Comments for: Two Cheers for Akrasia by Kevin Dorst Konstan3n Genin Carnegie Mellon University 2018 Eastern APA Savannah, Georgia
Probabilis3c Frames • W is a set of possible worlds . • P is a func3on mapping world w to P w, the uniquely ra1onal credence func3on at w . (Samet, 1997)
Probabilis3c Frames We want to talk about what credences you ought to have: [ P ( p ) = t ] := { w : P w ( p ) = t } In English: `one (you) ought now to be t - confident of p .’
Probabilis3c Frames For this setup to make sense it must be that the total evidence is a func3on of the world. Let E w be the set of worlds compa3ble with the total evidence in w .
Standard Sta3s3cal Setup: Sequen3al Binary Experiment Worlds = infinite sequences of coin flips. Eviden1al states = cones of possible extensions of finite sequences: observed so far
Higher-order Uncertainty Kevin argues that the following situa3on is ra3onally permissible: • and P w 6 = P w 0 P w ( w 0 ) > 0 . •
Eviden3al Internalism • Internalism: If E is your total evidence, then it entails that it is your total evidence: E w ⊆ { w 0 : E w = E w 0 } . • Externalism : Some3mes your total evidence does not entail that it is your total evidence. Gallow (2017)
The Internalist Argues against HoU 1. If then P w 6 = P w 0 E w 6 = E w 0 . 2. If then (Internalism). E w 6 = E w 0 { w 0 } ∩ E w = ∅ Ccl. P w ( w 0 ) = 0 .
Dorst denies Eviden3al Internalism Kevin argues for higher-order uncertainty by producing examples in which eviden3al internalism seems to fail.
Bianca the Biologist If the sample is poisonous, Bianca will eventually p verify that it is. If the sample is safe, she will never be able to verify that it is safe. ¬ p
The Internalist Response If total evidence is a feature of the world, then so is what evidence you have managed to figure out by now.
Bianca the Biologist Whether or not Bianca has realized that the sample is p, r poisonous is now a feature of the world. p, r p, r
Bianca the Biologist Total evidence now sa3sfies eviden3al p, r internalism. p, r p, r
Word Comple3on 2 Suppose there are at 0-2 possible comple3ons of the word fragment. 1 0
Word Comple3on How many comple3ons I’ve thought of is part of (2,2) the world. Number I’ve thought of (1,1) (2,1) (0,0) (1,0) (2,0) Number of Comple3ons
Word Comple3on Eviden3al internalism is (2,2) sa3sfied. Number I’ve thought of (1,1) (2,1) (0,0) (1,0) (2,0) Number of Comple3ons
Peer disagreement The situa3on Kevin describes could be modeled in two different ways.
Peer disagreement P 1 = P 2 = P 3 � me 1 = me 2 = me 3 P i (j) = 1/3 you 1 = you 2 = you 3 me 1 =P 1 =you 1 me 1 =P 1 =you 1 me 2 =P 2 ≠ you 2 me 2 =P 2 ≠ you 2 me 3 ≠ P 3 =you 3 me 3 ≠ P 3 =you 3 In the first situa3on Open and Uncertain are sa3sfied but there is no higher-order uncertainty.
Recommend
More recommend
