Source
Le Bihan & Read (2018), LIT-113 (Philosophy Compass 13(12)), read in full in NOTE-152 from the accepted manuscript. Page numbers are the manuscript’s printed numbers.
What was actually shown
The setting. A duality is a map between the solution spaces of two theories under which related models agree on their empirical substructures. The map may be many-one (fn. 4, p. 3). Each model is read naively: the world it represents is isomorphic to it (fn. 8, p. 5). Antirealism and structuralism, including OSR, are set aside (§4).
The taxonomy (§3, Fig. 1, pp. 7–9). Which worlds count as candidates for actuality gives six cases: all duals with or without the common-core world (underdetermination, cases 1 and 2), one dual with or without it (discrimination, 3 and 4), and the common core alone or nothing (nihilism, 5 and 6). The common-core approach is cases 1, 3 and 5, so it overlaps with discrimination in case 3.
The common core does not help on its own (§5). Moving from case 2 to case 1 by adding the common-core world “has made the situation worse” (p. 12): it adds one more empirically adequate candidate. Super-empirical virtues and Occam’s razor are “first and foremost” practical and remove no candidate. The identity of indiscernibles would take case 1 to case 5, “but why should we commit ourselves to such a metaphysical principle?” (p. 13).
Nor does an overarching theory (§8.1, pp. 17–19). If the duals’ solutions simply are solutions of a deeper theory, it adds no distinct ontology and the underdetermination remains. If it does add one, a further principle is still needed, and the identity of indiscernibles need not supply it, since the deeper theory need not have less structure.
Discrimination can be motivated case by case (§6, pp. 13–15). For AdS/CFT, preferring the CFT side is defensible because only it is fully constructed. Where one dual carries gauge redundancy, the identity of indiscernibles applied within that theory supports the other.
What could have come out otherwise. One of the familiar routes could have selected a world without an extra premise. On the paper’s analysis none does.
What this does not say
- That dual theories are physically equivalent, or that they are not. The verdict is conditional: given naive internal interpretation, a single world needs an extra principle.
- That it holds for the stronger notion of duality. De Haro & Butterfield (LIT-199) define duality by isomorphism of a bare theory. Their account of internal interpretation argues a route to case 5. This paper leaves that route open and does not engage it (NOTE-152, Connections).
- That structural realism dissolves the problem. Structuralism is set aside, not weighed (§4, fn. 18). The common-core approach is distinguished from both versions of structuralism (fn. 19).
- That pluralism is a live option. On pluralism, all dual structures are co-instantiated in one world. It is judged “prima facie consistent”, with an overdetermination problem unanswered and an “ontological problem” named but not discussed (pp. 19–21).
- That discrimination for AdS/CFT survives a full construction of the bulk side. The argument depends on the bulk being known only perturbatively.
Connections
- Belot’s treatment of gauge reduction as removing physically otiose variables (LIT-122; THEORY-441) is the intra-theoretic move this paper borrows in §6 and contrasts with the inter-theoretic use in §5.
- De Haro & Butterfield (LIT-199) give the common-core construction the paper assumes (fn. 13).
- THEORY-690 files their verdict for the stronger notion: in the cases they examine, inequivalent duals are not empirically equivalent, and a cautious realist commits to the common core. That is not a rival to this account, which is conditional on empirical equivalence and naive reading.