L16 Reading Guide, Swinburne on miracles

Reading: Swinburne, “Miracles and Historical Evidence”, pp. 455-460

Reading Guide

Swinburne seems to be defining miracles as events that cannot be explained by the laws of nature because these events are physically impossible. In contrast to Hume, he broadens the scope of evidence for miracles. The basic distinction is between historical evidence, which is divided into 3 types (testimony, our own apparent memories, physical traces), and scientific evidence (“our contemporary understanding of what things are physically impossible or improbable”, 455).

On p. 456, Swinburne articulates a basic principle and 3 subsidiary principles regarding how to weigh evidence. He takes these principles to be non-controversial, “adopted … by and large by all historical investigators” (457). In the course of discussing the first subsidiary principle, he uses a term that may be unfamiliar to some of you; ceteris paribus means “all other things equal”. His discussion of the 2nd subsidiary principle may be a bit difficult for some of you to follow, but I think the example of Smith and Jones should help you understand what he means. Basically, he’s talking about a process of making finer and finer distinctions. For example, we can lump all testimonial evidence together or we can make classifications within the category of testimony (e.g., testimony of children vs. testimony of adults, testimony of those who have a stake in the outcome vs. testimony of those who have no stake, Jones’ testimony vs. Smith’s testimony). The need to make narrower classifications is spurred by the conflict of testimony. After making these finer-grained distinctions, we then assess which class of testimony is more reliable.

Swinburne then turns to the question of whether historical or scientific evidence should take precedence when trying to determine whether a miracle has occurred. He agrees with Hume that “we should accept the historical evidence … only if the falsity of the” historical evidence “would be ‘more miraculous’, i.e., more improbable ‘than the event which he relates’” (459). However, he rejects the assumption that scientific evidence automatically takes precedence, an assumption that Antony Flew, an influential modern Humean, accepts and attributes to Hume. Flew identifies what he takes to be an important difference between historical evidence and the laws of nature, viz., that the latter can “in theory, though obviously not always in practice, be tested at any time by any person” (Flew, quoted in Swinburne, 457). Swinburne gives a plausible interpretation of why Flew takes this to be significant, but then argues that the contrast that Flew draws is incorrect. So, “Flew’s justification of his principle” – the principle that scientific evidence always outweighs historical evidence – “is mistaken” (457). But this does not establish that Flew’s principle is wrong. In the last 2+ pages of the article, Swinburne argues that Flew’s principle is wrong: “claims that some formula L is a law of nature, and claims that apparent memory, testimony, or traces of certain types are to be relied on are claims established ultimately in a similar kind of way … and will be strong or weak for the same reasons, and so neither ought to take automatic preference over the other” (457). This section is a bit technical, but I think the basic points are clear enough. Just pay attention to what the variables stand for – L = a specific law of nature, C = correlation between a specific type of (historical) evidence and the facts/the truth (basically, C specifies reliability), and E = the (alleged) event. When scientific and historical evidence conflict, we determine which takes precedence by considering (a) whether L or C is better established (see p. 458, 1st full paragraph for a statement of how a L or C is established) and (b) whether the L is universal or statistical and (c) whether the C is universal or statistical. Swinburne admits that “in general any one correlation C will be less well established than L, and since L will usually be a universal law, its evidence will in general be preferred to that of C” (459). Still, there may be exceptions to the general rule such that the historical evidence will trump the scientific evidence.