The Identity of Theseus's Ship

What makes a thing the thing that it is? At first glance, the question seems almost trivial. A chair is a chair because it is made in a certain way, has a certain structure, and is ordinarily used for sitting. Yet the moment we ask what happens when its parts are gradually replaced, the apparent simplicity disappears. Replace one leg, then another, then the seat and back, and eventually none of the original material remains. Nevertheless, we are inclined to say that it is still the same chair.

This suggests that the identity of an ordinary object is not exhausted by the matter composing it. Function, structure, continuity, and convention all seem to play a role. A pile of books can temporarily function as a seat without thereby becoming a chair, while a chair displayed in a museum remains a chair even when nobody is sitting on it. What we call a chair therefore seems to depend not merely on what it is made of, but on the organization and history of its parts, together with the conventions by which we classify objects.

The ancient Ship of Theseus makes this problem impossible to ignore. Imagine that every plank of the ship is gradually replaced until none of its original material remains. Most people will still be tempted to call it the same ship, because there was an uninterrupted history connecting the original vessel to the final one. But now suppose that the discarded planks were preserved and later reconstructed into another ship. Which vessel is the original?

The problem becomes even stranger if we think about the individual components rather than the ship as a whole. If the original planks are distributed among several newly constructed boats, each of those boats contains some of the very material that once coexisted in Theseus's ship. Yet it would be difficult to say that all of them are numerically identical to that ship. Sharing constituents is clearly not sufficient for sharing identity. The identity of a complex object cannot simply be the sum of the identities of its parts.

Perhaps, then, there is no single ingredient that constitutes identity. Different kinds of things may persist according to different principles. The identity of a ship may depend heavily on structural and historical continuity, while the identity of a person may depend on something quite different.

Consider transplantation. If I receive someone's kidney, I do not become the person who previously owned that kidney. The organ becomes part of my body while my personal identity remains continuous. The same seems true of a heart, a liver, or most other organs. Their biological importance does not make them the carriers of the person's identity.

The brain presents a radically different case. Imagine that my brain is transplanted into another person's body, while the original body receives another brain. If my memories, personality, intentions, and conscious experience continue with my brain, there is a powerful intuition that I have gone with the brain. The body that previously belonged to me now seems to belong, in an important sense, to someone else.

The thought experiment therefore suggests that the identity of a person is not governed by the same criterion as the identity of an organ. The relevant continuity may be psychological rather than merely material. A kidney can change owners without transferring a person, whereas a brain transplant appears to threaten the very boundary between two persons because it carries with it the organization responsible for memory, personality, and conscious experience.

This does not prove that the brain is metaphysically identical with the person. It instead demonstrates that identity depends upon what sort of entity we are trying to preserve. Asking what makes a chair the same chair and what makes a person the same person may be asking two different questions under the same grammatical form.

The same puzzle appears in a surprisingly modern form when we ask whether a computer program exists. Consider a simple program stored on a computer. There is a physical arrangement of bits on a storage device, but there is also the program understood as an abstract sequence of instructions. If the file is copied perfectly to another computer, there are now two physical arrangements of bits, yet we ordinarily say that both computers contain the same program.

This is peculiar. The two files are not numerically the same physical object. One can be deleted while the other remains. The program, however, appears to survive the destruction of any particular copy. It can be instantiated again on another machine, perhaps thousands of times, without producing thousands of entirely different programs in the conceptual sense.

We can therefore distinguish between a program and its physical implementations. The particular arrangement of bits is a physical object, while the computational organization represented by those bits can be instantiated in many different places. The relationship resembles that between a musical work and its performances: the performances are physical events, while the work can be instantiated repeatedly without being exhausted by any single performance.

But this raises a deeper question. If a program is not identical to any particular physical copy, what kind of thing is it? Is it an abstract object, like a mathematical structure? Is it instead a pattern that exists whenever the appropriate physical system instantiates it? Or is the distinction between the program and its implementation merely a useful convention of description?

The Ship of Theseus, transplantation, and computer programs may appear to be unrelated puzzles, but they point toward the same possibility: there may be no single universal rule governing the persistence of every kind of thing. Material continuity is important for some objects, structural continuity for others, psychological continuity for persons, and computational or functional continuity for programs.

This also explains why function and convention seem so powerful in ordinary objects. A thing is not always defined by the material occupying a particular region of space. What matters can instead be the organization of that material, the role it performs, the history connecting its successive states, and the conceptual framework within which we identify it.

Perhaps identity is therefore less like a hidden label attached to every object and more like a relation we trace through change. We ask which later thing stands in the appropriate continuity relation to an earlier thing, and the answer depends partly on what kind of thing it is. A chair can survive a replacement of its wood; a person can survive a replacement of much of their body; a program can survive the destruction of a particular copy. In each case, something persists, but what persists is not necessarily the same matter.

The deepest lesson of these puzzles may therefore be that asking, “Is this the same thing?” is incomplete. We should first ask, “What kind of thing is this, and what kind of continuity would make it the same?” Once that question is asked, the paradox becomes less mysterious. The world contains matter, but the things we recognize within it are also structures, processes, functions, histories, and patterns. Identity may emerge not from any single ingredient, but from the particular form of continuity that allows us to say that something has remained itself while changing.

References

1. VSauce. Do Chairs Exist?.
2. Plutarch. Life of Theseus.
3. Locke, J. (1689). An Essay Concerning Human Understanding.
4. Parfit, D. (1984). Reasons and Persons. Oxford University Press.
5. Dennett, D. C. (1991). Consciousness Explained. Little, Brown and Company.