7

Ship of Theseus is a thought experiment in which every piece of a ship kept in a harbor is replaced one at a time. The questions are: would the end result be the same ship or a new ship? If it is not the same ship; at what point did it stop being the ship of Theseus? If it is the same ship; would reassembling the removed pieces to form another ship result in another ship of Theseus? Surely the ship would not be in two places at once?

Rephrasing the thought experiment to improve relevance: our bodies grow and change. The old, worn out cells are replaced one at a time. The atoms that make up our cells and bodies too are replaced. Is the end result the same human from before or a new human?

ActualCry
  • 1,979
  • 6
  • 26

9 Answers9

8

The Ship of Theseus is one of the more illuminating thought problems in philosophy, and it having been around for something like 2500 years indicates how insightful the early Greek philosophers were.

What the thought problem shows is that if one takes a reductive approach to identity, where identity is taken as identicality down to the reduced elements of a structure, then NOTHING in our physical universe, except perhaps for some elementary particles, are identical over time. This presents a major problem for reasoning, as classical reasoning is only valid if A ≡ A, and this thought problem shows that A ≢ A.

One major approach to identity taken based on this A ≢ A problem, is to accept that there IS no valid identity in our world. That objects in our world are heaps, composites, without any essence. Therefore identity is only an approximate state, used for pragmatic purposes, and logic, while often giving useful answers, COULD be invalid when applied to these soft-edged "A"s.

Hume was an articulate advocate of extending this "heap" or "bundle" theory of non-essence of things to selfhood. He pointed out that thoughts and experiences changed radically over a short time, and postulated that we are just a bundle of impressions. Selfhood today is generally identified more with memories, and personality inclinations, than with experiences, and these two are more stable than experiences, so Hume's radical variability moment/moment is not widely accepted today. But these other two also both change over time, hence his point that A ≢ A for selfhood is still very true. Hume lived in an era when logic was considered more absolute than it is today. And the Pragmatic alternative, where some ideas that are invalid in absolutist terms might still be "good enough approximations to get by" was still a few centuries off. So Hume followed the absolutist thinking of classical logic about selfhood -- if A ≢ A for selfhood, then in classical terms assuming "self" is a fallacy.

But in a more pragmatic worldview, a selfhood that is not reductively identical moment to moment, and is appreciably different year to year, could still be a useful item to identify and operate with in the world and with reasoning.

There are alternatives to the logical "objects are not real" and the pragmatic "both objects and reasoning are useful but not fully reliable". One of those is available to non-physicalists, where one can postulate an essence to an object, over and above its reductive elements. This essence could be part of an abstract realm such as with Plato's Forms, or could be a feature of consciousness or spirit. An example of the latter is the dualist philosopher Richard Swinburne who postulates that selfhood has an essence, a "thisness" that is maintained over time, despite changes in personality/memory etc. Swinburne's essence is clearly a soul, although he does not describe it as such.

Another alternative is to dispense with the reductionism that Hume and The Ship of Theseus assume is what matters for identity. The philosophy of science today has accepted that there are emergent structures in our universe, and these emergent structures are functionally independent of their reduced components. One can then tie identity to functions that an object performs, rather than to its constituents. Once one does that, there would then still be a Ship of Theseus, and it would be the continually repaired vessel floating in the Thebes harbor. And a set of discard boards, reassembled -- would then NOT be the Ship of Theseus. The function that defined which was THE ship would then be the continuity of the repair crew activities. Per this approach, if the repair crew were to leave, and the ship boards were broken and scattered, then a new crew reassembling and repairing as best they could, would still not be able to make THE Ship of Theseus, as they would never be sure of the provenance of the components and design. But add in pragmatism, and Ship of Theseus V1.1 might be a good enough stand in.

Emergence and functionalism lead to pluralism in science. And pluralism of reference frames — leads to logic explosions. Therefore, a noteworthy outcome of this thought problem is that for all three of the bundle/heap/no-objects, functional/emergence/pluralist, and the pragmatic "identity is approximate" answers, classical logic cannot be relied upon in our world. The only ways that applying logic to things in our world would still be fully valid, would be if they have an essence -- and Plato's Forms and their equivalent are widely rejected today -- with the sole exception of possibly selfhood.

Dcleve
  • 13,610
  • 1
  • 14
  • 54
  • 1
    Thank you. One thing to consider is that elementary particles are considered 'identical' in the sense of 'indistinguishable': able to be substituted one for one with no change being visible. Reasoning requires that things be explained in terms of other sorts of things, so an explanation of variation can only rest on a platform of sameness. Reductive explanation forces us to conclude that there are no unique things. Otherwise, all explanations would have to be infinitely long. – Scott Rowe Apr 23 '22 at 13:37
  • Yes, elementary particles are the only exception to objects in our world NOT being reductively identical. At least we think not. The partial entanglement that sometimes persists between them has a possible logic effect on this that I have not fully parsed through, as a possible caveat. our declaration that reasoning requires reduction is in opposition to the consensus view of philosophy of science today, which is pluralist. See SEP Scientific Reduction, section 5. The pluralism of logic also makes any assertions about "logic must" invalid. – Dcleve Apr 23 '22 at 17:17
  • @ScottRowe Here is a link for logical pluralism: https://www.cambridge.org/core/journals/think/article/abs/guide-to-logical-pluralism-for-nonlogicians/EDFDFA1C9EB65DB71848DABD6B12D877 – Dcleve Apr 23 '22 at 17:17
  • 1
    @ScottRowe "'identical' in the sense of 'indistinguishable'" means to me having no identity. I have a (separate) identity because I can be distinguished even from my ("identical" ;-) ) twin brother, and because (hello, Descartes) I know I'm not him. Elementary particles have no identity in this sense, at least not after they have come close enough in space and time to each other. – Peter - Reinstate Monica Apr 23 '22 at 20:22
  • @Peter: Indeed, it has been semi-seriously proposed that there's only one electron, although this usually isn't taken too seriously because there are several physics-related objections to it (for example, electrons greatly outnumber positrons, which contradicts this hypothesis because positrons are equivalent to electrons moving backwards in time, and similarly we can identify scenarios where an electron worldline forms a closed loop, or appears to). – Kevin Apr 23 '22 at 20:35
  • @Dcleve Isn't the other side of "having been around for… 2,500 years indicates how insightful… Greek philosophers were" simply that there was nothing special about the Question; rather that 2,500 years-worth of students have not been insightful? – Robbie Goodwin Apr 24 '22 at 00:54
  • A modern variation that questions "selfhood == memories" is the "Star Trek transporter duplication" problem. – Barmar Apr 24 '22 at 02:19
  • @Peter-ReinstateMonica I didn't know you had an identical twin brother! Well, the difference between you two is your world line, since even if you go everywhere together (you're not Siamese twins though) you still don't occupy the same volume of space. This answer is fairly physical and obvious, and doesn't require introspection. – Scott Rowe Apr 24 '22 at 02:50
  • The Pluralism thing is interesting. My usual response when 2 people can't agree is to say that they are both wrong. Because if you can't give a convincing argument for a correct position, then you must not understand it. Any good explanation should be just so bone obvious that it is inarguable. – Scott Rowe Apr 24 '22 at 03:04
  • @ScottRowe Different world lines distinguish two electrons as well, until they come so close that their world lines become fuzzy and indistinguishable. One could argue that they only can do that because the electrons don't have separate identities... – Peter - Reinstate Monica Apr 24 '22 at 04:33
  • @Peter-ReinstateMonica So, don't get to close to your twin brother. You always want plausible deniability, in case he turns out to be an evil twin. I always keep my car away from other identical cars... Actually, it is a quite rare model for some reason. If you're worried about identity, perhaps the best thing is to be really unusual? – Scott Rowe Apr 24 '22 at 12:21
  • @ScottRowe Of anecdotal interest in this discussion is a jewelry heist in a famous Berlin department store committed by one of two (well) identical twins. They couldn't tell them apart by the DNA traces and had to let them go. – Peter - Reinstate Monica Apr 24 '22 at 12:30
  • @Dcleve "This presents a major problem for reasoning, as classical reasoning is only valid if A ≡ A, and this thought problem shows that A ≢ A." No. The law of identity does not say that A at t₁ and A at t₂ are identical.. – Speakpigeon Apr 24 '22 at 16:32
  • @Speakpidgeon -- if A at t1 is not identical to A at t2, then A is not equal to A, and one cannot reliably use A in logic operations. Since this aspect of A is true for any object in our world, that makes logic unreliable for our world, and the objects in it. – Dcleve Apr 24 '22 at 16:44
3

It comes down to definitions of words. If we say that the ship is the collection of atoms that it’s made of then replacing Aunty part of it means that the ship is divided into different parts that we can identify and track. Once all the parts have been replaced, whatever we have left is no longer in any way the original ship. On the other hand, if we say that the ship is an abstract idea embodied in a physical thing then replacing parts of the ohms thing doesn’t degrade the abstract idea and so the ship remains regardless of the replacements. So when you ask such a question you should specify how you choose to recognise a ship.

Frog
  • 655
  • 3
  • 6
2

"No man ever steps in the same river twice, for it's not the same river and he's not the same man." Heraclitus (apparently)

As @Frog suggests it is a matter of the meaning of words, and in this case what is meant by identity. If I talk about the river Wensum it has an identity that doesn't depend on the particular molecules of water that are flowing through it at any particular moment, but there is also a sense in which it has an identity as the state of the river Wensum at a particular moment. There is no great mystery here as long as we are clear what we mean by "ship" or "human".

Like the Wensum, the Ship of Theseus has an identity as the ship belonging to Theseus that remains the same throughout its life, regardless of how many components are changed. But that does not mean that the Ship of Theseus is remains identical (in a different sense of the word) throughout it's life.

With people it is more difficult as then identity can also mean your conscious identity, rather than just the person as an individual amongst others. This is not necessarily fundamentally encoded in the atoms, molecules or cells of your bodies, but in your "brain state", and arguably you might retain the same identity if that state were transferred to a simulation of your physical brain on an ultra-powerful computer (perhaps this has already happened ;o). This means the analogy with the Ship of Theseus a bit weak as there is an additional dimension to identity that is not present in a ship. Your body does not define your conscious identity in the same way that it defines your physical identity ("ah there's Dikran Marsupial"), which is the sense that seems to be being used with the Ship of Theseus.

So perhaps it is more a question of defining what we mean by "identity" (or at least being specific about which meaning we are talking about). Perhaps it is a matter of levels of abstraction. It may be that nothing is completely identical in an absolute sense, but they can be the same at a higher level of abstraction. The Wensum is the same river from one day to the next if we are talking about the river as a geographical feature, even if it isn't the same river in the sense of Heraclitus. This isn't a fundamental problem for logic or anything else AFAICS, it is just a problem with the ambiguity of natural language that we need to deal with.

Dikran Marsupial
  • 2,118
  • 10
  • 15
1

Here is an answer from Thomistic philosophy.

All things have matter and form but God who is actus purus, pure form.

The wood constitutes the matter. The ship constitutes the form. A change in matter without destruction of the matter consitutes no change in substance. The ship is still a ship--it doesn't matter if it is made of wood or metal.

If we are talking about some sort of "spirit" which the ship has, i.e. the Ship of Theseus as a unique vessel or substance, then whether or not it remains the same vessel depends on the definition of its own substance considered in itself, and not merely as a ship.

If we are asking about the substance of the Ship of Theseus, and not whether it remains a ship after a change in matter (which we are not), then iff the Ship of Theseus [SoT] must be made of Cyprus wood, and iff the SoT must be made of only the first planks it was made of and no other (including any defects that are introduced), then merely applying a graft to, say, a hole produced by a cannonball, is sufficient to make it a different ship than the SoT and is not merely an accident because the substance is directly tied to the original construction completely such that no new construction can be considered part of it.

I think everyone would laugh if the three ships that sailed into the New World were considered to have their unique substance destroyed because of a bit of patchwork, so on this basis I would claim that if we define the unique substance to be based on origin such that we can say, "The SoT was made of this, but now is made of that," then the default answer I would give by this definition is, "no, it is still the same ship whose appearance only has changed over time, but not its substance".

For a more objective answer, what we conclude is that this is an incredibly subjective question. Without some objective substance to analyze or examine, it is impossible to determine what it is, so the best, most morally certain answer, really is, "inconclusive: insufficient data".

AMDG
  • 119
  • 1
  • As an additional consideration: if the substance is defined by the name itself, "Ship of Theseus", then it has accidental being within the ship, so the substance would be destroyed by simply changing the name, e.g. "Ship of Harkinian", or by destroying the ship, e.g. burning it to ashes. This would also mean that any ship can become the Ship of Theseus by simply naming it that. – AMDG Apr 23 '22 at 14:51
  • "We had to destroy the ship in order to save it." Apparently... The answer I would give would be in terms of a world line. The baby born long ago was me because at every moment since, there has been no divergence. This is the commonsense answer also supported by Physics. Not sure why it is still a subject of debate. – Scott Rowe Apr 24 '22 at 02:43
  • @ScottRowe I don't quite follow. This is rather vague and appears unrelated to my answer. – AMDG Apr 24 '22 at 15:57
1

Suppose Alfred owns a ship moored in the port of Honfleur in France. Suppose Alfred has enough money to pay for repairs to be carried out every day so that a bit of the ship is replaced every day and the ship is kept pristine all the time. Suppose every bit of the ship is replaced at one point or another within one year.

At the beginning of the year, Alfred owns the ship. At the end of the year, it is still Alfred who owns the ship.

Apparently, nobody really has any difficulty identifying the ship as Alfred's ship.

Speakpigeon
  • 7,364
  • 1
  • 11
  • 26
  • Pragmatically, logic works most of the time in our world, even though A ≢ A means it is subject to logic explosion. However, in exotic cases, the question of whether "that ship is Alfred's ship" does frequently come up, and leads sometimes to extensive legal disputes. For one of the more common such examples, if Alfred runs a ship chop shop, there will be LOTS of questions about whether that ship is really Alfred's. – Dcleve Apr 25 '22 at 15:26
  • @Dcleve Sorry, I can't make sense of you comment. Logic explosion?! You sure?! – Speakpigeon May 11 '22 at 17:27
  • Any contradiction in a logic system can theoretically be taken advantage of by an inventive arguer to assert the truth of either side of every claim. This is a problem for logic systems, and is called a logic explosion. – Dcleve May 11 '22 at 17:57
  • @Dcleve 1. No, contradictions are just false and nothing therefore follows from any contradiction. - 2. What you mean is really called "the principle of explosion", not "logic explosion". The principle of explosion is just crap. – Speakpigeon May 11 '22 at 18:05
  • You refer to both "Alfred" and "the ship" as items that last thru time. But those identifiers do not satisfy A=A. You use the rationale that they are close enough. Which is also true, most of the time, but not A=A, and is subject to logic contradictions as a result. – Dcleve May 11 '22 at 19:01
  • @Dcleve Alfred is Alfred and the ship is the ship and no problem. Time is entirely irrelevant to the law of identity. You misunderstand it completely. You're not the only one. The universe is supposed to be 13.7 years old and yet the universe is the universe. This is no problem because we use the same word "universe" to refer to the universe at any one point in time and indeed to the universe throughout time. Same thing with every entity in the physical world, in particular people. Concepts are not thought of as existing in time. The law of identity works the same with or without time. – Speakpigeon May 12 '22 at 17:07
  • Both Alfred and the ship and the universe are not identical moment to moment, hence A=A is not true for them over time. They all exist in time, so delta time is intrinsic to each of their definitions. Therefore using classical logic on them, because delta time is included, violates the criteria for classical logic to be valid, and subjects one to a logic explosion. Your just declaring this not to be a problem, does not actually address the problem. If you claim that you are not applying classical logic to object that are relevant to this world, that is simply not true. – Dcleve May 12 '22 at 17:15
  • @Dcleve You just misunderstand and misrepresent the law of identity. I tried to explain, but you are clearly not interested. Conversation finished. Have a good day. – Speakpigeon May 12 '22 at 18:00
  • Speakpigeon -- I understood your claim -- that logic works in virtual logic space where A=A. But this does not address the problem, which you appear not to understand. There is a logic Alfred, A(L), where A=A. But there is a real Alfred, A(R), where A=/=A. Applying the logic model to the real world, leads to failures, because A=/=A for real world objects, invalidating the the preconditions for the logic model to apply. Meanwhile, your "explanation" primarily consists of unsupported declarations, then dismissals of everyone who does not agree with them. – Dcleve May 12 '22 at 19:09
  • @Dcleve 1. No, you don't understand the identity principle. 2. "But there is a real Alfred, A(R), where A=/=A" The idea that you can know something about the real Alfred ("A=/=A") directly without going through an abstract model of it is absurd. All we can do is talk about the model. This is why we need a model. If we knew the real world directly, we wouldn't bother to use any model. I suspect you are just going to continue pretend you enjoy absolute knowledge of Albert--a fictional character!--that nobody else has. – Speakpigeon May 13 '22 at 10:10
  • Your claim that one cannot test and falsify models of the world, because our knowledge of the world is indirect, is simply false. Read Popper -- probably Conjectures and Refutations will spell this out, but it will be in most of his works. We use our more certain inferences to test the less certain. We know that Alfred and his boat do not satisfy A=A. and your logic model is false when applied to the world because it assumes that precondition. Neither Alfred nor his boat are fictional. You do not understand indirect realism. – Dcleve May 13 '22 at 14:56
  • @Dcleve "Your claim that one cannot test and falsify models of the world, because blah-blah-blah" My claim?! No, not my claim. Your invention. Have a good day. – Speakpigeon May 13 '22 at 15:56
  • Your confusion about the supposed necessity of direct realism in order to test hypotheses is explicit in your comment "you are just going to continue pretend you enjoy absolute knowledge of Albert". If everyone disagrees with you, there are other more likely alternatives than that everyone else is dim witted. Accepting the possibility of error and that others can correct them, is the primary way most of us learn in this world. – Dcleve May 13 '22 at 16:52
  • @Dcleve "Your confusion about the supposed necessity of direct realism in order to test hypotheses is explicit in your comment* "you are just going to continue pretend you enjoy absolute knowledge of Albert"" It is neither explicit nor implicit. You are just making that up. Your pretence to have knowledge of the real world is implicit in your comment "But there is a real Alfred, A(R), where A=/=A.*" Discussion terminated. – Speakpigeon May 13 '22 at 17:18
0

The ship of me probably does not contain a single atom that it contained when I was born. Does that matter? Do atoms have identity? If not, then one is the same as another. My consciousness has been discontinuous. Does that mean that it is illusory? I do not see that this follows. So is there an ongoing identity within me that endures, albeit continually updated? Panpsychism proposes that the fundamental stuff of the universe is matter-consciousness. So the interchangeability of particles to form entities with identity is irrelevant. I am still me.

Meanach
  • 2,337
  • 7
  • 36
0

A thorough analysis and a model of the Theseus ship paradox were presented in an article by Dinis:

Dinis, Bruno. Equality and near-equality in a nonstandard world. Log. Log. Philos. 32 (2023), no. 1, 105–118. https://doi.org/10.12775/LLP.2022.018

The analysis is somewhat similar to modeling the sorites paradox.

Mikhail Katz
  • 1,216
  • 5
  • 16
0

First of all, we must accept that everything changes over time. So trying to solve the identity problem by reference to the material aspect of the thing does not lead us anywhere.

The only candidate for identifying something, a being, is its behavioural aspect. If something behaves the same, then it is the same.

This approach though, may lead to a phenomenological paradox. A thing may be or may not be the same for different people, based on its behavioural aspect according to what each one is focused or interested in. For example the new/repaired ship of Theseus is the same for the sailors, and the reconstructed from the old pieces is the same for a collector.

But that's life. It isn't mandatory that we all agree on everything!

Ioannis Paizis
  • 1,706
  • 1
  • 16
0

The phrase 'the ship of Theseus' is inherently vague, as is the phrase 'the same ship'. Whether something is 'the same ship' or not, depends on what you intended to mean by the term 'the same ship'. If you intended it to mean the original ship identical to the nearest atom, then the ship stops being the ship of Theseus almost at once. At the other extreme, if you intended it to mean 'all or part of any ship-like object that currently or once belonged to, or was associated with, Theseus' then even a few planks recovered from the seabed by archaeologists might be labelled the ship of Theseus. There is a spectrum of possibilities between those two extremes, and the phrase 'the ship of Theseus' can be interpreted as you like to cover, or exclude, any subset of them.

Broadly, the same reasoning applies to people. Am I the 'same' person I was yesterday? That depends on what you mean by the same. If you mean the same to the nearest atom, then no, I am not the same person. If you mean am I a person who very closely resembles the me of yesterday and has largely the same deplorable collection of attributes, to the point at which there is no discernible difference, then I would say yes, I am the same person.

There are countless other question you can put in a similar vein. If I whistle 'Lady be Good' is it the 'same' tune conceived by George Gershwin? If you and I both suffer from having to use Microsoft Word, are we using the 'same' software? In all cases, it depends on what you mean by the 'same'.

The ship of Theseus 'paradox' is a classic example of the scope for argument at cross purposes arising from the ambiguity of language. It appears paradoxical only if you overlook the fact that a word such as 'same' is equivocal.

Marco Ocram
  • 20,914
  • 1
  • 12
  • 64