Wittgenstein published one book in his lifetime, believed it had solved philosophy, left the subject for a decade, came back, and spent the rest of his life arguing against it.

That is not a change of mind about a thesis. Both philosophies agree that philosophical problems arise from misunderstanding language and that the job is dissolution rather than solution. What changes is the diagnosis of how language works, and the second diagnosis says the first one was a picture that had held him captive.

The life, where it bears on the work

He was born into one of the richest families in Vienna, and gave his entire inherited fortune away in 1919 — to his siblings, on the reasoning that they were already rich enough not to be corrupted by it.

He trained as an engineer, came to Cambridge in 1911 to work on the foundations of mathematics, and wrote the Tractatus in the Austrian army on the Italian front, finishing it in a prison camp at Cassino. Three of his four brothers killed themselves.

Believing philosophy finished, he taught primary school in rural Austria from 1920, where he struck a child and left under a cloud; worked as a gardener; designed a house in Vienna for his sister with an exactness that had the ceiling of one room raised by three centimetres after it was built.

He returned to Cambridge in 1929, submitted the Tractatus as his doctoral thesis, and told his examiners — Russell and Moore — not to worry, they would never understand it. He gave away his professorship's comforts, worked as a hospital porter during the war, and resigned the chair in 1947. His last words were that he had had a wonderful life.

None of this is anecdote for its own sake. The demand for exactness, the suspicion of comfort, and the conviction that philosophy is something you do to yourself rather than a body of results are visible in the work.

The first philosophy

Language pictures the world. A proposition is a model, its elements standing for objects, arranged as the objects are arranged. What can be said is what can be pictured; the rest — ethics, aesthetics, the meaning of life, the nature of the logical form itself — cannot be said, though it may show itself.

The argument is in the Tractatus, along with the numbering system that organizes it and the ladder he tells you to throw away.

The second philosophy

The Philosophical Investigations, assembled from his notebooks and published in 1953, two years after he died.

The opening move is a quotation from Augustine describing how he learned language: adults named objects, and he came to understand. Wittgenstein's objection is that this is a picture of language in which every word is a name and every sentence a description — which is the picture the Tractatus ran on. It fits some of language and is taken for all of it.

What replaces it:

Meaning is use. Do not ask what a word stands for; look at what is done with it. "Slab!" on a building site means something because of the practice it belongs to, not because it labels an object.

Language-games. Speaking is one activity among many — ordering, joking, thanking, measuring, praying — and there is no single thing all of them have in common.

Family resemblance. What do all games share? Not one feature. A network of overlapping similarities, like resemblances within a family. So a concept can be perfectly usable without necessary and sufficient conditions, which removes the demand that had driven philosophy since Socrates.

Rule-following. Any finite set of examples is consistent with more than one rule, so nothing in the mind can determine how to go on. What settles it is a practice, and a shared one — which produces the private-language argument: a language only one person could understand would have no distinction between following a rule and thinking you are.

The therapeutic aim. Philosophical problems are confusions, and the treatment is to see how the language went on holiday. The target is not a false theory but a picture that held us captive.

The objection

The style, and what has grown up around it.

He writes in numbered remarks, without argument in the ordinary sense, and it is genuinely difficult to state a thesis of his that he would have accepted as a thesis. He would say that is the point. It also means that disagreement is hard to conduct: a reader who is unconvinced has nothing to refute, and commentary tends toward exegesis of what the master meant.

The result is a body of secondary literature more concerned with reading him correctly than with whether he was right, and a habit among admirers of treating an objection as evidence of misunderstanding. That is a real cost and it is a cost of the method rather than an accident.

The narrower charge is that the second philosophy dissolves problems by changing the subject. Told that the problem of other minds is a confusion about the grammar of sensation-words, a philosopher who wanted to know whether other people have experiences has not been answered. Wittgenstein's reply is that the question was malformed. Whether that is a cure or an evasion is the live dispute.