DOES A THING EXIST WHILE IT'S BROKEN?⁽¹⁾

Take it for granted that I hold to Wiggins' D_II⁽²⁾: for every continuant, there is a sortal under which that continuant falls for the whole of its existence. Now imagine that we have given the name 'J' to a clay jug.

What do we say about the following (construe, throughout, 'broken' as 'disassembled enough to count, bar reassembly, as destruction')?

Case A: At time T₁ J is broken, the hundred or so parts that result are saved, and, at T₂, they're all glued back together.

I assume that we say J exists after T₂: the pieces have been put back in just the right way (I will simply ignore, here, examples where there are many ways in which the pieces might be reassembled).

Maintaining consistency with D_II, we might say (i) or (ii):

(i) J existed from T₁ to T₂. From T₁ to T₂ J was a jug, albeit a broken one.

(ii) From T₁ to T₂ no jug existed (in the place under consideration), and, perforce, J didn't. Only the pieces that used to constitute J existed continuously from T₁ to T₂.

It may seem that we might just take our pick, innocuously. 'jug' and 'J' go together in each alternative, and that's all that matters for D_II. But it's not that easy: each line has a consequence thought by some to be counter-intuitive.

For (i), reach the bad consequence by noting this: had J never been reassembled, the end of its existence would have been at T₁. Thus J's existence, in (i) as described, is what used to be called a "soft fact": criterially dependent on what will happen later.

(ii) seems to conflict with the intuition, often expressed, that a thing commences existence at most once.

So far, exposition. Let me approach a conjecture by way of two further cases and an observation.

Case B: J undergoes a ship of Theseus experience from T₁ to T₂: after much chipping and repairing, it comes to be made of entirely new clay.

Case C: J is broken, as in A; the pieces are put in a drawer; then, as in B, one by one the pieces are thrown away and replaced by new ones. At T₂ (none of the original pieces remaining) the new pieces are assembled into a jug.

My intuitions say that in B J exists after T₂, and that in C it does not.

The observation is this: anyone trying to make out the logical possibility of time travel has to introduce a distinction between "objective" time and what I will call "diary" time - the time series of the thing travelling in time (for example, an event in 1776 might be diary later than some event in 1986, for a time-travelling American patriot).

Now suppose that we adopt the idea of a diary time series for every continuant, regardless of any interest in time travel. Then look back at (i) and (ii) in our discussion of case A.

Under (i), nothing of interest would be said. J's diary time would coincide with objective time.

Under (ii), there would be a difference: in objective time J was out of existence for a period of time, but, in J's diary time, not: in that series there is no temporal gap between T₁ and T₂. (In objective time J's existence is discontinuous, in its own diary time, continuous).

And - to get back to the point - (ii) wins hands down in the game of answering the question: 'Why don't our intuitions treat case C just as a combination of A and B?'. If J were to be counted as continuing to exist in C, the events in its diary time would have to include an instantaneous total replacement of parts. But such instantaneous replacement paradigmatically counts against continued existence. Option (ii), combined with the idea of diary time, permits us to picture C very differently from the way in which we picture A and B, in just the direction that would suit our untutored judgements.

I conjecture, then, that we will have a more coherent account of continuant identity if we add to our descriptive metaphysics this principle: things don't exist when they're broken. There are no "soft facts" to worry about (that a currently nonexistent thing will come back into existence is, if a fact, surely a future fact). We may even save, by relativizing to diary time, the claim that a thing can only come into existence once. Rescue by trivializing, of course - just what we should do with modal intuitions.⁽³⁾

1. Doris Olin midwifed the central point here. Ann Wilbur MacKenzie collaborated with me on an essay about time travel, whence (her coinage) 'diary time'. Written in mid-80s.

2. Wiggins, David, Sameness and Substance. Cambridge, Mass. : Harvard University Press, 1980.

3. I still think this view has a certain amount of charm. But it requires one to swallow some funny consequences. One's parts, for example, may be older than oneself, even though they began life at the same time as one did oneself but there's no inconsistency, if the various diary times are kept straight. Still, it might be better to deal with the cases most notably, (C) in terms of a hierarchy of criteria, so that when form and function are not available, continuity of matter must carry the identity burden. For the idea of a hierarchy of criteria, and reference to Wiggins, whence the idea comes, see my "The Schematics of Continuant Identity", Dialogue XXV, Summer 1986.