Knowledge, Belief, and Opinion

Two people say the coin will land heads. One has watched a slow-motion recording of the toss. The other simply feels lucky. The coin lands heads, and both were right. Something separates them, and saying what it is turns out to be the work of a whole subject.

What is being analysed

The word "know" covers several things, and only one of them is under examination here.

Two other uses of the word are set aside rather than denied. Knowing how to swim is a capacity rather than a stored fact, and Gilbert Ryle argued that it cannot be reduced to knowing a set of propositions about swimming, since one can recite the propositions and sink. Knowing a person or a city is familiarity built by acquaintance, which Bertrand Russell distinguished from knowledge of truths about them.

Both are real and neither is the topic. When epistemologists ask what knowledge is, they mean the first sort, and the analyses proposed are analyses of the form "S knows that P".

The three notions

Belief in this sense has nothing to do with religion or with conviction. It is the ordinary state of accepting something, and most of what anyone believes was absorbed rather than decided. You believe there is a floor below you now, and you did not arrive at that by reasoning.

The definition is deliberately thin. What makes a statement true, and whether truth consists in correspondence to fact or in some other relation, is a large question that the philosophy of language takes up. Nothing in this tier depends on the answer. What matters here is only that truth is a property of what is believed rather than of the believer, and that whether a belief is true is not settled by how confident the believer is.

Opinion is not a third thing alongside belief. It is a belief with a particular standing: something held provisionally, or in an area where the holder does not expect everyone competent to agree. Note that opinions can be true. The person who feels lucky about the coin held an opinion, and it was correct.

The definition is openly incomplete, and any honest one at this stage will be. What can be fixed now is two constraints that survive every account. Knowledge requires truth: nobody knows something false, and finding out that a claim is false retracts the claim to have known it rather than merely revising it. And knowledge requires more than accidentally hitting the truth.

The Meno's question

Plato put the difficulty in a form that has not been improved on. In the Meno, Socrates observes that a man who has a correct opinion about the road to Larissa will guide you there as reliably as a man who knows the road. On the journey itself, the two are worth exactly the same.

So why is knowledge better? The answer offered compares true opinions to the statues of Daedalus, which in the story run away unless they are tied down. True opinions are like that: correct while they last, and liable to be talked out of, forgotten, or abandoned at the first contrary suggestion, because nothing holds them in place. Tie one down with an account of why it is so, and it becomes stable. In the Lamb translation, "this is why knowledge is more prized than right opinion: the one transcends the other by its trammels".

Whether the tethering account is right, the diagnosis is one that survived. The difference between knowing and being right is a difference in what supports the belief rather than in the belief's content, and the two people betting on the coin held the same view about the same coin.

Is belief really required

Where this goes

The coin example leaves a gap that has now been named. Both bettors believed, both were right, and one of them had something the other did not.

The next article takes up that something directly.