Ultimately, all that can be said is the following:
In one (objective) sense, the old ship is the Theseus.
In another (objective) sense, the new ship is the Theseus.
In a third (objective) sense, neither ship is the Theseus.
Any of these senses can be justified, by thinking of 'identity' in different ways. There is no unique answer to be discovered.
It is useful that it forces us to think, but it is meaningless in that it doesn't refer to anything.
The question is posed as if there is something intrinsic to the vessels that would cause one or the other to be identifiable, but identity is generated on the observers end.
I don't think it is that simple. The ship is built of entirely new materials. However, it is spatially and temporally continuous with the ship which consisted of the old materials. It is that spatio-temporal continuity that justifies saying of it that it is still the Theseus. The issue is an old philosophical issue; that of persistence through change.
I'm not quite following. What gives the ship with new parts this spatio-temporal continuity with the ship consisting of old parts?
Any of these senses can be justified, by thinking of 'identity' in different ways. There is no unique answer to be discovered.
But can we choose one definition of identity which: a) is generally useful and reasonable, and b) gives an answer to the question? Any definition I've thought up so far that does give a single definite answer to the problem comes into conflict with other understandings of identity that are more generally useful.
The problem is that, in language, usage precedes definition. It may sometimes be useful to regard one ship as the Theseus, sometimes the other, and sometimes neither. The word 'identity' therefore has to do at least three conflicting jobs, so it is logically impossible for it to have a single definition. The only solution I can think of is to specify the kind of identity meant, e.g. spatio-temporal, physical, continuous physical, absolute, or any other. 'Identity' on its own would be a generic term.
2. If, in addition to having its parts replaced one by one, the original ship is at some point temporarily dismantled, perhaps having its parts dispersed, does it still have spatio-temporal continuity after it is reassembled?
I'm happy to assume that the other two of the conflicting jobs may be given different words to identify them. But when you say 'spatio-temportal identity', 'physical identity', etc, are you not simply talking about properties of entities rather than the entities being discussed? I'm fine with the the sub-classes of identity discussed but then cannot a property of an identity be another identity?
Actually, spatio-temporal identity might better be thought of in terms of spacetime worldlines. If such worldlines diverge then converge with passing time... well why not keep them unified under one identity however disparate they become? It makes a lot less sense to keep the identity beyond separation of worldlines if they do not later converge again.
I think you are on an impossible quest here. 'Identity' is the property of being the same; but there is more than one way of being the same. (We are obviously not talking here about absolute identity, i.e. 'Theseus is identical to itself'.) Each of these ways of being the same is justifiable; there is no one 'real' way. If 'identity' is to have sub-classes, the word on its own must range over all the sub-classes; it cannot refer to a sub-class on its own, since it denotes the entire class. Therefore it must be generic. To give an analogy, there are male persons and female persons; there is not a third group simply called 'persons' (= 'real' persons).
But what if we do not know whether they will converge again? What if all the Theseus's parts are lying in a scrapyard awaiting disposal, but then some naval history enthusiast suddenly suggests reassembling it? Does the Theseus currently exist in that case?
In the case of the Theseus, we (perhaps) have an identity of the Theseus. Say the Theseus has its own special flag. That flag is unique, therefore has an identity. The flag is also a property of the Theseus.
What we're talking of here is different identities, or to put it another way identities of different entities.
My point being that, yes, while there may be a spatio-temporal identity corresponding to the Theseus, spatio-temporal location/extension is also a property of the Theseus and thus constitutes the Theseus identity.
I don't see any difficulty in this, and am not convinced this is an obstacle to defining a complete and unconditional Theseus identity. I think we simply have degrees of freedom in which to do the defining, namely in its continuity.
We can consider this in 3D. A group of islands, say the Farne islands, are spatially discontinuous. The water between the islands is not the islands. Yet we have no problem identitfying the group of islands as one single identity. If we have always lived on one island and no nothing of the existence of others, again this does not effect the objective identity 'the Farne islands'. This is merely an absence of knowledge of that identity.
Same goes for temporal discontinuity.
In that case we would be defining the Theseus as the ship (if any) that flies (or usually flies) that special flag. No problem there. But other people may wish to define it differently. There is no objectively 'right' definition.
I don't follow. What different entities? If there is to be one entity/identity uniquely called 'the Theseus', what are the other entities to be called?
Sorry, I do not know what you mean here. Can you please clarify. Are you saying that spatio-temporal continuity objectively/logically defines the Theseus identity? If so, I don't understand what you mean by "while there may be a spatio-temporal identity corresponding to the Theseus".
You mean that spatio-temporal continuity is crucial in defining identity, but that the definition of spatio-temporal continuity needs to be made precise? And that this process of achieving precision would itself be subjective?
But in some cases it may not be obvious which of two groups a particular island should be assigned to. Geography may suggest one answer, history another. Gibraltar is geographically part of Spain, but historically part of the United Kingdom (it has been a British dependency for nearly 300 years). Which country is it 'really' part of? Wars have been fought over such matters.
Are you saying that, in disputed cases of spatial identity, there is always one objectively true answer, which can be deduced by applying an agreed set of (subjective) rules?
But in extending to subjective resolution, I see no problem either. Gibraltar has well-defined geographical (i.e. spatial) properties. It also has well-defined continuity of political properties. If one wishes to include 'which country Gibraltar is part of', these objective properties may be considered in the objective identity.
Sure. I was continuing with the definition of identity as the totality of the properties that uniquely identify the entity.
I was saying that spatio-temporal identity, i.e. the identity 'the spatio-temporal extension of the Theseus', is not the identity 'the Theseus'.
I simply meant that extension is a necessary property of the identity of the Theseus. Whether you wish to consider the identity of the extension of the Theseus... well that's reasonable.
I meant that any continuity (formal, physical, semantic, purpose, ownership, etc, etc) could (and I think should) be part of the identity of any entity. However, in dealing with the continuity of the identity itself (e.g. whether the identity of the Theseus survives the repair and, if so, how so) seems to be open to suggestion.
However, the identity of that group of islands, however arbitrary, exists objectively, with or without a name.
But in extending to subjective resolution, I see no problem either. Gibraltar has well-defined geographical (i.e. spatial) properties. It also has well-defined continuity of political properties. If one wishes to include 'which country Gibraltar is part of', these objective properties may be considered in the objective identity.
On the subjective front, while the properties that uniquely define an entity may be subjectively chosen, thay may still objectively exist. This is the arbitrary nature of identity. The Universe has an identity and this isn't arbitary, but the universe can be said to consist of systems. Well, which systems (i.e. what are the identities of the systems)? There are infinite ways of subdividing the Universe but the identities of these arbitrary (and subjectively chosen) systems still objectively exist since they are uniquely specified by a complete set of objective properties.
I agree. (I take it that by 'extension of the Theseus' you mean 'extension of the ship originally called the Theseus'.)
I am confused by your terminology here. If identity has the property of extension, how can extension have identity? Can you please clarify.
1. If we decide to make physical continuity the criterion of identity, which ship is the Theseus while the repair process is incomplete?
2. Is any kind of continuity appropriate? If the 'King of England' dies and his son immediately becomes 'King of England', we have continuity of title and role, but nobody would say the old king 'is' the new king.
Isn't that just a tautology? (If the group of islands is there, then it exists.)
I should probably expand upon this. Any objective identity should be specified by only the minimum number of objective properties that uniquely define it. Of course, this has to be the minimum number that always uniquely defines that entity, possibly including in hypothetical situations such as that of the OP.
This causes a huge problem in considering questions such as this. If we do not ever consider a property that was not required to uniquely define the entity at a given time of consideration, we may struggle to identify it at another time when it is required. For instance, there might be a property that we have not considered that specifies exactly which ship is the Theseus.
Conversely, if we consider properties not included in the necessary minimum, we may struggle to identify the entity because we look for continuity in something that should not be considered. For instance, we may consider the physicality of the Theseus when this is in fact irrelevant, and thus be tempted to choose the wrong boat.
If I understand you correctly, you are saying that we should aim to specify the objective identity of the Theseus in such a way that there is always one and only one Theseus. But why must we do this?
Why can we not decide in advance which properties we consider relevant, and, if on analysis it turns out that at a certain time no ship possesses all those properties, conclude that there is no Theseus at that time? And if it turns out that more than one ship possesses all those properties, why can we not then pragmatically add further properties on an ad hoc basis until only one ship has them all?
I am not sure how the physicality could be 'irrelevant'. If we decide at the outset that physicality is relevant to our needs, i.e. that it is useful, then it is relevant. The choice is a subjective one. Likewise, I don't see how the ship we choose could be the 'wrong' one, if our analysis leads us to it.
1. According to your argument, we have one entity called 'the Theseus' and another entity called 'the extension of the Theseus'. The latter (like the former) has properties, which are also entities in themselves, and so on. Should we invoke Occam's Razor to get rid of this infinite regress? For the purposes of this thread at least, why do we need to consider extension as an entity? Can't we just consider it as a property?
2. Suppose that, after the total repair of the original Theseus and the reassembly of its old parts, the ship made from the old parts is then irrevocably destroyed. Later, the ship with the new parts is completely repaired a second time, and the parts from the first repair are reassembled as a third ship. Can we now call the third ship the Theseus on the grounds of physical identity? Or would this be inconsistent, as we relied on spatio-temporal identity at the first stage?
For instance, if there is only one ship in the world and it is called the Theseus, what it is made of doesn't jump out at you as being a defining characteristic. If it is made of soap, so be it. If another ship is built, let's say it is identical but made of straw, then physicality becomes relevant.
Identity 2: Form (inc. extension, exc. temporal) of (original) Theseus -> Both ships are the Theseus.
Whatever you define, you will get a valid answer, but you can't hold you definition to be better than any other.
If there is only one ship, then yes, physicality is irrelevant. But you stated earlier: '....we may consider the physicality of the Theseus when this is in fact irrelevant, and thus be tempted to choose the wrong boat'. This implies the existence of more than one, in which case it is always open to us to deem physicality relevant.
How can this be so? How can both ships be 'the' Theseus? Isn't it necessarily unique? I can understand how part of each ship can be part of the Theseus according to Identity 4, but not how both ships can ever be the whole Theseus.
I agree entirely. This has been my view right from the beginning of this thread. Objectively there exists a complex set of entities; it is up to us (individually or collectively) to decide subjectively how to group, categorise and name them.
