Safety, Sensitivity and “Distant” Epistemic Luck

Abstract

Prominent instances of anti‐luck epistemology, in particular sensitivity and safety accounts of knowledge, introduce a modal condition on the pertinent belief in terms of closeness or similarity of possible worlds. Very roughly speaking, a belief must continue to be true in close possibilities in order to qualify as knowledge. Such closeness‐accounts derive much support from their (alleged) ability to eliminate standard instances of epistemic luck as they appear in prominent Gettier‐type examples. The article argues that there are new Gettier‐type examples which are grounded in “distant” epistemic luck. It is demonstrated that sensitivity and safety theories cannot handle such examples.

Luck: Its Nature and Significance for Human Knowledge and Agency

“…7 . Those who group fake barn and hidden helper cases with the Classics include Plantinga (1993), Hiller and Neta (2007), Stone (2013), Freitag (2014), and Schafer (2014). Theorists who group fake barn cases with the Classics include Howard-Snyder, Howard-Snyder, and Feit (2003), Lackey (2009), Reed (2009), Turri (2012), and Pritchard (2012a).…”

No Luck in the Distance: A Reply to Freitag

“…At all other times, it shows 6:00. Jim, not aware of the clock's malfunctioning, looks at the clock at 3:00, thereby picking up the true belief that it is 3:00 (Freitag, , p. 54).…”

“…The odometer presents any number of miles up to 69,999 miles correctly; with the 70,000th mile, however, the odometer rolls over to 60,000. Ignorant of this defect, Jenny does not take this possibility into account (Freitag, , pp. 57–58).Freitag's intuition is that Jenny does not know that the car has less than 69,000 miles.…”

“…The reason why Freitag thinks that safety and sensitivity cannot rule out O dometer as a case of knowledge is the following. On the one hand, Jenny's belief that the car has run less than 69,000 miles is safe because there are no close possible worlds in which Jenny believes that proposition falsely (according to Freitag, , p. 58, the relevant world ordering is fixed by how many miles the car has run: “In 68,400‐worlds, as in 68,600‐worlds, Jenny continues to hold her belief and this belief is still correct. The worlds in which Jenny believes falsely, via the same method, that the car has run less than 69,000 miles, e.g., 70,000‐worlds, do not, by my reckoning, count as relevantly close”).…”

“…I n a recent article in this journal, Wolfgang Freitag argues that Gettier‐style cases that are based on the notion of distant epistemic luck, i.e., “luck which discloses itself in distant possible worlds alone” (Freitag, , p. 45), cannot be ruled out as cases of knowledge by modal conditions such as

S afety (if S knows that p , then S believes that p truly in the actual world, and in nearly all, if not all, close possible worlds in which the conditions for belief formation are the same as in the actual world, S only believes that p when p is true)

or

S ensitivity (if S knows that p , then in the closest possible worlds where p is false and in which the conditions for belief formation are the same as in the actual world, S does not believe that p ) In this article I will argue that: Safety and sensitivity can be easily fixed just by adding a clause to the effect that one must not believe closely related false propositions. Freitag provides no convincing reason for the existence of “distant” epistemic luck.

…”