The Elusive Persistence: Stage Theory, Lewis's Worms, and the Baroque Labyrinth of Modal Realism — Epoche C1
Simón Bolívar rode into Caracas in August 1813 at the head of the Admirable Campaign and was acclaimed El Libertador ; in March 1830 he resigned the presidency of Gran Colombia, and in December of that year he died at Santa Marta, having watched the federation break apart. The metaphysical question is what makes those two men one man. It is an old question — Plato already reports Heraclitus as saying one cannot step twice into the same river ( Cratylus 402a) — and the twentieth century's most systematic answer belongs to David Lewis, whose On the Plurality of Worlds (1986) defends modal realism: the doctrine that there exist countless concrete universes, spatiotemporally isolated from ours and from each other, in which every way things could have been is a way some world actually is. The essay this one expands claimed that Lewis's "stage theory" solves the problem of persistence and then argued that it fails. Both halves need repair, and the first repair is a correction of fact. Lewis is not a stage theorist. He is a worm theorist, and the difference between those two positions is the pivot on which everything here turns. The stage theory is the work of Theodore Sider and Katherine Hawley, both writing in 2001. Setting out the two views properly — and giving Lewis the argument he actually gives — turns out to leave a real objection standing, but a narrower and more interesting one than the original charge of incoherence. The problem the machinery was built to solve Start with the puzzle that forces the whole apparatus, because without it the machinery looks like extravagance for its own sake. Lewis calls it the problem of temporary intrinsics, and states it in On the Plurality of Worlds (§4.2). An intrinsic property is one a thing has purely in virtue of how it itself is, independently of anything else: shape is the standard example, whereas being an uncle is extrinsic , since it depends on a sibling's having a child. Now: Bolívar is upright in 1813 and recumbent in 1830. Both are intrinsic. If one and the same wholly present thing has both, we have contradicted ourselves. Lewis canvasses exactly three escapes. Make the properties relational. Say that shape is really a two-place relation between a thing and a time: Bolívar bears upright-at to 1813 and recumbent-at to 1830, and there is no contradiction because the relata differ. Lewis rejects this, and his reason is not fastidiousness: the proposal denies that shape is intrinsic at all. If being bent is a relation to a time, then a thing's shape is not a matter of how that thing is in itself, which Lewis takes to be plainly false. Presentism. Say that only the present exists, so that only present properties are had at all, and the 1813 properties are simply not there to conflict with anything. Lewis rejects this because he holds past and future things to be as real as present ones; the escape costs the reality of everything that is not now. Perdurance. Say that Bolívar is spread out in time as well as in space, and has different temporal parts at different times — the way a road has different spatial parts in different counties. The 1813 part is upright, full stop; the 1830 part is recumbent, full stop. No single thing bears both properties, so nothing is contradicted, and shape stays intrinsic. Lewis takes the third, and it is he who fixed the vocabulary that the field still uses. Something persists if it exists at more than one time; it endures if it persists by being wholly present at each of those times; it perdures if it persists by having a different temporal part at each. The original version of this essay described Bolívar's stages as parts of "the same enduring person". On Lewis's own definitions that is a contradiction in terms: endurance is precisely the view he is rejecting. The word wanted is "perduring", and the slip matters, because the intuition of a robustly unified individual that the essay went on to defend is the endurantist intuition, and it needs to be argued for rather than smuggled in through an adjective. Worms and stages: the correction Everyone in this debate accepts temporal parts. The disagreement is about what the name "Bolívar" picks out. On the worm theory — Lewis's — Bolívar is the whole four-dimensional aggregate, the "space-time worm" extending from 1783 to 1830. He is not wholly present in 1813; only a temporal part of him is. When we say Bolívar was a revolutionary in 1813, we say that the worm has a temporal part located in 1813 which is a revolutionary. Sameness over time is straightforward numerical identity: there is one worm, and the 1813 and 1830 stages are two parts of it, in the same undramatic sense in which the bow and the stern are parts of one ship. On the stage theory — Sider's Four-Dimensionalism (2001) and Hawley's How Things Persist (2001) — "Bolívar" denotes a momentary stage. The persisting object is not what we refer to; it is at most the series along which our references travel. A past-tense claim is then made true not by identity but by a temporal counterpart relation: "Bolívar was a revolutionary" is true, said of the 1830 stage, because some earlier stage suitably related to it was a revolutionary. Identity across time drops out of the semantics entirely. So the compressed essay's central attribution was doubly astray. It ascribed the stage theory to Lewis, and then, in describing it, correctly said that on Lewis's view the person is "a trans-temporal mereological sum of stages" — which is the worm theory, and the denial of what had just been ascribed to him. Where Lewis does work with stages is in "Survival and Identity" (1976), his response to Derek Parfit, and there he is explicit: a person is a maximal aggregate of person-stages, so the person is the sum and not the stage. Why temporal counterparts are not modal counterparts, for Lewis The word "counterpart" appears in both halves of Lewis's system, and the original essay's argument depended on treating the two occurrences a