Source
Cuffaro & Hartmann (2021), LIT-037, read in full in NOTE-077.
What was actually shown
The two frameworks. ST: normalised state vectors, Hermitian Hamiltonians and unitary evolution, with open-system dynamics obtained by tracing out an environment. GT: density operators under general linear maps, in general non-unitary and not necessarily completely positive (§2.2.2, after Sudarshan–Mathews–Rau and Jordan–Sudarshan). Each instantiates a “view”: the closed view models an open system by coupling it into an isolated whole; the open view writes the environment into the system’s own law.
The tie (pp. 29, 35–36). On the descriptions both frameworks make sense of, ST’s dynamics re-describe losslessly in GT via ρ = |ψ⟩⟨ψ|, and GT’s completely positive dynamics derive from ST by Stinespring dilation. “Neither dependence nor determination can help” (OntFund-1). Under generalized Nagel–Schaffner reduction each justifies the other (EpFund-1). The paper sets aside the greater expressiveness of GT (non-CP maps) as a premise, because using it would be circular (pp. 29, 33, 35).
The break. OntFund-2 asks whether a framework’s objects actually represent their targets as its motivating metaphysics says. On orthodox readings a state vector encodes an open system’s possible manifestations in interaction with a further system; on Everettian readings reduced density operators are as real as the universal state, and a non-unitary universal density operator is consistent with Everett (§3; pp. 30–33). So ST’s objects match the open view, and GT is ontically more fundamental. EpFund-2 orders concepts by which is constructed from which in practice: a pure state is an idealisation of ρ = (1−ε)|ψ⟩⟨ψ| + εσ (Eq. 4.2), and a global state is reached only through subsystems (pp. 36–38).
Explanation. ExFund-3 on its face favours ST, whose explanations are simpler; it is rejected by one reductio (Newton against GR). ExFund-1 is deflationary, and the authors suggest discarding the notion (p. 41).
What could have come out otherwise. OntFund-1 or EpFund-1 could have found an asymmetry, a translation that loses content in one direction on the common domain. They did not, which is why the verdict rests on interpretation.
What this does not say
- That the open theory is more fundamental tout court. The ontic verdict holds only relative to the partition in §3, which counts Everettian and orthodox readings as the candidate-fundamental ones and sets hidden-variable and collapse theories aside as different frameworks (fn. 41). “There are no other options” (p. 26) is asserted (NOTE-077, C5 weak, C9 weak).
- That Everett favours non-unitary universal dynamics. The paper shows only that they are consistent with Everett, and concedes nothing makes that description “any more likely to be true” (p. 32).
- That epistemic order confers fundamentality. EpFund-2 rests on an epistemic fact about us, imperfect preparation. The paper disclaims foundationalism but does not show why the order should rank frameworks.
- That some phenomenon forces the general theory. The authors concede that “no phenomena are known that would absolutely force us” to a not-CP or non-Markovian fundamental dynamics (fn. 83).
- That arrow-of-time or information-loss puzzles are solved. §5.1 asserts they lose their conceptual difficulty and leaves the details to others.
Connections
- PBR (LIT-062; THEORY-605) is placed by the paper inside the ontological-models framework, which it judges of limited use for orthodox readings, since those are neither ψ-ontic nor ψ-epistemic (p. 20). The terminological point matches Hance et al. (THEORY-586).
- Rovelli’s relational reading (THEORY-621) shares the denial of a closed universal system, for different reasons.
- The worked case, intertranslatable frameworks differing in fundamentality only through interpretation, bears on the record’s accounts of theoretical equivalence and dualities: THEORY-687, THEORY-690 and THEORY-459.