James is the best prose stylist in American philosophy and the reason pragmatism is a household word, and those two facts are connected to the thing he is most criticized for. He wrote to be understood by people outside the subject, and the formulations that made him readable are the ones that made him vulnerable.

He trained as a physician, never practiced, taught physiology at Harvard, then psychology, then philosophy, and spent much of his twenties in a depression severe enough that he considered suicide. The way out of it is in the work: he decided, as an act rather than a conclusion, to believe that his will was free.

What he actually claimed about truth

The sentence that drew forty years of attack is that truth is what works, and the honest thing to say is that he did write things close enough to it to deserve the trouble.

His considered position is narrower. A belief is not a static picture matching a fact; it is a tool a person uses, and its truth consists in the process by which it is validated — the way it leads through experience, connects with other beliefs, and survives contact with what happens. Truth happens to an idea; it is made true by events.

Russell's response was that this makes truth depend on the believer's convenience, and Moore's was similar, and they were pressing on a real weakness. James's replies never fully closed it, partly because he kept reaching for examples about religious belief where the payoff is subjective by construction. Peirce objected from the other side and renamed his own position to escape the association.

What survives is the negative half, which is robust: a difference in belief that makes no difference anywhere is not a difference worth arguing about.

The will to believe

His most attacked essay and more careful than its reputation.

The claim is not that you may believe whatever suits you. It is that there is a specific class of question — where the options are live, forced and momentous, and where the evidence cannot settle it — in which withholding belief is itself a choice with consequences, and not a neutral one.

His examples are the useful part. Whether someone can be trusted is often only determinable by trusting them. Whether a venture will work may depend on whether anyone believes it will. In these cases the demand to wait for evidence guarantees the evidence never arrives, which makes evidentialism a policy with its own costs rather than a neutral default.

The objection is that the conditions are elastic and that almost anything can be dressed as live, forced and momentous. Fair. The residue is still worth having: the refusal to decide is a decision, and it should be argued for rather than assumed to be the safe position.

The stream of consciousness

The Principles of Psychology (1890) took twelve years and is still read, which almost no nineteenth-century psychology is.

Its central move is a description. Consciousness does not come in units. It is not a chain of ideas or a sequence of states; it is continuous, personal, always changing, always selecting. The word he coined for it — the stream of thought — entered the language and entered the novel, and Joyce and Woolf are downstream of a psychology textbook.

The chapters on habit, on attention, and on the self are the ones still worth reading. The account of the self is particularly good, and it divides the self into the known and the knower in a way that connects directly to personal identity without pretending to have solved it.

Radical empiricism

The late position, and the most philosophically ambitious.

Ordinary empiricism takes sense data as basic and then has to explain how they get connected — by the mind, by inference, by habit. James's objection is that the relations between things are given in experience just as directly as the things are. You do not experience two items and then supply the and; the conjunction is part of what is experienced.

The consequence is a metaphysics in which mind and matter are not two substances but two ways of sorting one kind of stuff. It is the least popular part of his work and it anticipates a good deal of later philosophy of mind.

The objection

That the flexibility is doing too much work, and that his generosity toward positions he found humanly valuable was not always distinguishable from endorsement.

The charge lands hardest on religion. He wanted an account on which religious belief could be legitimate without settling whether any of it was true, and the account he produced makes truth sensitive to what belief does for a believer — which is exactly what would be needed to reach the conclusion he wanted. The book where he does this best is the one where he most carefully avoids the question: The Varieties of Religious Experience.