Handout on Sosa

Below the fold are the notes given out in class on Ernest Sosa’s article, “Philosophical Scepticism and Epistemic Circularity.” [UPDATE: This is now the revised version of the handout given out in 2010.]

Reliabilism (Goldman): Basically, the idea that whether a belief counts as knowledge depends on whether it was arrived at through a process that is actually reliable.
Not a process that we know to be reliable, or a process that is reliable in every possible world, but a process that is reliable in this world.

Reliabilism and Skepticism: Consider two “people,” Pinky and The Brain.
Pinky is a normal person interacting in the world in normal ways.
The Brain has exactly the same sense-experiences as Pinky, but is a brain in a vat.
According to reliabilism, the beliefs Pinky gets through his senses are justified, because relying on his senses generally gives him true beliefs.
The beliefs the Brain gets through his senses are not justified, because relying on his senses does not generally give him true beliefs.
Note that Pinky’s sensory beliefs are justified even when they’re false, because the method he’s using is generally reliable.
And the Brain’s beliefs are unjustified even when they happen to be true, because the method he’s using is generally unreliable. “Reliable” is different from “true” here.

Sosa’s argument: His favorite theory of justification, reliabilism, is circular.
But any theory of justification has to be circular. A non-circular theory is logically impossible.
So it’s not really a knock against reliabilism that it’s circular. A circular theory like that is the best one we could possibly get.

Again, Pinky doesn’t have to know that his senses are reliable in order to get justification through them. They just have to be reliable (in Pinky’s actual situation).

Objection to the reliabilist point of view:
Pinky’s beliefs are justified because he’s not a brain in a vat.
Does Pinky know that he’s not a brain in a vat?
Well, he can come to the belief that he’s not a brain in a vat through reliable means. He can rely on his senses, which (ex hypothesi) are actually reliable.
What good is this for us? It gives us a way out of skepticism—if we aren’t brains in vats.
But isn’t the whole point that we’re trying to figure out whether we might be brains in vats? It seems circular.

Sosa’s terminology:
“Formal internalism”: The only justification of a belief is through reasons or arguments.
It’s impossible to attain this without circularity or endless regress.

Three kinds of “formal externalism”:
E1. Coherentism (justification through membership in a coherent body of beliefs)
E2. Foundationalism of the given (justification based on what’s present in your mind)
E3. Reliabilism (justification through processes that are actually truth-conducive)
Usually when philosophers talk about “externalism” they mean something like E3.

E1. Sosa argues that coherence alone, plus truth, isn’t enough for knowledge; a coherent belief system could be completely cut off from the world even if some beliefs happen to be true. The Brain has a coherent belief system.

E2. The question is, what is it about the given that makes us justified in believing in it?
We can be wrong about some things that are given to us. We can all tell that an image of a triangle is an image of a triangle just by envisioning it; can we all tell whether an imagined shape has ten or twelve signs just by envisioning it?
Sosa argues that introspection yields justification when (and because) it’s reliable. So E2 collapses into E3.

Stroud’s criticism of E3:
MR: In order to understand one’s knowledge satisfactorily one must see oneself as having some reason to accept a theory that one can recognize would explain one’s knowledge if it were true (p. 272).
Sosa sees this as coming from:
(a) understanding p is having reason to accept some q that would explain p if q were true;
(b) we want to understand how we know things.
But from this it only follows that you have to have a reason to accept your explanatory theory, not that you must see yourself as having a reason to accept an explanatory theory.
Sosa sees the additional demand as motivated by the desire to find a belief underlying everything. That’s “formal internalism,” which leads to circularity or infinite regress.

But what good is understanding if we don’t know we understand? We don’t want to attain understanding by dumb luck.

Well, by the reliabilist’s lights we can come to know that we understand our knowledge. We just have to use the same knowledge-gathering methods that we’re coming to understand.
For instance, I can use my senses in testing that my senses are reliable. (Eye exams.)

Alston: This is circular and unsatisfying. Instead, we should rely on the practical reasons for accepting our belief-forming practices.
Sosa: But we actually have to rely on those belief-forming practices to know that those practical reasons obtain.
Circularity remains inescapable.

Self-support is the best we can do; if the method actually is reliable, if our beliefs about how it works are true, and if we correctly believe that it’s reliable, are we really so badly off? (See p. 282.)
Especially because it’s the best we could possibly do. The better alternative isn’t logically coherent.