Category: consciousness
Type: thought experiment
Associated thinker(s): Hilary Putnam
Original publication or approximate period: Putnam, “The Meaning of ‘Meaning’” (1975)
Difficulty: Intermediate
Themes: meaning, externalism, natural kinds
Related experiments: Swampman; Beetle in a Box; Brain in a Vat
Imagine a planet called Twin Earth. Around 1750, before chemistry had been developed on either planet, its rivers contain a clear liquid with a different chemical composition from Earth’s water. Putnam calls the Twin Earth liquid “XYZ”; it looks and behaves like water, and ordinary speakers do not know its composition.
Compare a person on Earth with a Twin Earth counterpart in the same psychological state. Both use the word “water,” but one liquid is H2O and the other is XYZ. Putnam says the story does not depend on assuming that the two people are exact molecular duplicates.
Are the references of our words, and the contents of our thoughts, fixed entirely by what is in our heads, or can they depend on our environment and history?
Putnam argues that the extension—the kind of liquid picked out—by “water” differs between Earth and Twin Earth even when the speakers are in the same psychological state. The example challenges the idea that reference is fixed by that state alone. Putnam’s broader slogan that “meaning isn’t in the head” should not be read as saying that meaning simply is extension; the original example focuses on a natural-kind term and its reference.
The example became important to content externalism, the view that some mental contents are individuated partly by relations to the environment. This does not mean that mental processes are physically located outside the person. Putnam’s brain-in-a-vat argument is a distinct argument from 1981, though it draws on externalist ideas; it is not the same thought experiment as Twin Earth. That later argument and the scope of externalism needed to support it remain disputed.
Externalists argue that reference can depend on a term’s environment and relevant linguistic or causal history, not just on the speaker’s current psychological state. Internalist responses differ: some reject the externalist inference, while moderate views may accept broad content and also posit a narrower content shared by the twins. Two-dimensional semantics is one proposed approach, not a single agreed internalist solution. The significance of externalism for explanation, self-knowledge, and thought remains debated.
Putnam stipulates the XYZ liquid and the historical setting; the example does not establish a general rule that any chemically impossible liquid will work. A literal molecule-for-molecule duplicate also raises a complication because human bodies contain water; Putnam says the argument does not depend on imagining an exact duplicate. The argument is clearest for natural-kind terms and does not show that all meaning is environmentally determined or that minds extend beyond the body.
One can substitute other natural kinds, such as aluminum or elm, or vary how long a subject has lived in a new environment. These are useful extensions, but questions about the role of learning and history should be treated separately from Putnam’s original water example.
Putnam’s 1975 essay was consulted at the Twin Earth and 1750 passages. Relevant sections of the cited SEP entries were read for the scope of content externalism, narrow-content responses, and the separate brain-in-a-vat argument. The article distinguishes Putnam’s original example from later extensions; it does not claim that externalism or its implications are settled.
Research check: 2026-11-01. Claim-level evidence and access limits are recorded in the source-review ledger.