Artifacts of Theseus: Fact and Fission
Carter (William)
Source: Australasian Journal of Philosophy, 61.3 (September 1983)
Paper - Abstract

Paper StatisticsNotes Citing this PaperLink to Latest Write-Up Note


For my Write-up, which is fairly extensive, though incomplete: Click here for Note, or read on …



Write-up1 (as at 19/04/2018 00:12:58): Carter – Artifacts of Theseus

Note that this paper ("Carter (William) - Artifacts of Theseus: Fact and Fission") is essentially an argument against the Constitution View.
  1. Introduction
    • Standard “Ship of Theseus” case, mutated to bicycles.
    • Definitions:
      • The “continuity survivor (CS)” is the one with spatiotemporal continuity (part replacement).
      • The “reassembly survivor (RS)” is the one made of the replaced parts.
      A footnote points out (Gupta) that there may be issues with replacing the frame. This is probably referring to a point that comes up later as another aside – do things have essential parts?
    • Question: which of the surviving bicycles is the original? There are 5 options:-
      1. a-Theorist: CS is, RS isn’t. Supporters: Wiggins, Salmon.
      2. b-Theorist: RS is, CS isn’t
      3. c: both are
      4. d: neither is
      5. e: no fact of the matter. Supporters – Chisholm, Nozick, Rorty
      The paper focuses on a- and b-theorists, though it spends a lot of time considering Chisholm’s arguments.
  2. Section 1
    • The a-Theorist supports spatio-temporal continuity.
    • The b-Theorist has a modal counter-argument.
    • Definitions:-
      • W1: reassembly without replacement; no continuity competitor. RS only.
      • W2: reassembly with replacement. RS + CS.
      • t1=time that the reassembly takes place.
      Argument: RS(W1) = RS(W2); B(W1) = RS(W1); Transitivity of identity: RS(W2) = B(W1). So, the RS is the original and the CS isn’t.
    • Key premise: RS(W1) = RS(W2). Rejected by a-Theorists. Distinguish original construction from reconstruction. Reconstructed objects have a different origin and are non-identical to the pre-deconstructed object.
  3. Section 2
    • Salmon and Hunks of Matter. Distinguish the “is” of identity from the “is” of constitution. The same Hunk constitutes different bicycles in different worlds.
    • Two questions:-
      1. Does B(W1) exist throughout the disassembly/reassembly process?
      2. Does H(W1) - the Hunk of matter - exist throughout the same process?
      Df: td is the time at which disassembly is complete. This raises the question whether disassembly is instantaneous or piecemeal. It seems that as replacement occurs in W2, it cannot be instantaneous. The TE at the start of Section 3 confirms that disassembly is gradual. But Section 1 seems to imply that reassembly is instantaneous “at t1”.
    • Four obvious alternative answers:-
      1. a.1 Both B and H do
      2. a.2 B doesn’t, H does
      3. a.3 B does, H doesn’t
      4. a.4 Neither B nor H do.
      I don’t think the choice of “a” as a prefix has any connection with a-theorists. As such, it is an unfortunate choice.
    • Note: that “Hunk” is a piece not a portion, so is contiguous and can lose bits. The standard constitution account has it that things are constituted by pieces rather than portions (which can exist in scattered states, and cannot lose bits). "Ayers (Michael R.) - Locke on 'Masses of Matter'" raises a problem against Locke’s “Masses of Matter” – they need boundary conditions, but Locke’s masses are portions not pieces, so Ayers is probably confused.
    • Carter suggests initially hat we must reject all but a.4, as neither a B nor an H can exist dispersed.
    • However, Wiggins (an a-Theorist) claims that a thing starts to exist only once, and that a clock does survive the disassembly/reassembly process. And, there are no intermittent objects.
    • Chappel distinguishes between Parcels (Portions) and Pieces (Lumps, = Hunks). Reject the rejection of a.1 and a.2: this was originally rejected on the grounds that Hunks can’t exist dispersed, but considered as Portions they persist. However, this just confuses Pieces with Portions (if we do this, why not go the whole way and confuse Bicycles with Hunks)?
    • Note: that this distinction is already present in Gibbard. It is an important distinction, though not universally accepted. He features in the references to the paper, so why is he not mentioned here?
  4. Section 3
    • Df: a-theorists who agree with Chappell are Constitutionists.
    • One question: can there be atemporal constitution; ie. If a Portion of matter at any time constitutes a bicycle, does it always do so? Carter replies “no” on the grounds that in W2 shortly before td, we have a pukka bicycle with mostly replaced bits, and a heap of bits, which don’t quite make a bicycle because there are still a few bits of our Portion embedded in our existing bicycle. Our Portion consists in the heap + the embedded bits, and this collection doesn’t presently constitute a bicycle.
    • Note: this is even more obvious in the case of human beings, the matter changing every 7 years, and being incorporated into the bodies of many other human beings, or into other things.
    • In 4D, we might ask whether all things exist atemporally – ie. Do the future stages exist? I need to review Presentism and the arguments against it. Does Presentism just assume that future things don’t exist, or argue against their existence as it does against that of past things? Can there be 4D versions of Presentism? Ie. Do all stages exist, or just the present stage?
    • A second question is a refinement of the above. Does a portion which, when it clearly constitutes a bicycle always constitutes the same one, always constitute a bicycle?
    • Two TEs:-
      1. Bicycle completely disassembled, and its bits left knocking around, and then completely reassembled.
      2. Bicycle completely disassembled, but its bits are temporarily incorporated as parts of 100 bicycles before being ultimately removed and reassembled into a single bicycle.
      What to make of these TEs? Answers:
      1. A: it is somewhat plausible that the bicycle continues to exist in a dispersed state, but Carter doubts it.
      2. A: same but stronger doubts – there are 100 bicycles, not 101.
  5. Section 4
    • Carter doubts Constitutionists can distinguish 1A from 2A and affirm 1A but deny 2A.
    • But is Constitutionism a deep problem? Is it mostly accepted? Carter is worried.
    • Might the “is” of Constitution be the “is” of identity after all. Richard Cartwright and Chisholm think so. If so, the Constitution answer to the Ship of Theseus paradox is incorrect.
    • Locke’s Parcels – mereological essentialism for Parcels. So, a Constitutionist must say that the W2 Continuity bicycle (CS) is constituted by different Parcels at different times.
    • Note: they might well, but don’t most Constitutionists claim that things are constituted by Pieces (which may themselves be constituted by Parcels)? Lumps can lose or gain parts.
    • Chisholm denies that anything persists – there’s just a succession of related objects.
    • Note: I think this is due to confusing objects with their stages. 4D may be the only answer to the denial of “strict and philosophical identity” applying to ordinary objects. Is / was this Van Inwagen’s view – in saying that only simples and people (souls?) exist. Or was it Unger? Or both? And have they changed their minds?
    • Carter admits to skating over Chisholm, and only mentions him to cast doubt on Constitutionism. Each chunk of matter is attributively a bicycle (ie. Is bicycle-shaped, bicycle-propertied). Carter accepts this, and thinks it “does for” Constitutionism.
    • Note: we need to watch out for the attributive “is”, which may not be the “is” of identity any more than the “is” of Constitution.
    • Explanation: TE – in W2, between td and t1. td is the time of complete disassembly, t1 the time of complete reassembly. So td – t1 is the period of reassembly. Carter claims we always have a bicycle before us throughout this time. Which? We have CS, which is already fully replaced at td (is it?), and is a full bicycle throughout the period, and we have RS, which is being built in the period. Maybe I’ve misunderstood. On p. 249, Carter says “at t1 the parts that were replaced are reassembled”. So, it may be instantaneous re-assembly, but we’re not looking at RS in this TE but at CS. Good. So, Carter considers the stages of CS from td to t1. For each stage there is a (different) co-located Parcel of matter, labelled pi, and he considers p90, which is attributively a bicycle.
    • We now get the “too many bicycles” objection (TMB). There’s only one bicycle present, and it’s p90, so in saying “p90is a bicycle” we have the “is” of identity.
    • Note: this just begs the question. Why can’t we have co-located objects of different sorts, with different persistence conditions: the bicycle and the Parcel of matter that instantaneously constitutes it, and adjust our counting rules?
    • Carter thinks the b-theorist (who argues that RS is the original) will be enthused by this result. Presumably because Chisholm is arguing against the existence of ordinary objects – only Parcels of matter really exist, and the b-theorist tracks the Parcel (though I suspect he tracks pieces … he’s not interested in all the odd atoms that come and go).
    • We now return to W1 (RS only, no CS). So, prior to td, a W1-inhabitant points to a Parcel, which persists throughout td to t1, so a.1 is true, since B=H.
  6. Section 5
    • Df: A Parcel is Cyclesque only if its parts are configured as and function as a bicycle. The Constitutionist must deny that a Cyclesque Parcel is a bicycle, or fall foul of the TMB argument (Note: unless he changes his counting rules).
    • Argument from Supervenience2: Carter thinks that there’s an argument from Supervenience3 to the effect that a Cyclesque Parcel is (identical to) a bicycle. Carter cites Kim, though I suppose that only the terminology, and not the argument, is Kim’s. I think the argument is due to Chisholm in the dialogue with Shoemaker in Care/Grim. This argument is important, and the aim is to show that if the thing constituted is not identical to the thing constituting, then we have a violation of generally-accepted supervenience4 claims. I need to consider it carefully some time, probably after reading the Chisholm/Shoemaker piece. Enough to say that it’s really a more technical version of the TMM/TMB argument.
    • Terms: Closure, Microdeterminate & Microreducible.
    • Note: I’m not sure what the closure of a family of properties is, nor why or whether it’s important.
    • It’s possible to quibble that the argument has the wrong constituting entities – Parcels / Portions rather than Pieces / Hunks; but, I expect an analogous argument from supervenience5 can be derived for Hunks. Does Olson (or Blatti) refer to these papers?
    • Shoemaker rejects the notion that the constituting items are themselves ships (but why?). We have a modus ponens versus modus tollens situation. If we don’t like the thought of co-located entities, we reject the premise (constitution) that implies it; but if we’re impressed by the importance of different persistence conditions (and the rejection of deviant identities) then we allow co-location of entities of different sorts. The crux of the whole matter comes when we consider whether we can allow (virtually) co-located entities of the same sort. Carter’s review of Hudson’s book ("Jones (Nicholas K.) - Too Many Cats: The Problem of the Many and the Metaphysics of Vagueness") addresses this “Problem of the Many”. Maybe we should (with Lewis) change / acknowledge our counting-rules. I still think there’s a big reductio available for the TMM argument in the haze of mostly overlapping atom-complements (more cogent than Tib/Tibbles, because while a tail-complement may not be a cat – though a tailless cat is a cat – atom-complements of cats certainly are cats).
    • There’s an interesting footnote on a problem for the Supervenience6 argument, in that for an object to be what it is, it has to have an appropriate history (is this the same argument as Putnam’s in "Putnam (Hilary) - Brains in a Vat"?). However, Carter claims that Shoemaker agrees with him that objects don’t need to be of a certain minimum “temporal size” in order to qualify as a thing (rather than a stage of a thing?). Is this really Shoemaker’s line, and what does he intend by it?
  7. Section 6
  8. Section 7
  9. Section 8
  10. Section 9
  11. References



In-Page Footnotes

Footnote 1:

Text Colour Conventions (see disclaimer)

  1. Blue: Text by me; © Theo Todman, 2018



© Theo Todman, June 2007 - Sept 2018. Please address any comments on this page to theo@theotodman.com. File output:
Website Maintenance Dashboard
Return to Top of this Page Return to Theo Todman's Philosophy Page Return to Theo Todman's Home Page