In this section, entitled "Of Miracles," Hume argues that we have no compelling reason even to believe in miracles, and certainly not to consider them foundational to religion. Our knowledge of miracles derives exclusively from the testimony of others who claim to have seen miracles. Since we receive this testimony secondhand from the experience of others, we should treat it as less reliable than our own experience.
Belief, Hume asserts, should be proportioned to evidence. In those cases where all evidence points to one particular conclusion, we can be almost certain that that conclusion is correct. However, when there is evidence both for and against a certain conclusion, we can regard that conclusion only with a certain degree of probability, to the extent to which the evidence for it outweighs the evidence against it.
In the case of miracles, our evidence in favor of their existence comes from the testimony of witnesses, and our evidence against them comes from their contrariety to the laws of nature. Both our faith in the testimony of others and our knowledge of the laws of nature are founded in experience. Human testimony tends to accord itself with reality, and the laws of nature tend to be constant. Since a miracle, by definition, is a violation of the laws of nature, it can only be credible to the extent to which the testimony in its favor is more forceful than the laws of nature that contradict it.
Hume provides four reasons to think that there has never been sufficient evidence in favor of a miracle to render it probable. First, no miracle is supported by testimony of a sufficient number of trustworthy people to rule out the possibility of falsehood. Second, while we should normally believe that which most closely accords itself with past experience, the sensations of surprise and wonder often lead us to unreasonable beliefs. There are countless instances of tall tales of all sorts that stem not from reasonable inquiry but from a love of wonder. Third, Hume remarks that most reports of miraculous events occur amongst barbarous or ignorant people, who may not be sophisticated enough to disbelieve fabricated testimony. Fourth, since every religion claims the veracity of its own miracles as against the miracles of every other religion, the evidence of all other religions opposes the evidence in favor of a miracle in any one particular religion. For instance, what a Muslim might consider a miracle would be considered a heresy by anyone of different faith.
Hume asserts that no testimony can ever count as a probability, let alone a proof, of the existence of miracles. All testimony in favor of miracles is based in experience, and this same experience opposes this testimony with contrary testimony and with the laws of nature. While God may be all-powerful and could contradict the laws of nature, we cannot ascribe any attributes or actions to him except for those that experience teaches us.
Hume concludes that religion is based in faith, not in reason. There is no rational ground for trusting in miracles, and he suggests that all the miracles found in the Bible are more likely the fabrications of their authors than a true revelation of the facts. The same can be said for prophecy as can be said for miracles. Religion is based in faith because it requires a kind of miracle--a willing subversion of our own natural reason--in order to assent to it.