One repair is easy; every repair is not
Suppose one plank in an old boat rots and a carpenter replaces it. Calling the result a new boat would make routine maintenance oddly destructive. Replace another plank next year and the same answer seems natural. Nothing dramatic happens between one repair and the next.
Continue long enough, however, and no original timber remains. The vessel may retain its name, route, owner, registration, crew, and scratches in roughly the expected places, but all its material has changed. The Ship of Theseus asks whether it is still numerically the same ship: one object existing at different times, not merely a later object that looks and works like the first.
Plutarch reports that Athenians preserved the ship associated with Theseus by replacing decayed wood. He says philosophers used it to dispute whether a changing object remains the same. The historical boat gives the puzzle a memorable setting, but the problem does not depend on accepting every part of the Theseus legend. Any sufficiently patient repair shop can produce it.
There is no plank carrying the identity
The first pressure comes from gradual change. If replacing plank one preserves the ship, and replacing plank two does too, each individual step appears harmless. Yet the end state differs completely in material from the beginning.
This resembles a vague-boundary problem, but identity makes the stakes sharper. A heap can lose grains without offering a precise last moment of heapness. The ship raises a second question: has one object continued, or has a sequence of replacements become increasingly convincing?
Similarity cannot answer. Two factory-made mugs may match in every visible respect while remaining two mugs. Numerical identity obeys a stricter rule: if A is identical to the original ship and B is also identical to it, then A and B must be identical to each other. That rule becomes important as soon as a second vessel appears.
Hobbes sends the old planks back to sea
Thomas Hobbes sharpened the case by imagining that someone saves every removed plank and later rebuilds them in their original arrangement. Now two boats sit in the harbor.
The maintained vessel has continuous history. It stayed in use while new pieces entered an existing structure. The rebuilt vessel has original matter. Its wood sailed on the first voyages, even though the collection spent years as a pile before reconstruction.
Both claims sound plausible because matter and history usually travel together. The thought experiment pulls them apart. They cannot both deliver numerical identity to different objects, so an answer needs a principle more specific than “it is basically the same.”
This is a philosophical puzzle, not a formal contradiction. Unlike Russell’s paradox, it does not derive both a statement and its negation from one rule. It shows that several ordinary rules for using “same” can point in different directions.
Matter remembers, but it is not enough by itself
Original material carries evidence. Old wood may preserve tool marks, damage, and traces of use that a replica lacks. Museums, conservators, and courts can have good reasons to care about that history embodied in matter.
A strict material rule is still costly. If changing one essential piece creates a numerically new object, a repaired roof becomes a sequence of roofs and a hammer may die when it receives a new handle. Living bodies continuously exchange material, making persistence through ordinary biological change difficult to explain.
Material also needs arrangement. A pile of authentic planks is not yet a ship. Reassembling them restores a form, but form alone cannot identify an individual either. Ten vessels built from the same plan share a structure and function without becoming one enormous ship located in ten places.
The distinction between an object and its matter is the territory studied under material constitution. A statue and the lump of material composing it may occupy the same space, yet appear to have different conditions for survival: melting destroys the statue while the material remains.
Continuous history explains repair—but can branch
The maintained ship has a strong causal claim. Each stage develops from the previous one through ordinary repair, use, and record keeping. No rival object replaces it overnight. Ownership documents, crew habits, and public recognition all follow the connected path.
This explains why gradual maintenance differs from theft followed by an exact duplicate. It also explains the identity of organizations. A company can keep contracts and institutional records after every employee has changed. The continuing entity is not a fixed collection of people.
Continuity can survive pauses: dismantling a ship for restoration need not destroy it. But the longer and more complete the interruption, the less automatic the judgment becomes. Hobbes’s rebuilt ship creates the hardest case because causal history appears to branch. One path follows the working vessel; another follows the original matter.
Some philosophers model persistence with temporal parts. The ship at launch, halfway through repair, and after the final replacement are different time-slices of one object extended through time. This makes change easier to describe, but branching still demands a decision about which later path belongs to the earlier whole.
The kind of object changes what continuity matters
Hobbes’s own examples suggest that “same matter” and “same thing of a kind” can answer different questions. A river keeps its name and course while its water changes. A commonwealth can continue as citizens enter and leave. A ship can remain fit for sailing through repairs even though its matter is different.
Software offers a modern version. An application may receive thousands of edits until no original line remains. We often preserve its identity through a release history, maintainers, compatibility promises, and the role it serves. Fork the code into two projects and the question changes: both descend from the earlier program, but neither can acquire exclusive identity merely by being similar.
That is also why a computed object need not be unreal. Our article on whether we are living in a simulation considers how digital objects and relationships could be genuine within their world. The Ship of Theseus adds a narrower problem: even real things can have disputed persistence conditions.
People make the puzzle morally important
People change physically and psychologically. Biological accounts emphasize the continuity of a living organism. Psychological accounts emphasize overlapping memories, intentions, character, and other mental connections. Ordinary lives give both kinds of continuity enough overlap that the distinction can remain hidden.
Copying exposes it. Imagine two future people receive equally accurate copies of your memories and personality. Each remembers your childhood, recognizes your friends, and expects your plans. Similarity cannot select one: the successors are equally similar. Psychological continuity appears to branch, but numerical identity cannot turn one person into two people who are identical to each other.
This does not prove that people are boats or that identity is unreal. It shows why different questions need different standards. Medical responsibility may track the living body. A promise may depend on psychological continuity. Family, law, and personal concern may combine several relations rather than discovering one hidden identity particle.
The Ship of Theseus therefore earns no universal percentage—no plank number at which an old object automatically becomes new. Its useful answer is a request for precision. Are we tracking original matter, one causal history, a functioning artifact, an institution, or a person? Once those come apart, “Is it the same thing?” is only the beginning of the question.
