Saul Kripke

Picture a lecture room at Princeton University, January 1970. Word has gone around, so the room is full. In walks the speaker: twenty-nine years old, rumpled, bearded, already a legend among logicians for work he did in high school. He carries nothing. No manuscript, no notes, not an index card.

Then he starts talking — in complete, precise, quotable paragraphs — and keeps talking, across three lectures, until he has dismantled the reigning theory of how words attach to the world, rebuilt it from scratch, resurrected a piece of Aristotle's metaphysics that modern philosophy had buried with honors, and thrown a spear into the leading materialist theory of the mind. For garnish, he explains why some truths that had to be true — that could not possibly have been otherwise — were discovered through telescopes and laboratory flasks.

Stop on that last claim, because it should feel wrong. We're taught to sort truths into two piles. Pile one: things you can know just by thinking, like 2 + 2 = 4 — and surely those are the necessary ones, the couldn't-have-been-otherwise ones. Pile two: things you learn by looking, like it's raining in Omaha — the accidents, the could-have-beens. Two piles, one distinction. The young man at the front of the room says: no. The piles cross. There are truths you can only learn by looking that are as necessary as arithmetic. Necessity is a fact about the world; knowability is a fact about us; and nobody ever had the right to solder them together.

The transcripts of those three unscripted evenings became Naming and Necessity — routinely called the most influential work of philosophy of the late twentieth century, and by some reckonings the most influential philosophy lectures since the Second World War. The speaker was Saul Kripke: the rabbi's son from Nebraska who never earned a graduate degree, published astonishingly little, and changed the subject anyway.

The Life

Saul Aaron Kripke was born on November 13, 1940, in Bay Shore, New York, and raised in Omaha, Nebraska, where his father, Myer Kripke, led the city's Beth El Synagogue and his mother, Dorothy, wrote children's books on Jewish life. It was a warm, bookish, argumentative household — a good greenhouse for what came next, because what came next was one of the most extreme intellectual childhoods on record.

The prodigy stories are the kind you'd assume were exaggerated, except that these ones check out. He taught himself ancient Hebrew around age six. He had read the complete Shakespeare by nine. He worked out algebra on his own before anyone showed it to him — he later said, with characteristic matter-of-factness, that he could have invented it — and he was reading Descartes and doing serious mathematics before he finished elementary school. Then came the part with no real precedent: while a teenager at Omaha Central High School, about sixteen when the ideas came together, he proved the first completeness theorem for modal logic — the logic of necessity and possibility — showing that a formal system for those notions captures exactly the truths it should. The paper, "A Completeness Theorem in Modal Logic," appeared in the Journal of Symbolic Logic in 1959, when its author was eighteen. Professional logicians had to reckon with a schoolboy from Nebraska. A story from those years has a university writing to offer him a teaching post, and the boy replying that his mother said he ought to finish high school first — a tale told in several polished versions and likely buffed in the retelling.

He went to Harvard and took his bachelor's degree in mathematics in 1962 — the only degree he would ever hold. No master's, no doctorate: like a handful of other figures in this book, he simply never needed one. The most famous absurdity of his college years is that while still a Harvard sophomore he was teaching a graduate logic course down the road at MIT — an undergraduate by day, a graduate professor by, so to speak, evening.

The career that followed ran through the aristocracy of American academia without ever passing through its usual gates: the Harvard Society of Fellows, a stint teaching at Harvard, then Rockefeller University in New York, then Princeton from 1977, where he held the McCosh chair in philosophy until retiring in 1998. In 1970 came the three January lectures at Princeton — delivered, as everything of his was, without notes, from a mind that seemed to store finished prose. A tape recorder ran; the transcript, lightly edited, was published in 1972 and as a book in 1980, and Naming and Necessity has never been out of philosophy's bloodstream since. His last act was a late-career move to the City University of New York's Graduate Center, which in 2007 founded the Saul Kripke Center to preserve and edit his work — necessary because Kripke's perfectionism meant that the mountain of what he had thought dwarfed the molehill of what he had published.

The man himself was famously, endearingly odd: the unworldliness, the rumpled suits, philosophy at full intensity at any hour with anyone who could keep up, exacting standards that kept masterpieces in desk drawers for decades. He could be difficult — impatient, prickly about how his ideas were used — and extravagantly generous, reconstructing a student's fumbled objection into something better than the student had meant and then answering that. In his vicinity, the highest compliment was to be argued with for three hours.

One controversy needs a line here, because this book has already told it in full. Ruth Barcan Marcus — chapter 254 — had argued in the late 1940s and in 1962 that names are simple tags, not descriptions, and had proven formally that true identities are necessary; a young Kripke sat in the audience at her famous 1962 Boston lecture, and in the 1990s a bitter priority dispute erupted over how much of the "new theory of reference" was really hers. That chapter weighs the claims; the settlement reached there stands here. What belongs to Kripke beyond any dispute is the possible-worlds semantics that made modal logic rigorous, the causal picture of reference, and the systematic case for the necessary a posteriori — the arguments this chapter now unpacks.

Kripke died of pancreatic cancer on September 15, 2022, in Plainsboro, New Jersey, at eighty-one — the Omaha prodigy gone at last, his unpublished archive still being mined.

The Big Ideas

The target: names as disguised descriptions. To feel the force of Naming and Necessity, you need the theory it demolished. How does the name "Aristotle" reach across 2,300 years and hook onto one particular Greek? The going answer, descending from Gottlob Frege (chapter 184) and Bertrand Russell (chapter 195), was: through description. On Frege's version, a name carries a sense — a bundle of identifying information — and refers to whatever fits it. On Russell's, a name simply abbreviates a description: "Aristotle" is shorthand for something like "the teacher of Alexander the Great who wrote the Nicomachean Ethics." A refined "cluster" version, associated with John Searle, said a name attaches to whatever fits enough of the associated descriptions. It was elegant, it explained how names could be informative, and nearly everyone believed it. Kripke took it apart with three kinds of argument.

The modal argument, and the rigid designator. Start with a simple test sentence: Aristotle might not have taught Alexander. True, surely — Aristotle could have died young, or stayed home in Stagira. But if "Aristotle" just means "the teacher of Alexander," the sentence means "the teacher of Alexander might not have taught Alexander," which is either nonsense or false. Something has gone wrong, and Kripke's diagnosis became one of philosophy's most-used tools. Descriptions are flexible: in different possible worlds — different complete ways history could have gone — "the teacher of Alexander" picks out different people, or no one. Names are not flexible. "Aristotle" picks out the same man in every possible world in which he exists; when we say Aristotle might not have taught Alexander, we hold onto the very man and vary his career. Kripke coined a term for this: a name is a rigid designator — an expression that designates the same object in every possible world. Descriptions describe; names stick. That single distinction, between rigid and non-rigid ways of picking things out, runs through everything else he did.

The Gödel case: reference survives the failure of description. The modal argument shows names don't behave like descriptions; a second argument shows we don't need to know a correct description to refer at all. Here is Kripke's teaching centerpiece. Ask an ordinary person what they know about Kurt Gödel, and you'll likely get one fact: "the man who proved the incompleteness of arithmetic." Now run a fantasy. Suppose the proof was actually the work of an unknown Viennese named Schmidt, who died in obscure circumstances, whereupon his friend Gödel got hold of the manuscript and published it as his own. In this story, who does the ordinary person refer to when they say "Gödel"? The description theory has to answer: Schmidt — because Schmidt is the man who fits the description "prover of incompleteness." But that's plainly wrong. They are referring to Gödel, the fraud — and saying something false about him, namely that he proved the theorem. Your one identifying description could fail completely, and your name would still land on its man. Reference, Kripke concluded, doesn't route through what's in your head.

So how do names work? Baptism and chain. Kripke's replacement — he modestly called it a picture, not a theory — is causal and historical. Somewhere at the start, a thing gets baptized: parents name a baby, astronomers name a comet, a community's usage crystallizes around a label. From then on, the name passes from speaker to speaker like a relay baton, each user intending to refer to whatever the person they got it from was referring to. You can use "Gödel" successfully knowing almost nothing true about him, because you stand at the end of a chain of usage reaching back to the man himself. Reference is borrowed through community; no single mind has to hold the credentials. If this rhymes with something you've read recently, it should: Hilary Putnam (chapter 262) was arriving at the same externalist insight from the side of common nouns — Twin Earth, "meanings just ain't in the head," the division of linguistic labor — at the very same moment, and each man acknowledged the other. Names and natural-kind words alike are anchored to the world by history, not by mental descriptions.

The necessary a posteriori: Kant's distinction severed. Now for the deepest cut. Since Immanuel Kant (chapter 155), philosophers had worked with two distinctions treated as near-synonyms: a priori versus a posteriori (knowable without experience versus only through it) and necessary versus contingent (couldn't have been otherwise versus could have). Kant and two centuries of successors assumed the pairs line up: the necessary is the a priori. Kripke performed what may be the cleanest conceptual surgery of the century: the first distinction belongs to epistemology — the theory of knowledge, a fact about knowers; the second belongs to metaphysics — a fact about reality. There is no law that they coincide, and they don't. Take Hesperus, the evening star, and Phosphorus, the morning star. Astronomers discovered — a posteriori, by observation — that they are one object, Venus. But "Hesperus" and "Phosphorus" are both rigid: each tags Venus in every possible world. So there is no possible world where Hesperus isn't Phosphorus; the identity is necessary, yet it took astronomy to learn it. (That true identities are necessary had been proved formally by Barcan Marcus decades before; Kripke's semantics and examples made the world believe it.) Likewise water and H2O: chemistry didn't record an accident, it uncovered what water is — a necessity found in a laboratory. Kripke even completed the chiasmus with the contingent a priori: whoever fixed "one meter" by the standard meter bar in Paris could know a priori that the bar was one meter long, though the bar could have been longer — a contingent fact known from the armchair.

Essence, respectable again. Once necessity is a feature of things rather than of our descriptions, an old Aristotelian idea walks back in the front door: things have essential properties — features they couldn't exist without — and accidental ones they merely happen to have. Kripke argued that origin is essential: this very wooden table could not have been made from a different block of wood (a table made from other wood would be a different, exactly similar table), and you could not have come from different parents. W. V. O. Quine (chapter 236) had spent decades arguing that such talk was incoherent — that necessity made sense only relative to how a thing is described. Kripke's possible-worlds semantics answered the scruple with mathematics: give modal talk precise models, make names rigid, and "this object is necessarily thus" comes out exact and evaluable. The metaphysics Quine had shown the door returned with formal credentials.

The mind-body cartridge. Kripke ended the lectures by firing his machinery at the identity theory of mind — the then-dominant materialist view that pain just is a brain process, say the firing of C-fibers, the way water is H2O: a contingent scientific identity. Kripke's semantics forbids the escape hatch. "Pain" is rigid; so is "C-fiber firing"; and identities between rigid designators, if true, are necessary. So the theorist must say pain couldn't exist without C-fiber firing. Yet it certainly seems possible — a world of pain without those fibers, or fibers firing with no pain. For heat and molecular motion, Kripke showed how to explain such an appearance away: what's really possible is having the sensation of heat without molecular motion, since we pick heat out by how it feels, and that feel could have accompanied something else. But run that move on pain and it jams — because pain is nothing but how it feels. Anything that feels like pain is pain; there's no gap between the appearance and the thing. So the appearance that pain and brain process could come apart can't be explained away, and stands as evidence they're not identical. Together with Thomas Nagel's bat (chapter 281), this argument was the 1970s' anti-materialist artillery, and its direct descendant is David Chalmers's zombie argument and the "hard problem" of consciousness.

Kripkenstein: the skeptic about meaning. In 1982 Kripke published his second bombshell, Wittgenstein on Rules and Private Language — his reconstruction of a paradox he found in Ludwig Wittgenstein (chapter 208). You've added numbers all your life. Now compute 68 + 57. You say 125. A skeptic asks: what fact about you makes it the case that by "+" you ever meant addition — rather than quaddition, a rule that agrees with addition for every number below 57 and outputs 5 otherwise? Your past calculations won't settle it: they were finite, and all involved smaller numbers, so they fit both rules perfectly. Your dispositions won't settle it: they're finite too, they include mistakes, and anyway a disposition only says what you would do — while meaning addition is about what you ought to answer. That's the normativity gap: no pile of descriptive facts about you seems to add up to a standard of correctness. The skeptic's conclusion is vertiginous — there is no fact as to what you meant, ever, by any word. Kripke's Wittgenstein then offers not a rebuttal but a skeptical solution: stop looking for a meaning-fact inside the individual and look at practice. The community licenses saying "she means addition" of anyone whose answers march in step with everyone else's; meaning lives in shared use and mutual checking, which is why a rule followed in principle by one person alone — a private language — is impossible. Is this really Wittgenstein? Kripke himself flagged the question, offering the argument only as Wittgenstein "as it struck Kripke" — philosophers promptly nicknamed the hybrid author Kripkenstein — and scholars of the Anscombe-era executors' circle objected sharply that the historical Wittgenstein embraced no such skepticism. Either way, the book ignited fifty years of literature on rule-following and the normativity of meaning — and it has a sharp new edge: a system trained on finitely many examples faces the quus problem in industrial form, which is one sober way of asking what, if anything, a large language model means.

The technical monument. Beneath all of this stands the teenage work: Kripke semantics. A modal statement gets evaluated in a model — a set of possible worlds plus an accessibility relation telling you which worlds count as possible from which; "necessarily P" is true at a world if P holds at every world accessible from it. The jungle of rival modal systems that logicians had built blind suddenly became a zoo with labeled cages: make accessibility reflexive and transitive and you get the system S4; make it a total equivalence and you get S5; each axiom mirrors a shape the relation can have. The same framework tamed intuitionistic logic and the logic of provability, and computer science quietly adopted it whole: the model checking that verifies chips and protocols runs on Kripke structures, a silent debt the verification industry repays daily. And in 1975 Kripke did it again for truth: where Alfred Tarski (chapter 224) had exiled "true" to an infinite hierarchy of languages to dodge the liar paradox ("this sentence is false"), Kripke's "Outline of a Theory of Truth" showed how a language can contain its own truth predicate if that predicate is partially defined — grounded sentences get truth values through a fixed-point construction, while the liar simply never gets one — founding the modern era of paradox studies. Much more never left his desk: the 1973 Locke Lectures appeared only in 2013 as Reference and Existence, and a vast archive still awaits editing at the Kripke Center.

Before & After

Coming off Kwame Gyekye's Accra (chapter 283), the pendulum swings back to the analytic engine room — but notice that both chapters ask what a community's language commits it to; the questions travel.

Backward, Kripke gathers threads from across this book. Aristotle (chapter 25) supplied the essences he restored to respectability. John Stuart Mill (chapter 171) had insisted names denote without connoting — a minority line Kripke vindicated a century on. Kant (chapter 155) forged the a priori/necessary alliance that Kripke's surgery severed. Frege (chapter 184) and Russell (chapter 195) built the description theory he demolished — while keeping their puzzle (how can an identity be informative?) at center stage. Wittgenstein (chapter 208) provided the text out of which Kripkenstein was quarried. Tarski (chapter 224) built the truth hierarchy Kripke relaxed. Quine (chapter 236) posed the modal scruples the semantics answered. Barcan Marcus (chapter 254) tagged the names and proved identity's necessity first — the priority reckoning lives in her chapter. Putnam (chapter 262) was the revolution's other wing. Nagel (chapter 281) was the other barrel of the anti-materialist gun.

Forward, the fan opens wide. Most immediately: Kripke treated possible worlds as useful formal apparatus — ways things might have been, not places. The philosopher in the next chapter, David Lewis (chapter 285), looked at the same machinery and drew the breathtaking conclusion Kripke refused: the other worlds are real — flesh-and-blood universes as concrete as ours — and his counterpart theory rewrites Kripke's rigid designation accordingly. That collision is next. Beyond it: Chalmers and the zombies carry the mind-body cartridge into the present; Kit Fine leads a revival that asks whether essence is even deeper than necessity; the "two-dimensionalists" try to reunite what Kripke cut apart; and every semantics textbook and model-checking suite ships with Kripke models inside.

Why It Still Matters

Every time software verification proves a chip design or a flight-control protocol can't enter a fatal state, the proof runs over a Kripke structure — the teenager's semantics, industrialized. The devices this book was written on were checked by his mathematics.

The causal theory of names quietly explains features of your daily information life. Why can everyone talk about a person online while believing mostly false things about them? Because reference doesn't depend on true description — the name travels the social chain regardless, which is exactly why misinformation about someone is still about them, and why correcting the record is possible at all. Courts lean on the same insight: what a trademark or a name refers to is fixed by history of use, not by what any individual believes.

The necessary a posteriori reshaped how we talk about science: chemistry and genomics read naturally as uncovering what things essentially are — water's composition, a species' lineage — not just cataloguing regularities. And wherever people argue about whether some category (a species, a disease, a kind of person) has an essence or is a social construction, they are fighting on ground Kripke resurveyed.

And Kripkenstein has become uncannily contemporary. The plus/quus skeptic asks how any finite history of examples can determine a rule — which is now a live engineering question, since a neural network's training data underdetermine what function it has learned, with the deviant cases waiting beyond the data like quus beyond 57. When alignment researchers worry that an AI has learned "something that agrees with what we wanted on the examples we checked," they are living inside Kripke's 1982 book.

Over to You

You learned that water is H2O from a book, not a flask — yet Kripke says it's a necessary truth discovered empirically. Can you get comfortable with truths that couldn't be otherwise but could never be reasoned out from the armchair? What else might be like that?

Could you have had different parents? Kripke says no — a child of different parents would be someone else. If that's right, how much of you is essential, and where exactly do the accidents begin?

The skeptic says nothing about your past fixes what you meant by your words. If meaning really rests on nothing firmer than community agreement, what happens to the lone dissenter who uses a word correctly while everyone else drifts?

Kripke argued the appearance that mind and brain could come apart cannot be explained away. Do you trust that appearance — or do you suspect it's exactly what a brain would feel like from the inside?

Where to Go Next

Start with the source: Saul Kripke, Naming and Necessity — three transcribed lectures, conversational and surprisingly readable; philosophy's rare revolution you can actually hear happening.

A modern way in: John P. Burgess, Saul Kripke (Polity, Key Contemporary Thinkers) — a clear, honest guide to the whole system, technical monuments included.

The thread forward: Kripke used possible worlds as a tool and left their nature discreetly open. Next comes David Lewis (chapter 285), who took the tool literally — every possible world as real as this one — and built the boldest metaphysics of the century on it.