Unformatted text preview: A Tension in TW’s Usage of Probabilistic Relevance in KAIL 11/22/06 (B.F.) I mentioned in my discussion of TW’s stuff on the explanatory power of knowledge vs true belief (see my chapter 3 notes) that TW has a comparative notion of explanatory relevance that is radically different than his notion of evidential relevance. This is despite the fact that TW uses probabilistic relevance in his treatments of both concepts. I will use the notation e (C,D) to denote “the degree to which D ( explanans ) is explanatorily relevant to C ( explanandum )”. While TW was not claiming that e (C,D) = ρ(C,D) , he was claiming that ρ is a “guide” to e . And, it was crucial to his main example (which was supposed to bolster his case for the explanatory power of knowledge over true belief) that the following principle is violated by e (1) If D entails C but E does not entail C , then e (C,D) > e (C,E) . In fact, TW gives an example in which (he claims) ( i ) D entails C , ( ii ) E does not entail C , but ( iii ) e (C,D) < e (C,E) . And, TW thinks (therefore) that it is a virtue of ρ that it is capable of violating (1) in exactly this way.(1) in exactly this way....
View Full Document
- Fall '06
- TW, probabilistic relevance, evidential relevance