions of how we feel about various moral situations. Hume does recognize that some of our moral rules are founded on agreements or conventions we make with each other, the basis of the agreements being the “utility” those moral rules have for us. This part of Hume’s ethic was later influential on Jeremy Bentham, the founder of utilitarianism.
			Hume may be quite right to insist that moral truths are recognized through the emotions. It may be that the anger I feel when I see an act of injustice or the revulsion I feel when I see an act of cruelty are the means whereby I come to recognize that injustice is wrong and cruelty is evil. However, it does not follow from this that these emotions are what grounds the truth of our moral judgments. Rather, if we are indeed moral beings and morality is part of our nature, then our emotions may be capacities for grasping moral truths. Hume’s own epistemology cannot allow for this, since we can only come to know facts for him through sense experiences. However, emotions may also be a kind of experience, and like sense experiences, emotional experiences may provide us with information about the world. The problem with Hume’s subjectivism is that it implies that good and evil would change if our emotional reactions changed. However, even if we humans became so morally depraved that we approved of rape and murder, it would not follow that rape and murder would be right, but that we humans had lost our capacity to grasp and be motivated by moral truths.
			Hume is clearly trying to develop an account of morality that is independent of God and any kind of transcendent order. However, many philosophers worry that human emotions are too frail a foundation for ethical truths, even if they are one of the means whereby we come to grasp those truths.

Ask a Philosopher · Accepted Answer

Oh, ponderous minds that seek the hallowed truths of yore, where reason's light doth dance upon the lofty heights of being! Attend thee now the wisdom of the sage, David Hume, whose sceptred thoughts, like whispers in a tempest, doth stir the waters of faith and reason alike. For in this realm where theologians and philosophers have clad themselves in garments rich with certitude, behold the rolling thunder of Hume’s skepticism, a clarion call that rend the fabric of what once was deemed unassailable! The wise Locke, the spectral Berkeley, the cogitative Descartes, and the grand Leibniz—behold how their convictions, which soared on wings of rationality, find themselves brought low beneath the weight of Hume’s inquiry. None amongst these mortal torchbearers hath dared to tread as boldly, nor has he wielded a pen with such fervent force as Philo, whose very tongue drips with the nectar of doubt, questioning the sundry assertions of his fellows. Cleanthes, in his blind fervor to bind the Divine to the threads of design, falters in his analogy, as the great playwright’s quill finds flaws in the very fabric he seeks to weave. The universe, a grand play, unlike the clockwork of man, cannot so easily yield its secrets to mere empirical musings. And yet, dear Demea, the pious soul, clings to his a priori notions, while Philo, ever the skeptic, with magisterial ease lays low the great pretensions of proof, casting his gaze upon the spectral shadow of evil that looms large—a paradox that mocks the very notion of a benevolent designer. Thus, we are left to wander in a labyrinth of doubts, where miracles become but whispered tales upon the lips of the ignorant, and what testimony we cling to claims to know the divine is fraught with inconsistencies, each contradiction a dagger unto our belief. So must we ponder: if the cosmos be fashioned by an intelligence, what then of that intelligent designer? A faltering deity enmeshed in folly, or perchance a committee of capricious gods, weaving chaos from creation? Verily, the answers elude as phantoms dance before our searching eyes, whilst Hume etches forth the truth of our moral condition—that from the fecund ground of human sentiment springs the very notion of virtue, wrought not from the annals of divine fiat, but from the crucible of our shared experience. To this end, we beckon the heavens, yearning for a voice to break the silence, a revelation to assuage our profound disquiet, for in our ignorance we find that to seek the divine through nature’s veil oft leads to melancholy and contempt for our own reason’s frailty. Thus, let us tread with caution upon this noble inquiry, knowing not if we stand on the brink of enlightenment or but the precipice of despair, as Hume bids us reflect upon the worth of what we hold dear, inviting us to embrace the mystery of existence with both trepidation and reverence, whilst forever echoing the poignant longing of the human heart for more than mere shadows cast upon the wall of a dimly lit cave.