The topic of Hume’s section on miracles is not whether miracles are possible. After all, almost anything is possible, according to Hume. Rather, what is under discussion is whether we could ever have adequate reason to believe that a miracle has happened. In particular, he is concerned with whether we could ever have adequate reason to believe testimony about miracles, that is, someone’s report to have witnessed a miracle.
The section is divided into two parts that correspond to the distinction between proof and probability (see ¶¶ 4–6 in Part I, pp. 73–75).
In Part I, Hume imagines that the testimony in favor of a miracle amounts to a proof. As I understand him, that means that the testimony comes from someone who has always been accurate: whenever such a person’s reports have been checked, they have always matched the facts.
In Part II, Hume weighs the probability that testimony about a miracle is accurate against the probability that it is inaccurate. He does that by looking at the characteristics of people who report having witnessed miracles. Since all of the other people who report having seen a miracle are unreliable, according to Hume, we should regard the next person who reports having witnessed a miracle as also being unreliable.
Part II is especially difficult to follow, so I will pay special attention to it.
Introduction: Hume says he has come up with an argument similar to the one Tillotson used against transubstantiation. (¶¶ 1-2, pp. 72–73)
Rules of reasoning.
Even if the testimony for a miracle amounts to a proof, the balance of the evidence would favor not believing it. This is the main argument of Part I. Also, note the definitions of a miracle as “a violation of the laws of nature” and “a transgression of a law of nature by a particular volition of the Deity.” (¶¶ 11–13, pp. 76–77)
Four reasons why, in real cases, the testimony about miracles has been bad.
Firstly, no one who has claimed to witness a miracle had the intellect, character, and personal stake in the truth to be reliable. Nor have reported miracles been public enough to rule out fraud or error. (¶ 15, p. 78)
Secondly, there are benefits to claiming to have witnessed a miracle; this motivation makes the testimony suspect. (¶¶ 16–19, pp. 78–79)
Thirdly, miracles are always first reported by ill-educated people; the reports cease happening as civilization advances. (¶¶ 20–23, pp. 79–81)
Fourthly, different religions make equally good claims about miracles that are incompatible with one another. What makes them incompatible is that they are evidence for different gods or that different churches are God’s true representative on Earth. (¶¶ 24–28, pp. 81–86)
Note: Hume gives several examples of reported miracles (Vespasian, Cardinal De Retz’s story, and the tomb of Abbe Paris). These have two qualities: (a) the evidence favoring them is very good and (b) they are not accepted in standard Christian (especially Protestant) practice. These are meant to illustrate the fourth point. Hume is taking it for granted that his audience would dismiss these putative miracles, despite the strong evidence in their favor. Thus they are examples of the fourth point, that different religious traditions have incompatible stories about miracles.
These problems are especially acute when miracles are used to establish a new religion. (¶¶ 29–34, pp. 86–87)
Summary (¶ 35, pp. 87–88)Was this originally meant to be the end?
Apparent qualification: Hume is willing to concede that there could be testimony about a violation of the course of nature that is worth believing. Example of total darkness for eight days (¶36, p. 88). But he insists that he could never have adequate testimony to believe that Queen Elizabeth rose from the dead (¶37, pp. 88–89). What’s the difference?
Another summary (¶¶ 38–39, p. 89)
Christianity is based on faith, not reason. So miracles never should have been used to establish the truth of Christianity in the first place. (¶¶ 40–41, pp. 89–90). Is that sincere or a bit of misdirection?