What would you call a counterfactual theory of Justified True Belief?

Question

Ever since I heard the characterization of knowledge as justified true belief, the proper meaning of the word "justified" has always seemed clear to me: it should mean that you acquired the belief in such a way that you would not have acquired it if it had been false, which means that you can be certain of its truth.

Unfortunately, when I've tried to classify my view of justification, I've identified multiple theories that all seem in line with mine:

1. Nozick thinks that "P knows S" means that P believes S, S is true, P would not believe S if S were false, and if S is true then P will believe S. This is almost exactly my view, except I don't understand that point of the fourth condition. We've already assumed that S is true and that P believes S, so what is the force of saying that if the S is true then P will believe S? EDIT: Can someone clarify what kind of conditional is used in Nozick's fourth condition?
2. Goldman's Causal Theory of Knowledge states that "justified" means that the truth caused your belief. That's also in line with what I'm saying, because I see "A caused B" as "if A had not occurred, B would not have occurred".
3. Infallibilism states that that the justification of your belief must necessitate its truth. Again, that's consistent with my view: the justification for your belief must be such that it could never lead you astray. In other words, if it leads you to believe something, that something must be true. And if that something had not true, then you wouldn't have acquired the belief.

So what school of epistemology do I belong to? How can my views be consistent with three different theories? Or am I missing subtle distinctions between them? Are there other views closer to mine?

Answer 1

First, let's compare Infallibilism with your interpretation of Nozick. The counterfactual claim that "if S were false, P would not believe that S" is weaker than the claim that "P has some justification R for S, such that S is a necessary consequence of R". In the Infallibilist version, P actually has to have something that constitutes a justification that entails the believed proposition - we don't commit ourselves to saying what that justification is, but we still need to say that it is in principle decided what that justification is. By contrast, Nozick's theory lets us be neutral as to what it is about their situation that makes them suitably justified. The key point is simply that the falsity or truth of the proposition are necessarily related to the worlds in which the knower does or does not come to acquire the beliefs that they have.

Goldman's "Barn Facades" example serves to demonstrate the distinction here (driving down the road I see some wooden cutouts made to look like barns, but mistake them for the real thing; there is also one real barn among them, and I correctly identify it as a barn but on the grounds of objectively dubious visual evidence). Nozick's requirements would be satisfied as long as some kind of explanation could be given as to why it turned out that we did in fact necessarily wind up in front of the only real barn that there was, without requiring any additional reasoning or reflection on my part as the knower. I don't actually need to have that additional knowledge myself on Nozick's account to be in a position of knowledge about the barn.

By contrast, the infallibilist will be unsatisfied, because I only relied on my visual perception in making my claim, and these additional facts about the metaphysics of the situation will not have been reflected in how I came to the conclusion that I did. Even though it metaphysically could not have been any other way, and so the truth of my belief is in some sense a necessary consequence of how I came to believe it, the patterns of reasoning I employed to reach my conclusion would not count as a formally valid argument in themselves without explicitly reflecting on the additional premises that Nozick's theory allows to be implicit in the situation.

What might be gleamed from this is an "internalist/externalist" distinction in epistemology that seems relevant to deciding your position. The question as to whether either approach can be in line with Goldman's theory, however, doesn't seem obviously epistemological to me at all. Rather, the puzzle is to flesh out the notions of modality and causality that we're taking to be live in our methodology, and this is hardcore metaphysics. Do we think that Hume's notion of causation as constant conjunction is the best we've got, or do we want to take a more involved notion of counterfactuals or mathematical models as usefully informative about regularities in the world? Is what is possible covered by what is caused, or maybe vice versa, both or neither?

Goldman's theory can be read as drawing on the metaphysics of causality rather than the metaphysics of modality. This may have various advantages and disadvantages (probability methodology seems to be harder to come by, but causal explanations might be more naturalistically coherent than appeals to abstract possible worlds), and I guess it's up to you to decide which framework might be more suited to your needs.

Answer 2

The table below shows 9 possible ways that the truth value of a proposition can map to belief, where - indicates that belief is invariant for that truth value.

True   |   False 
0          0         Always believes False
0          1         Inverse of justified belief
0          -
1          0         Justified belief, according to Nozick
1          1         Always believes True
1          -
-          0         Condition 3 holds, but not 4
-          1         
-          -         Invariant to truth

Note that, in the 3rd last case, belief never results when the proposition is false, but that does not guarantee that belief will result when the proposition is true. Imagine a black box that returns true/false to any query and that, when it says true, you can be sure that its correct, but, when it says false, you can't.