What Justification Is

The bettor who watched the recording had something the lucky one lacked, and naming it is not hard: he had a reason. The difficulty starts when you ask what makes something a reason for believing, rather than a reason for hoping, or a cause of believing, or a reason for saying so out loud.

Three attitudes, not two

Before the reasons, a correction to the picture inherited from the previous article. Beliefs were treated there as something a person either has or lacks, which leaves out the position most people occupy on most questions.

The three-way division matters because the middle option is the one that most needs defending. Someone who suspends judgement on a question can do so well or badly. Suspending judgement about whether the Earth is round is not an admirable refusal to overreach; it is a failure to respond to what is available. Any account of what makes belief appropriate has to also say when suspension is appropriate, and an account that only grades beliefs will miss half the cases.

Reasons of the right kind

The definition is doing more work than it looks. It excludes a large class of considerations that genuinely count in favour of holding a belief, and that have nothing to do with whether the belief is true.

Believing that your treatment will work may help it work, and that is a reason to believe it. Believing that your business will succeed may sustain the effort that makes it succeed. Blaise Pascal argued that the expected payoff of believing in God outweighs the expected cost, whatever the probability, and the argument does not claim to show that the belief is true. William James argued that in certain cases where evidence cannot settle the matter, and something important turns on it, believing is permissible on other grounds.

These are practical or prudential reasons. They can be strong, and they can be decisive about what to do. What they cannot do is make a belief more likely to be correct, which is why they are set aside here. Epistemology grades beliefs by a single standard, and wanting something to be true has never been evidence that it is.

Justification

Three features of the notion should be fixed early, because each is a place where beginners go wrong.

Justification comes in degrees while knowledge does not. Evidence can be weak, decent, or overwhelming, and the same belief can be better justified on Tuesday than it was on Monday because more came in. Any account has to specify how much is enough, and no account does so precisely.

Justification does not guarantee truth. A doctor who orders the right tests, reads them correctly, and diagnoses the most probable condition is justified, and can still be wrong, because the improbable sometimes happens. A theory on which justification guaranteed truth would make almost nothing justified, since almost nothing is certain.

Justification is a status relative to a person and a time. The same claim can be justified for the specialist and unjustified for the layman who happens to have heard it, because they stand in different relations to the evidence.

Having a reason and using it

There is a distinction here that the rest of the subject leans on constantly, and it is easy to state.

A person can be in possession of excellent reasons for a claim and believe the claim for entirely unrelated reasons. The detective who has assembled conclusive evidence against the suspect, and who believes the suspect guilty because he dislikes him, is in an odd position: the evidence justifies the conclusion, and his believing it is not justified by that evidence, because his believing it has nothing to do with the evidence.

Epistemologists mark this by separating two things. There is justification a person has available for a claim, which is a matter of what their evidence supports. And there is the standing of the belief they actually hold, which additionally requires that the belief rest on that evidence rather than merely coexist with it. The second is what knowledge needs. The relation of resting on is itself a difficult topic, and the intermediate tier treats it under the structural theories.

What is now in place

Belief, truth, and justification are the three notions that the standard analysis of knowledge is built from, and all three have now been stated.

Before that analysis is assembled, it is worth looking at where justified beliefs actually come from, since the theories on offer are largely disagreements about which sources can supply support and how.