Contribution
Belot makes the ontological status of the quantum state a question to be answered inside one primitive-ontology theory, Bohmian mechanics with particles in 3-space. There the macro-object problem is under control, and the dualism of material degrees of freedom plus ψ is explicit. He sets out and prices four options, introducing the dispositionalist one and its particle-in-a-box counterexample.
Key insight
Primitive ontology solves the macro-object problem by attributing properties to regions of spacetime, but it does not remove the quantum state. It is dualistic, because the material degrees of freedom are “not dynamically autonomous” (p. 3). So the primitive ontologist still owes an account of ψ, and every familiar category — object, law, property — fits ψ only with a specific cost.
Assumptions
- Non-relativistic Bohmian mechanics for n spinless particles in Euclidean 3-space. Wave functions are identified up to measure-zero differences and complex multiples (p. 3, fn. 8). The guidance law is eq. (1), and p(q) = |ψ|²/‖ψ‖² is read as typicality (p. 4).
- The primitive-ontology reading of the particles, as n worldlines in spacetime, not Albert’s single world-particle in configuration space (p. 4).
- The maxim that an acceptable account must model the experiments that support it (p. 1): the measurement problem plus the macro-object problem.
Key results
- ψ as field (pp. 5–7). ψ is unlike a field even for one particle. (i) It does no conservation-of-energy or momentum work. (ii) There is no back-reaction: the Schrödinger dynamics ignores which particle history is actual. (iii) It is not a boost-invariant scalar: ψ(x,t) ≠ ψ(x′,t′), differing by a phase that depends on x, t, v and m (fn. 19). A field on configuration space buys only a locality in configuration space, which is not the locality anyone wanted. It also raises the question of why the world appears 3-dimensional, with little motivation for taking on the burden.
- ψ as law (pp. 7–9). The model is the classical Hamiltonian, which no one expects to be acted on by the motion (Dürr, Goldstein and Zanghì 1997; Goldstein and Teufel 2001). The reading implies (1) time-indexed law-propositions and (2) grades of necessity: ψ is “more like an initial condition” than the Schrödinger equation. The quantum-gravity programme (time-independent, perhaps unique ψ, Hartle–Hawking) would remove both, but “in the absence of such a theory” one is stuck with them. (1) is the serious obstacle and (2) is tolerable. A residual worry: there is little imaginative resistance to supposing ψ different (p. 9).
- ψ as property, or dispositions (pp. 10–13). The guidance law maps ψ(t) to a vector field Xt on Q, a holistic velocity disposition of the system. Bare or holistic dispositions are not the real worry. The real worry is content. For a free particle in [0,1], every φn(x) = sin(πnx) gives zero velocity everywhere, so infinitely many inequivalent energy eigenstates share one dispositional history. The distinctions in energy and in typicality (φ1 versus φ2) are effaced (“dismaying”, p. 12). Belot’s rescue: conjecture that the map is injective generically (fn. 41), and bite the bullet elsewhere — no determinate fact about energy or typicality. The problem arises only for closed systems (fn. 42).
- Conclusion (§3, pp. 13–14). Each option relies on contested metaphysics. On the orthodox quasi-Quinean view the multi-field reading is tentatively most plausible. On van Fraassen’s “heresy”, which Belot favours, the aim is understanding, and the other options are worth pursuing.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Existing primitive-ontology approaches are dualistic: the material degrees of freedom are not dynamically autonomous | strong | the form of the guidance law (eq. 1), plus the literature cited in fn. 5–6 |
| C2 | ψ differs from classical fields in conservation role, back-reaction and boost behaviour | strong | standard physics; (iii) from the Galilean transformation of ψ (fn. 19) |
| C3 | ψ-as-law in current theory entails time-indexed laws and grades of necessity | moderate | informal argument (p. 8) |
| C4 | Time-indexing, not grades of necessity, is the serious cost of ψ-as-law | weak | informal argument from accounts of laws (Lewis vs Lange) and statistical mechanics (p. 9) |
| C5 | On the dispositionalist reading, inequivalent ψ can share one dispositional history | strong | explicit counterexample, particle in a box (pp. 11–12) |
| C6 | The map ψ(t) ↦ Xt is injective for generic solutions | weak | conjecture, Goldstein p.c. (fn. 41); not proved |
| C7 | The orthodox should prefer the multi-field reading | weak | informal judgement from minimal revision (p. 14) |
Concepts
- Primitive ontology approach: fully describing a history attributes interesting properties to regions of spacetime (p. 2; Goldstein 1998).
- Macro-object problem: supplying truth conditions for ordinary properties of macro-objects (p. 2).
- Multi-field: a configuration assigns properties to sets or fusions of n points of 3-space (p. 5; Forrest 1988).
- Dispositionalist interpretation: the complete history is q(t) together with Φt, the system’s velocity dispositions, which is Xt on Q (pp. 10–11).
- Orthodoxy versus heresy: interpretation as a search for the most plausible truth, versus the delineation of interesting alternatives for understanding (p. 14).
Connections
Belot is Chen’s (LIT-204) source for the multi-field reading, and Chen’s strong nomological section answers Belot’s time-indexing worry with the same quantum-gravity programme. Allori (LIT-200) cites these objections (fn. 22) and answers the time-dependence cost by demoting ψ to an ingredient of the law. Poltorak (LIT-164 §5.2) cites Belot as the map of the primitive ontologist’s options. Dorato’s powers-of-events reading (LIT-176) is the dispositional relative. Belot’s particle-in-a-box argument bears on any reading that identifies ψ with the velocity field it induces. Gao (LIT-121) is the ψ-ontic alternative outside primitive ontology.
Metaphysics. The work presupposes a dualistic primitive ontology and weighs theses about ontological categories (object, law, property), the nature of laws (time-indexing, grades of necessity, Lewis versus Lange), bare and holistic dispositions, and the dimensionality of fundamental space. It ends with a metaphilosophical claim about the aim of interpretation. The metaphysics tag is justified. The primary tag should stay quantum-foundations.
Bearing on the record
It is a load-bearing node for the ψ-as-law debate. It supports no THEORY document. Nothing for ML practice.
Limitations
- It is restricted to one theory, Bohmian mechanics; GRW primitive ontologies are not worked through.
- It is non-relativistic.
- The dispositionalist rescue rests on an unproved genericity conjecture.
- The closing verdict is explicitly “inconclusive”.
- The configuration-space field reading is dismissed for lack of motivation rather than refuted.
Open questions
- Is ψ(t) ↦ Xt generically injective (fn. 41)? A proof or counterexample would settle the dispositionalist reading.
- Can the dispositional state be made deterministic by adding time-derivatives of Φ (fn. 39)?
- Does a fundamental quantum-gravity theory deliver a time-independent, unique ψ, which would remove ψ-as-law’s main cost?
Corrections to the seeded skim
- On ψ-as-law, the dossier lists time-dependence and grades of necessity as parallel costs. Belot ranks them. Time-indexed laws are “the more daunting obstacle” and undermine the analogy with the time-independent classical H (pp. 8–9). Grades of necessity “should not … constitute the same sort of sticking point”, since Lange-style accounts supply them. The Past Hypothesis comparison is offered in support of grades of necessity (p. 9), not as a further cost.
- The dispositionalist reading is Belot’s own. Fn. 7 says every option canvassed is in the literature “except the dispositionalist account”, although the property idea is credited to Monton (2006). It is a history Φt of velocity dispositions of the whole particle system, represented as a time-dependent vector field Xt on Q. It is not Esfeld’s view, which the dossier brings in.
- The dispositionalist reading also loses determinism if (q, Φ) is taken as the instantaneous state (fn. 39). Its rescue rests on a conjecture, credited to Goldstein by private communication (fn. 41): that the map ψ(t) ↦ Xt is injective for generic solutions. Belot does not prove it.
- Belot adds that the multi-field reading is “the safest” but “the least interesting” (p. 13), and that the field objections (i)–(iii) apply “even in the one-particle case” (p. 6).