Same-kind coincidence and the ship of Theseus.
Suppose that, in the course of repairs, the planks a wooden ship is made of are all (gradually) replaced by duplicate planks, in such a way that the end result is a ship exactly like the ship we started with. Suppose also that each original plank is destroyed as soon as it has been replaced. Suppose, finally, that throughout the process of replacement of the planks we have a working ship. Then, it seems, the original ship still exists, albeit with a new set of planks, after all the original planks have been replaced.
Suppose on the other hand that a ship is disassembled, plank by plank, and that the planks removed from the ship are not replaced as they are removed. Suppose also that the disassembled planks are subsequently reassembled, in such a way that the end result is a ship just like the original ship, made of all the same planks (in all the same places). Again, it seems, the original ship still exists.
So, it appears, (i) a ship will survive total (gradual, structure-preserving) plank-replacement, at least in cases in which the replaced planks are destroyed upon removal. And (ii) a ship will survive the disassembly and subsequent reassembly of its planks, at least in cases in which the planks removed from the ship are not replaced by other planks. If that is so, I shall argue, two things of the same kind (viz., ships) can occupy the same place at the same time.
On the island of Nassos lives a sea-captain called Stathis. On that same island there is a nautical museum that includes in its collection the (surprisingly well preserved) ship of Theseus.(1) The museum was intended to be a magnet for tourists, but failed signally to attract any; so it has been closed for years. Although Stathis does not know it, the ship he owns and captains is, amazingly enough, a plank-for-plank duplicate of the ship of Theseus in the Nassos museum. As the years pass, Stathis notices that his ship needs a plank replaced here, and a plank replaced there. He accordingly gives his first mate, Klops, some money, and sends him off to buy the replacement parts needed to repair his ship. Unlike Stathis, Klops knows that inside the locked-up museum a perfect duplicate of the ship of Stathis is gathering dust. The unscrupulous Klops breaks into the museum, steals the appropriate planks from the ship of Theseus, and pockets the money Stathis gave him to buy the replacement planks with. After (unwittingly) putting in the stolen planks, Stathis destroys the original ones (say, using them for firewood). This goes on for years, until not a single plank of the ship of Theseus is left in the museum, and every plank originally in the ship of Stathis has been destroyed.
At this point, by (i), the ship of Stathis still exists, even though none of its original planks do. So does the ship of Theseus, by (ii). After all the planks that were originally in the ship of Stathis have been replaced by planks that were in the ship of Theseus, there are two ships--Stathis' and Theseus'--made of the same planks, and occupying the same place at the same time. Those ships, on the face of it, constitute a counterexample to the Lockean principle that only different kinds of things can be in the same place at the same time. (For brevity, I shall henceforth call this principle the Lockean principle, or Locke's principle.)
In the De Corpore (II, 7, 2), Hobbes supposes that the ship of Theseus gradually has all its planks replaced, and that the discarded planks are subsequently reassembled in the same way they were when they constituted the ship of Theseus. There are now two ships. The first ship is spatio-temporally continuous with the ship of Theseus, but is made of different matter from the matter the ship of Theseus was originally made of. The second ship is made of the same matter (put together the same way) as the ship of Theseus was originally made of Which of these ships, Hobbes asks, is the ship of Theseus?
According to Simons (1987, pp. 198-204), it depends on what we mean by "ship". On his account, there are (what he calls) form-constant ships, which can survive the replacement of their planks, but cannot exist after their planks have been disassembled (even if those planks are subsequently reassembled). And there are matter-constant ships, which can exist after being disassembled and subsequently reassembled, but cannot survive the replacement of (any of) their planks. Before the ship has any of its planks replaced, there are two different artifacts made of the same planks--the form-constant ship of Theseus, and the matter-constant ship of Theseus. Only the first of these artifacts will survive plank-replacement, and only the second will survive disassembly and subsequent reassembly. Moreover, Simons holds, the sortal "ship" is ambiguous between "form-constant ship" and "matter-constant ship". Because it is ambiguous in this way, we are inclined both to say that the ship whose planks were replaced is the ship of Theseus, and to say that the ship whose planks were reassembled is the ship of Theseus. What we should say, according to Simons, is that there were two different ships all along--in different senses of "ship"--one of which survives as a repaired ship, and one of which survives as a reassembled ship.
If Simons has solved the puzzle of the ship of Theseus, then the case described in [sections] I is not a case of two ships in the same place at the same time, and does not constitute a counterexample to the Lockean principle. As long as we stick to one sense of "ship", there will be only one ship at the end of my story--Stathis's, if by "ship" we mean "form-constant ship", or Theseus's, if by "ship" we mean "matter-constant ship". And surely, if we are counting the number of ships in a place at a time, we must stick to one sense of "ship". Doing otherwise would be like saying that there are ten words on a page, on the grounds that there are five word-tokens, and five word-types(2), or saying that there are five banks within a quarter mile of King's College, on the grounds that there are four money-banks and one river-bank.
If Simons is right, after the last of Stathis's ship's original planks have been replaced, there aren't two ships, but there are two artifacts, in the same place at the same time. This, however, is perfectly consistent with the Lockean principle. The two artifacts (a form-constant ship, and a matter-constant-ship) have different persistence conditions, and accordingly belong to different kinds; and the Lockean principle does not exclude things of different kinds being in the same place at the same time.
In fact, though, I do not think Simons' account of the sortal "ship" allows us (satisfactorily) to explain away the apparent counterexample to the Lockean principle set out in the last section. For it is very doubtful that "ship" is ambiguous between "form-constant ship" and "matter-constant ship". Suppose Angelike works in the Nassos shipyard. One night, when she comes home from work, her husband asks her what she did today. She replies that she repaired one ship (replacing a few of its planks), and reassembled another. On the face of it, her remark is perfectly in order. But it would not be, if "ship" were ambiguous between "form-constant ship" and "matter-constant ship". In as much as the ship she repaired would be a ship in one sense, and the ship she reassembled would be a ship in another, she would not be entitled to say that she had refitted one ship and reassembled another. (Suppose Angelike were a real-estate magnate, and had just sold a river-bank and bought a money-bank. She could not on that account say that she had sold one bank, and bought another).
If "ship" is ambiguous between "form-constant ship" and "matter-constant ship", then Stathis's ship and Theseus's ship don't provide a counterexample to Locke's principle. But the converse doesn't obviously hold. "Ship" might be an unambiguous generic predicate, true of both form-constant ships and matter-constant ships. Or it might be what Field (1973) calls a referentially indeterminate predicate, that partially denotes both form-constant ships and matter-constant ships. In neither case would Klops's misdeeds mean trouble for Locke. If "ship" works like a disjunctive predicate, picking out anything that is either a form-constant ship or a matter-constant ship, it will turn out that different ships can occupy the same place at the same time--but again the different ships will be different kinds of ships. If on the other hand "ship" is a referentially indeterminate predicate of the sort discussed by Field, then a statement about ships will be true if it is true for every partial extension of "ship", false if it is false for every partial extension of "ship", and neither true nor false otherwise. (In other words, a statement about ships will be true (false) if and only if no matter how we resolve the indeterminacy about the extension of "ship", it would still come out true (false).) So if "ship" partially or indeterminately refers to both form-constant ships and matter-constant ships, it will be not only untrue, but false, that Stathis ends up with two ships, since he never has more than one form-constant ship, or more than one matter-constant ship.
But does "ship" determinately or indeterminately denote both form-constant ships and matter-constant ships? I think not. The trouble with supposing that "ship" (determinately) denotes both form-constant ships and matter-constant ships is that we end up with too many ships. Stathis will have two ships, not just after he has replaced his ship's original planks, but even before. Indeed, it will be impossible to build just one ship (since one cannot build a matter-constant ship without building a form-constant one, and vice versa).(3) It will also be impossible to repair a ship, without bringing into existence--or at any rate completing the assembly of--another one; and impossible to replace any of a ship's planks, without destroying another ship. All these consequences are highly counterintuitive.
The suggestion that "ship" is referentially indeterminate avoids a ship surplus, but has problems of its own. Whether the predicate "ship" is ambiguous or generic, there will be no problem about our being able to say truly that a certain ship has survived, or will or would survive, plank-replacement, or plank-disassembly-and-reassembly. If, however, "ship" indeterminately denotes both form-constant ships and matter-constant ships, and if Field has given the right semantics for indeterminate predicates, then statements like "This ship could have its planks replaced" will turn out to be neither true nor false. The same goes for "This ship could be disassembled and subsequently reassembled".
Perhaps, though, "ship" is a referentially indeterminate predicate whose reference is made determinate contextually, in accordance with what Lewis (1983, pp. 244-6) calls a rule of accommodation. Then I will be able to say truly that the ship I just built could survive plank-replacement, and able to say truly that the ship I just built could survive the disassembly and subsequent reassembly of its planks. (The context, will, as it were, automatically deliver a precisification that allows what I said to come out true.) Still, if all "ship" can refer to is form-constant ships and matter-constant ships, then no context, however accommodating, can make it come out true that a ship both could survive plank-replacement, and could survive disassembly and reassembly.(4) But it is true! Suppose a man called Francesco built the Santa Maria (one of the ships that went to America with Columbus in 1492). Suppose further that the Santa Maria was refitted and repaired by Francesco on various occasions (having at least some of its original planks replaced) before ending up in a museum in Madrid. Suppose finally that the planks of the Santa Maria have been disassembled, shipped to a Genovese museum, and reassembled there (in the way they were assembled in Madrid) as part of an exhibition commemorating Columbus's voyage to America. Only a philosopher in the grip of a theory would hesitate to say that the ship on display in the Genovese museum is the Santa Maria--the ship that Francesco built and later refitted and repaired, the ship that spent most of the last three centuries in a museum in Madrid, the ship that has survived both plank-replacement and disassembly-cum-reassembly.(5)
There is a certain irony here for Simons's account of the sortal "ship". He notes that that account may encounter resistance from those who feel it multiplies artifacts beyond necessity. The central problem with Simons's account is not, however, what it includes repairable though unreassemblable "ships" and reassemblable though unrepairable "ships"--but what it leaves out--repairable and reassemblable ships. (I use scare quotes here because I am inclined to think form-constant ships and matter-constant ships in virtue of being unrepairable (at least, unrepairable in any way that would involve plank-replacement) or unreassemblable, are not ships proprie loquendo.) If ships both can survive plank-replacement, and can survive plank-reassembly, then there is no reason to suppose that when Stathis's ship and Theseus's ship are made of the same planks, they are ships in different senses, or ships of different kinds, or ships according to different precisifications of the indeterminate sortal "ship". If they are not any of those things, a defender of the Lockean principle is not in a position to accept that (at the end of the story) Stathis's ship and Theseus's ship are two different things in the same place at the same time, and explain away the apparent incompatibility of this judgment with the Lockean principle.
There is a different way in which the Lockean principle might be defended against the alleged counterexample set out in [sections] I.(6) It might be maintained that, as the planks that had constituted the ship of Theseus are put into the ship of Stathis, they are (as Jonathan Lowe put it) exclusively appropriated by the latter ship. Thus at the end of the process of plank-replacement, the planks that had constituted the ship of Theseus constitute (only) the ship of Stathis: the ship of Theseus no longer exists.
I agree that there is, as it were, room in logical space for the view that a ship can survive the disassembly and subsequent ordinary reassembly of its planks, but not disassembly and subsequent reassembly-cum-incorporation of its planks into an already existing ship. It is just that this view does not accord with my actual intuitions about ship identity. Suppose that after the last plank from the ship of Theseus has been put into the ship of Stathis, the disappearance of the ship of Theseus from the museum is discovered by the museum's fabulously rich curator, who offers a massive reward for its return. If Stathis is apprised of Klops's misdeeds, and returns the ship(s) he is sailing on to the museum, I think he would expect, and be entitled to, his reward. I don't think the curator, upon learning the facts of the case, would refuse to pay the reward on the grounds that Stathis had returned the planks of the ship of Theseus to the museum, but not the ship itself. He's got the right planks to constitute the ship of Theseus, put together in the right sort of way--so he's got the ship of Theseus. Of course, the planks were not put together with the intent of reassembling the ship of Theseus, but I don't see why this should matter. And if it did, we could easily enough change the story so that the parts were put together with that intent. Suppose that an antiquarian in Ravenna has hired Klops to smuggle the ship of Theseus out of Nassos for him. Suppose there is no way for Klops to steal the ship of Theseus from the museum (in one piece) without getting caught; the most he can take away on any given occasion is a plank or two. Suppose also that there is nowhere Klops could store the stolen planks without arousing suspicion. So he hits upon the stratagem of gradually replacing the damaged planks of the ship of Stathis by duplicate planks of the ship of Theseus (Klops carries out the repairs himself). When all of the original planks of the ship of Stathis have been replaced by planks from the ship of Theseus, Klops steals the ship(s), sails them to Ravenna, and delivers them to the antiquarian, who rewards Klops handsomely for his ingenuity in getting him the ship of Theseus. If a ship couldn't survive reassembly-cum-incorporation, then in the modified story, Klops and the antiquarian would be operating under a delusion like the man in David Kaplan's story who thinks he has stolen the Nina when all he has in his garage is a replica of the Nina made of different planks.(7) But--it seems to me--Klops and the antiquarian are not deluded in thinking that that Klops has smuggled the ship of Theseus out of Nassos.(8)
Let us suppose for the sake of argument that when a plank is taken from the ship of Theseus and put into another ship, it is exclusively appropriated by that other ship. Even on that assumption, it should still be possible for the ship of Theseus to reappropriate that plank later on. Suppose a magnate with a large fleet of ships wants to smuggle the ship of Theseus out of Nassos and into Chioggia. First he disassembles the ship of Theseus; then he puts one plank from the ship of Theseus into each of his many ships. When one of the ships gets to Chioggia, its ship-of-Theseus plank is removed, and set aside; when all the ship-of-Theseus planks have arrived at Chioggia, they are reassembled. If the operation goes as planned, and the reassembly is carried out correctly, the magnate surely will have succeeded in smuggling the ship of Theseus into Chioggia. This could be so only if the ship of Theseus can reappropriate the planks it (temporarily) lost through incorporation into other ships.
Similarly, suppose for the sake of argument that the ship of Stathis, at the end of our story, has exclusively appropriated all of the planks of the ship of Theseus. It would still be possible to get the ship of Theseus back, by gradually replacing the planks that used to constitute it (and now constitute the ship of Stathis) by new planks, setting aside the old planks, and reassembling the old planks once they have been completely replaced. So-on the assumption that as a result of Klops's misdeeds, the ship of Theseus no longer exists--the curator of our museum, once he had the ship of Stathis in his possession, could get the ship of Theseus back in the way just described. Moreover, if he wanted to have the ship of Theseus--and no mere duplicate of it--back in his museum, it would be sensible of him to do so. But, I want to say, it would be daft of him to go to all that trouble, even supposing that what he wants is the ship of Theseus, rather than a duplicate of it. It would be daft because once Stathis has returned the ship(s) to the museum, the curator doesn't need to do anything else to get the ship of Theseus back.
So, at least, it seems to me. But I could admit that--in view of the vagueness of the sortal ship, and in line with a rule of accommodation--if someone says that the ship of Theseus does not survive the disassembly and subsequent "incorporation" of its planks into a pre-existing and fully functional ship, she says something true. As far as finding a counterexample to Locke's principle is concerned, it is sufficient if there is some (permissible) way of construing--or reconstruing--"ship" on which that sortal is true of things that can survive both (total) plank replacement, and reassembly-cum-incorporation (cf. note 5).
Presuming I am right about this, there remains something to be explained. Call a part a Thesean part just in case, at the beginning of our story, it was part of the ship of Theseus. It is at least arguable that, when the first few planks are removed from the ship of Theseus and put into the ship of Stathis, the ship of Theseus loses Thesean parts. Suppose now, that--as I think--at the end of our story, the ship of Theseus and the ship of Stathis share the same parts. Then, it would seem, when the last few planks are taken from the museum and put into the ship(s) Stathis is sailing, the ship of Theseus does not lose Thesean parts. If anything, it gains Thesean parts. How is it that, while the first plank-transfers definitely do not result in the ship of Theseus gaining Thesean parts, and arguably result in its losing them, the last plank-transfers definitely do not result in the ship of Theseus losing Thesean parts, and arguably result in its gaining them?
The answer is that at some point in the series of plank-transfers the ship of Theseus "moves" from the museum to the sea. I think it works something like this: suppose we start with a new, just-assembled ship S. Then we remove one of its planks, and put that plank into another ship. It is obvious that the parts of S left behind have, as it were, a better claim to (wholly) constitute S than the part removed. (The part removed has no claim at all to (wholly) constitute S, while the parts left behind might be thought to (wholly) constitute S--after, and not before, the other part has been removed, of course.) Moreover--although I think our intuitions on this matter are none too clear--it is at least arguable that the parts left behind have a better claim to (wholly) constitute S than the sum of the parts left behind and the part removed. If another plank is removed from S and put into another ship, once again, the parts left behind will clearly have a better claim to (wholly) constitute S than the parts removed (considered individually or jointly), and the parts left behind will arguably have a better claim to (wholly) constitute S than the sum of the parts left behind and the parts removed. Obviously, though, the process cannot continue indefinitely: at some point, the parts left behind won't suffice to wholly constitute S. What then will become of S? It depends on the fate of the removed (and transferred) parts. They may have gone out of existence. Then we shall say that nothing constitutes S, and that S no longer exists. Or the parts may have been dispersed and (multiply) appropriated, but not destroyed--as in the magnate smuggler example above. Then---presuming that we do not think ships can go on existing in a dispersed and multiply appropriated state--we shall say once again that nothing constitutes S, and that S no longer exists--through it might exist again, if the planks are reappropriated and reassembled. Or the removed parts may have been put back together in the same way they were originally put together in S. Here there are two cases. If the parts have been reassembled without being incorporated into a pre-existing ship (as happens at the end of the magnate smuggler case) those parts will have a serious claim to (wholly) constitute S--perhaps a better claim to constitute S than the sum of the removed and reassembled parts and the parts left behind. Moreover, they will not constitute any ship other than S. Alternatively, the parts may have been reassembled while being incorporated into an already existing (duplicate) ship. In that case, those parts- together with the parts of the duplicate ship that have not yet been replaced by the parts of S left behind--have a serious claim to (wholly) constitute S perhaps a better claim than the sum of the removed and reassembled parts and the parts left behind. If the parts removed, together with the not yet replaced parts of the duplicate ship do constitute S, they will also constitute another ship (the duplicate ship), as happens in the Stathis/Theseus case.(9)
We can now see how it might be true that the first plank-transfers from the museum to the harbor result in the ship of Theseus losing Thesean parts, while the last plank-transfers result in the ship of Theseus gaining Thesean parts. Early on, what gets left behind (at a time) has a serious claim to (wholly) constitute the ship of Theseus (at that time) and what is removed has no claim at all to do that. If what gets left behind early on does wholly constitute the ship of Theseus, then, early on, plank-transfers result in the ship of Theseus losing (Thesean) parts. Late on, what gets left behind (at a time) has no claim at all to (wholly) constitute the ship of Theseus (at that time), and (if I am right) what has been removed and transferred, together with what has yet to be replaced, has a serious claim to (wholly) constitute the ship of Theseus. If the transferred Thesean planks, together with the yet to be replaced parts of the duplicate ship, do (wholly) constitute the ship of Theseus, then any subsequent plank-transfers will result in the ship of Theseus gaining (Thesean) parts.
It follows that, in arguing against the Lockean principle, I need not deny the (not implausible) view that the ship of Stathis initially exclusively appropriates parts of the ship of Theseus. I can say that although the ship of Stathis initially exclusively appropriates those parts, the ship of Theseus eventually gets them (and all its other parts) back--though not by getting them back from (not by taking them away from) the ship of Stathis.
In his discussion of the ship of Theseus puzzle, Noonan (1985; 1989, pp. 152-68) argues that before the first of that ship's original planks are replaced, there is a ship made of those planks that will undergo repair, and a (different) ship made of those planks that will undergo reassembly. The line of thought that leads him to this conclusion is the following:
(1) A ship will survive the replacement of its planks in cases in which the replaced planks are not subsequently reassembled.
(2) A ship will survive the removal of all its planks, and their subsequent reassembly, in cases in which the removed planks are not replaced.
(3) Whether or not the replacement of some planks affords the continued existence of a ship cannot depend on what happens to the replaced planks; whether or not the reassembly of some planks affords the continue existence of a ship cannot depend on whether the planks were replaced as they were removed.
Let h, h', and h" be three possible histories of the world that diverge only after time t--the time at which the ship(s?) of Theseus come(s?) into existence. In h the original planks of the ship(s) of Theseus are replaced, and the replaced planks are subsequently reassembled (in the same way they were assembled before). In h' the same plank-replacement events occur that occur in h, but the replaced planks are not subsequently reassembled. In h" the same plank-removal and plank-reassembly events occur that occur in h, but the removed planks are not replaced. By (1) and (3), it is true at t on h that a ship made of Theseus's ship's original planks will survive as a repaired (plank-replaced) ship. By (2) and (3), it is true at t on h that a ship made of Theseus's ship's original planks will survive as a reassembled ship. Given that the replacing planks and the replaced and reassembled planks will be in different places at the same time, the ship that will be made of the replacing planks and the ship that will be made of the reassembled planks cannot be the same ship. So we may conclude that
(4) It can happen that this ship and that (distinct) ship are made of the same planks at the same time, and are in the same place at the same fume.
From (4) it is a short step though, as we shall see, one that Noonan does not take to
(5) Two ships can be in the same place at the same time and thence, by the considerations adduced in [sections] II, to the denial of the Lockean principle.
So we have a new (mostly Noonan) argument against the Lockean principle. While the argument set out in the first section moves "forward" from ships that are made of different planks at an earlier time, to ships that are made of the same planks at a later time, the new argument moves "backwards" from ships that are made of different planks at different times, to ships that are made of the same planks at an earlier time.
While the new argument needs premisses (1)-(3) to get to the denial of the Lockean principle, the old argument needs at most the first two of those premisses. (The old argument does not need a premise quite so strong as (1), as may be seen by comparing (i) from the first section with (1).) For this reason, I think the opponent of the Lockean principle is on safer ground with the old argument than with the new. After all, someone might accept (1) and (2), and deny (3).(10) She might, for example, reason as follows:
If the ship of Theseus's original planks hadn't been replaced and
reassembled, we would say that the builder of that ship had built
just one ship from those planks. (We don't think that shipbuilders
ordinarily make more than one ship at a time from the same set of
planks.) But how many ships there are at an earlier time cannot
depend on what plank replacements or reassemblies will or will not
happen at a later time. So even if the original planks of the ship of
Theseus are both replaced and reassembled, the builder built just
one ship from those planks--in which case (3) is false.
The point is not that this is a conclusive argument against (3), or that there are conclusive arguments against (3); it's simply that someone arguing against the Lockean principle is advised to get its negation from the weakest possible set of premisses.(11)
I noted earlier that Noonan does not in fact move from (4) to (5), and thence to the rejection of the Lockean principle. Though he does not say a great deal about why he does not, he suggests that the move from (4) to (5) looks obvious only if we are counting ships by identity, and perhaps we may (legitimately) count ships by some relation weaker than identity, with the result that the ship of Theseus that will be repaired and the ship of Theseus that will be reassembled, though distinct, count as one ship (Noonan 1989, pp. 165, 167).
I am suspicious of the legitimacy of counting distinct ships as one ship. Suppose it is true that after the planks that used to constitute Theseus's ship have replaced the planks that constituted Stathis's, there is just one ship made of the planks that used to constitute Theseus's ship. If just one ship is made of those planks, it seems, there should be an answer to the question, "has it [the one ship] ever been in the Nassos nautical museum?" If, however, a ship made of those planks has been in that museum, and a (different) ship made of those planks has never been in that museum, there seems to be no (sensible) answer to that question.
Suppose, however, that this ship and that (distinct) ship may be legitimately counted as one ship while they coincide (that is, while they occupy the same region of space). I don't think this will suffice to save the Lockean principle as usually understood. After all, contemporary defenders of that principle such as Simons (1987, pp. 221-8) and Wiggins (1968, pp. 90-5) take it to rule out the possibility that this K and that (distinct) K are in the same place at the same time. Locke himself writes:
For we never finding, nor conceiving it possible, that two things of
the same kind should exist in the same place at the same time, we
rightly conclude that whatever exists anywhere at any time
excludes all of the same kind. (Locke 1975, II, 27)
Locke says here, not just that there can't be two Ks in the same place at the same time, but also that if one K is here now, then every (other) K is elsewhere or elsewhen.
The considerations advanced in this paper certainly do not constitute a refutation of the Lockean principle. I have argued that the defender of the Lockean principle is committed to some counterintuitive judgments about the number and identity of ships in certain possible situations. No such argument can establish the falsity of Locke's principle, since it is possible that if we abandon the principle, we are committed to equally counterintuitive, or even more counterintuitive judgments. (It is no idle possibility that, whatever we decide about things of the same kind coinciding, we shall end up having to say something counterintuitive about the number and identity of ships in certain possible situations). I hope, though, to have provided a serious challenge to the Lockean principle.(12)
(1) Which ship is that--the one made of the original planks, or the one made of the planks that replaced the original ones? For the purposes of this example, it doesn't matter.
(2) It may be that word-types, unlike word-tokens, aren't in space, and hence aren't on a page. But even someone who thinks that word-types are multiply spatially located universals should not say that there are ten words on a page, if there are only five word-tokens and only five-word types. At least, she shouldn't say that, if, as Simons supposes, "word" is ambiguous between word-tokens and word-types (as opposed to generically denoting both).
(3) At least, it will be impossible unless the ship never changes its matter without losing its form, or vice-versa. If the ship never changes its matter, and never gets disassembled and reassembled, then the form-constant ship and the matter-constant ship will have the same world line. It is controversial whether in such a case the form-constant ship and the matter-constant ship are identical, or are coincident but modally discernible, and hence distinct.
(4) Here we must not confuse "both could F and could G" with "could both F and G".
(5) It is admittedly less clear that the ship Francesco built could survive the replacement of all of its parts, and subsequent disassembly-cum-reassembly, than it is that that ship could survive the replacement of some of its parts, and subsequent diassembly-cum-reassembly. That is because it is less clear that the ship Francesco built could survive the replacement of all its parts, than it is that that ship could survive the replacement of some of its parts. (Just as true, I would say, but less clear.) But if a ship can survive the replacement of all its parts, then it can survive the replacement of all its parts, and subsequent disassembly-cum-reassembly. Suppose I have been spending lots of money getting plank after plank of my ship replaced. You point out that at this rate I shall soon have replaced all the planks, and ask why I don't just buy a new ship. I answer that although buying a new ship would make more sense from a narrowly economic point of view, I am sentimentally attached to my old ship, which was the first one I ever owned. At some later time (after I have replaced all the original planks) I have to move: I accordingly have the boat disassembled and have the bits shipped to my destination. You ask me why I'm going to all that trouble, instead of selling my old ship, and buying a new one in the place I am moving to. I say that I remain sentimentally attached to the ship I have had for all those years. You couldn't very well reply that if I am sentimentally attached to that ship, the last thing I should do is disassemble it, because that ship will be gone forever once its planks are disassembled.
In order to defend the Lockean principle, it is not sufficient to deny that a ship can survive the replacement of all its planks. I could grant that when you say that a ship cannot survive the replacement of all its planks, you say something unfalse, or even something true, in accordance with a rule of accommodation. As long as there are things that can have all their planks replaced, and subsequently undergo disassembly and reassembly--call them ships-considered-as-repairable-and-reassemblable they will make trouble for the Lockean principle, whether or not such things are determinately denoted by the sortal "ship".
(6) Thanks to Jonathan Lowe for pointing this out.
(7) In Kaplan's story, a philosopher has been commissioned to disassemble the Nina, transport the disassembled parts to the Smithsonian institute, and reassemble the parts there. The unscrupulous philosopher carefully replaces each plank he removes from the Nina by a new one before removing the next. Afterwards, he transports the original parts to the Smithsonian, under the delusion that the ship in his garage is the Nina (see Salmon 1982, p. 221).
(8) Does the inclination to think that (in the case just described) the ship of Theseus makes it to Ravenna depend on thinking of that ship as a historically significant object, as opposed to an ordinary workaday ship? Not obviously. Suppose that A lives in Portugal. While vacationing in Oregon, A sees a new wooden ship, and buys it from its builder B, who agrees to deliver it to Portugal. B disassembles the ship into planks, loads the planks onto a truck, and drives the truck to New York. In the New York harbor there is another (duplicate) wooden ship, belonging to B. Rather than reassembling the planks of A's ship in New York, B loads them onto his New York ship, and sets sail for Portugal. The weather is so bad on the voyage that all of the carrier ship's original planks are damaged or destroyed, and need to be replaced by (duplicate) planks from A's ship. There is nothing obviously wrong with saying (I want to say) that (one of) the ship(s) B delivers to A in Portugal is the ship A bought in Oregon, in spite of the fact that the latter ship is not a historically significant object.
(9) If this is account is right, then--as Jonathan Lowe pointed out to me--ships can "move" in funny ways. If, say, the presence or absence of a single plank can make the difference between constituting the ship of Theseus and constituting a part (or ex-part) thereof, then the transfer of a single plank from the museum to the sea can result in the (whole) ship of Theseus' moving from the museum to the sea. This (conditional) claim does not seem counterintuitive to me, though not all my readers may agree. We could avoid accepting the conditional claim by insisting that the ship of Theseus retains ownership (though not exclusive ownership) of all its original planks throughout the series of plank transfers. If we go this route, however, we risk overcounting how many ships there are in certain possible situations: see Hughes (forthcoming).
(10) For instance, Nozick (1981, pp. 29 43) and Shoemaker (1984, pp. 115-8).
(11) The mostly Noonan argument does, however, bring out the interesting fact that there are arguments for there being two ships in the same place at the same time that do not explicitly or implicity presuppose a closest continuer theory of identity. Even an opponent of the Lockean principle who accepts the closest continuer theory of identity should be pleased by this.
(12) Thanks to Marcia Mayeda and Jo Wolff for encouragement, and to Jonathan Lowe, an anonymous referee, and the Editor, for many helpful suggestions.
Field, Hartry 1973: "Theory Change and the Indeterminacy of Reference". Journal of Philosophy, 70, pp. 462-81.
Hobbes, Thomas 1994: The Elements of Law, Natural and Politic. Part I: Human Nature; Part II: De Corpore Politico. ed. J.C.A. Gaskin. Oxford: Oxford University Press.
Hughes, Christopher (forthcoming): "Aquinas on Continuity and Identity". Mediaeval Philosophy and Theology.
Lewis, David 1983: "Scorekeeping in a Language Game", in his Philosophical Papers, Volume 1. Oxford: Oxford University Press, pp. 233-50.
Locke, John 1975: An Essay Concerning Human Understanding, ed. P. Niddich. Oxford: Clarendon Press.
Noonan, Harold 1985: "The Closest Continuer Theory of Identity". Inquiry, 28, pp. 195-229
-- 1989: Personal Identity. London: Routledge and Kegan Paul.
Nozick, Robert 1981: Philosophical Explanations. Oxford: Clarendon Press.
Salmon, Nathan 1982: Reference and Essence. Oxford: Blackwell.
Shoemaker, Sydney and Swinburne, Richard 1984: Personal Identity. Oxford: Blackwell.
Simons, Peter 1987: Parts: A Study in Ontology. Oxford: Oxford University Press.
Wiggins, David 1968: "On Being in the Same Place at the Same Time". Philosophical Review, 77, pp. 90-5.
|Printer friendly Cite/link Email Feedback|
|Date:||Jan 1, 1997|
|Previous Article:||A consistent relativism.|
|Next Article:||Weak deflationism.|