ADDENDUM: OTHER CHANGES TO “CONCEPTS OF EVIDENCE”
1. Potential Evidence. This is the basic concept in the sense that the others are defined in terms of it. The definition in the paper is this: e is potential evidence that h, given b, if and only if (1) p(h/e&b)
> k; (2) p(there is an explanatory connection between h and e/h&e&b) > k; (3) e and b are true; and (4) e does not entail h.
The latest definition retains conditions (3) and (4) but replaces (1) and (2) with this:(A) p(there is an explanatory connection between h&e/e&b) > 1/2
The new definition is stronger than the earlier one in two respects. First, it replaces the vague idea of k (the threshold for “high” probability) with a specific number /. The reason for this is that on my view evidence requires a good reason for belief which, as I argue in The Book of Evidence (pp. 115ff), requires probability greater than /. Second, condition (A) above requires the probability of an explanatory connection between h and e to be greater than /, given e and b alone (rather than given h and e and b, as in (2) above). The argument behind this idea is presented beginning on page 152 of my book. In addition I prove that (A) obtains if and only if the product p(there is an explanatory connection between h and e/h&e&b)xp(h/e&b) is greater than /. Accordingly, the new condition (A) entails the earlier conditions (1) and (2), but not conversely.
2. Veridical Evidence. In the paper the definition is this: e is veridical evidence that h, given b, if and only if (i) e is potential evidence that h, given b; (ii) h is true; and (iii) there is an explanatory connection between h and e. In the book two concepts of veridical evidence are suggested, depending on whether condition (iii) must be satisfied. This allows one to deal with various strange examples, including the “severed head” case, in which e is potential evidence that h, and h is in fact true, but there is no explanatory connection between h and e.
In the severed head case, Henry drops dead ofa heart attack and later is decapitated. Is the fact that (e) he was decapitated evidence that (h) he died? In the weaker sense it is, and in the stronger sense it isn't, since there is no explanatory connection between h and e.
3. ES-evidence. In the book I introduce a fourth concept of evidence: ES-(epistemic-situation) evidence. It is relativized to a type of abstract situation in which one knows or believes that certain propositions are true, and one is not in a position to know or believe that others are, even if such a situation does not in fact obtain for any person. This type of objective evidence I define in terms of “veridical evidence,” as follows: e is ES-evidence that h (with respect to an epistemic situation ES) if and only if e is true and anyone in ES is justified in believing that e is (probably) veridical evidence that h. Using the example in my original paper, we would say that, relative to an epistemic situation ES of the sort the doctor was in on Monday, the fact that Alan's skin had yellowed (e) is ES-evidence that (h) Alan had an i.c.b. Anyone in that ES would be justified in believing that e is (probably) veridical evidence that h (even though in the case in question it turned out not to be veridical evidence).
4. Probability. In chapter 5 of my book I introduce a concept of “objective epistemic probability” needed to define “potential evidence.” It is construed as a measure of how reasonable it is to believe a proposition. On this conception, reasonableness of belief admits of degrees and is subject to the formal rules of mathematical probability. Certain physical and/or mathematical facts or states of affairs may make it reasonable to degree r to believe something, independently of the beliefs of particular individuals or of epistemic situations. Suppose that this coin, which has two sides marked heads and tails, is balanced; it will be tossed randomly three times; there are 8 possible outcomes in 3 tosses, 3 of which involve exactly 2 heads. These facts make it reasonable to the degree 3/8 to believe that the coin will land heads exactly twice. That there is this degree of reasonableness is an objective, nonphysical, normative fact determined by the physical and mathematical facts of the case. It is abstract in the sense that it is divorced from particular individuals and types of epistemic situations. In the book I indicate how this concept of probability is different from other objective interpretations of probability, including Carnap's logical concept and the propensity interpretation.