Evidentialism: Essays in Epistemology by Earl Conee

Author:Earl Conee [Conee, Earl]
Language: eng
Format: epub, pdf
ISBN: 0199253722

end p.160

turn on some facts about reference class correlations (or propensities), there must be some way in which suitable reference classes are determined. Specifying how this is done is no doubt difficult, but in principle it must have a solution. Likewise, people make justification judgments with understanding and confidence. These are clearly determined by some fact about truth-conduciveness of associated types. So although the task may likewise be difficult, in principle a solution must exist.

The point at which this reasoning is plainly contestable is the claim that justification judgments depend in some way on something's truth-conduciveness. Epistemic justification is basically a matter of the kind of reasonableness that is determined by evidence. Reasonable belief need not tend toward truth. So judgments about justification can be understood, and correct, without depending for their truth on something truth-conducive.

Furthermore, even if justification is connected to truth in some notable way and there is some solution to the reference class problem, there is no good reason to think that reliabilists can make use of that solution to construct a satisfactory reliabilist theory. There is no good reason to think that exactly those beliefs that can be said in some context to be probably true are justified.

2. Process reliabilists do not have to specify a single relevant type for each belief. Most generally, what they must do is to specify some reliability rating associated with each belief. One way a process reliabilist might specify a reliability rating, without identifying a single relevant type, takes two steps. The first step is to specify multiple relevant process types for each belief. The second step is to specify some combinatorial technique that uses the various reliabilities of the specified types to yield a reliability rating for each belief. (An unpublished manuscript by Allen Plug explores this possibility.) Another alternative is to specify a combinatorial technique that does all of the work. That is, the technique takes the reliabilities of all of the types tokened by the process, and yields a reliability rating for the belief.

Nothing that we know of makes either of these alternatives appear to be more promising than the search for a single relevant type. In fact, they appear less promising. To identify a finite number of relevant types for a given belief-forming process, one would at least have to know approximately where to look. Yet, as things stand, common sense types of numerous varieties have some appeal to reliabilists, and so do scientifically identified types. The field seems wide open. And finding a combinatorial technique that gets a reliability rating out of some multiplicity of types and is plausibly associated with the belief's level of justification seems as formidable a task as finding a single relevant type.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.