Safety and sensitivity
Also called: the sensitivity condition, the safety condition, modal conditions on knowledge
Two modal anti-luck conditions on knowledge: a belief is sensitive if one would not hold it were it false, and safe if it could not easily have been false.
Safety and sensitivity are the two leading modal conditions offered to exclude epistemic luck from knowledge. A belief that p is sensitive if, were p false, one would not believe it by the same method. A belief is safe if one could not easily have believed it falsely: in nearby situations where one forms the belief in the same way, it is true.
Sensitivity is the core of Nozick’s tracking theory and descends from Dretske’s conclusive reasons. Its best-known cost is closure: I can sensitively believe that I have hands but not that I am not a handless brain in a vat, though the first entails the second. Sosa proposed safety as the better condition, and Williamson and Pritchard developed it. Safety preserves closure but has difficulties of its own, among them that beliefs in necessary truths come out trivially safe, as they come out trivially sensitive.
Both are standardly stated as conditions on a believer. The Tracking Minimum is a condition on a source’s production of assertions, and it is deliberately neutral between them. The case that fixes its strength shows why. A broken thermometer sits in a room held at constant temperature by a thermostat. It is perfectly safe: it would not easily read falsely, since the temperature would not easily change. A safety condition therefore fails to exclude it, and a theorist who prefers safety to sensitivity might take that as a reason to reject any constraint that does. But the thermometer is excluded on other grounds: nothing about the temperature figures in the explanation of what it displays.
Because the minimum constrains production rather than belief, the standard difficulties for modal conditions do not arise for it in the same form. A source can fail to assert a necessary truth in nearby alternatives where its evidence fails to support it, which is why the minimum is stated disjunctively, in terms of the truth or the evidential support. Its test is interventionist: vary the evidence and see whether production follows; vary a proxy and see whether it does not.
Related terms
Questions that turn on it
Further reading
- Nozick, R. (1981). Philosophical Explanations. Cambridge, MA: Harvard University Press.
- Sosa, E. (1999). How to defeat opposition to Moore. Philosophical Perspectives 13: 141–153.doi.org/10.1111/0029-4624.33.s13.7
- Williamson, T. (2000). Knowledge and Its Limits. Oxford: Oxford University Press.doi.org/10.1093/019925656X.001.0001
- Pritchard, D. (2005). Epistemic Luck. Oxford: Oxford University Press.doi.org/10.1093/019928038X.001.0001