In 1973 the philosopher Hilary Putnam asked his readers to imagine a planet exactly like Earth in every respect but one. He called it Twin Earth. Everything is identical to ours — same continents, same languages, same history, same people with the same memories and the same brains. There is only one difference. The clear, drinkable liquid that Twin Earthlings call "water" is not H₂O. It is a complicated chemical, looking and tasting and freezing and boiling exactly like water, that we will abbreviate XYZ.
Now Putnam asks the question: when an English speaker on Earth says "water" and an English speaker on Twin Earth says "water," do they mean the same thing?
The thought experiment looks innocent. The argument that follows shattered an assumption philosophers had relied on for centuries — that the meaning of the words you use is determined by what is going on in your head. Putnam's slogan for the conclusion has become famous in twentieth-century philosophy: "meanings just ain't in the head."
The setup, in more detail
Putnam's Twin Earth essay (originally "The Meaning of 'Meaning,'" 1975) asks us to imagine a moment in 1750, before chemistry was advanced enough for anyone to know that Earth's water was Hâ‚‚O or that XYZ was anything other than water. In 1750, an Earthling Oscar and his Twin Earth duplicate Twin-Oscar are molecule-for-molecule identical (we adjust the body chemistry slightly for the sake of the puzzle). They have the same brain states, the same beliefs, the same memories, the same uses of the word "water." They both call the stuff in lakes and rain and drinking glasses by that name.
But what they are referring to is different. When Oscar says "Bring me a glass of water," he is asking for Hâ‚‚O. When Twin-Oscar says it, he is asking for XYZ. The two requests, if they could be honored across worlds, would be honored differently.
Putnam draws the conclusion: meaning cannot be just a matter of what is in the speaker's head. Two heads can be in identical states and still mean different things. What you mean depends in part on the actual stuff in your environment that your words latch onto.
Why this overturns a long tradition
For most of the modern period, philosophers had assumed something like the following picture. To know what a word means, you ask the speaker. The speaker has some concept in mind — a definition, a mental image, a cluster of associations. That mental content is the meaning. Frege and Russell each refined versions of this picture. So did much of mid-twentieth-century philosophy of language and mind, often in the form of descriptivism — the view that what a name or term refers to is whatever fits the description the speaker associates with it.
Putnam's thought experiment shows that descriptivism cannot be right, at least for natural-kind terms. Oscar in 1750 cannot describe water to the level that distinguishes H₂O from XYZ. He doesn't know any such description. Yet his word "water" still picks out H₂O and not XYZ. The reference is not riding on the description. It is riding on something else — the actual physical stuff in his environment, plus the social practice of using the word that links his usage to that stuff.
This is what philosophers now call semantic externalism: the meaning of at least some terms is partly external to the individual mind.
What externalism actually claims
It is easy to overstate the conclusion, so it is worth being precise.
It does not say that words have no relation to mental states. Twin-Oscar still has the concept of a clear drinkable liquid; he just lives in a world where that concept latches onto XYZ rather than Hâ‚‚O.
It does say that two people in identical mental states can refer to different things, and that purely internal descriptions cannot, in general, fix reference. The world helps fix it.
It generalizes. Putnam, and Saul Kripke in Naming and Necessity (1980), extended this kind of reasoning to natural kinds (gold, tigers, electrons), proper names, and indexical terms. Tyler Burge's Individualism and the Mental (1979) extended it to social-linguistic facts: what you mean by "arthritis" is partly determined by how the medical community uses the word, even if you yourself are confused about it.
Your words borrow their grip on the world from a community and a world you did not invent.
That is a claim with consequences far beyond the seminar room.
Why it matters outside philosophy
Philosophy of mind. If meaning is partly external, then mental content is partly external too. This complicates any project trying to explain the mind purely as something happening between your ears. Whatever your brain is doing, what your thoughts are about may depend on what your environment contains.
Science and language. When chemists discovered that water is Hâ‚‚O, they did not change the meaning of the word "water." They discovered something that had been there all along. Ordinary speakers were already referring to whatever-this-substance-is, even before they could define it.
Communication and trust. Putnam called the underlying social fact the division of linguistic labor. Most speakers do not know exactly what gold is or what arthritis is; they defer to specialists who do. Language works because we are willing to be partially wrong, partially deferring, partially borrowing — a more humble picture of speech than the modern individual self likes.
The lesson
Twin Earth does not require any actual planets. It is a thought experiment — a controlled philosophical scenario designed to test an assumption by pulling on the variables one at a time. The thought experiment succeeds: it produces a robust intuition (the two Oscars mean different things) that the prior theory could not accommodate.
What it leaves behind is a humbling picture of language. You did not build the meanings of the words you use. They came to you embedded in a community, hooked into a world, refined by specialists you have never met. When you speak, you draw on a well you did not dig. Putnam's slogan keeps its bite because the alternative — that meaning is wholly your private property — is the one we keep, against all evidence, slipping back into.
