NOTE-140: Gao — The Meaning of the Wave Function

Contribution

The book assembles in one place Gao’s twenty-year programme (from 1993). There are three main theses. (1) A protective-measurement argument that ψ is ontic, which Gao says avoids PBR’s preparation independence and Hardy’s ontic indifference and also applies to deterministic theories (ch. 4). (2) A new ontological reading of ψ as the state of random discontinuous motion of particles in 3-space (chs. 6–7). (3) An argument that this ontology favours real collapse, with a concrete energy-conserving discrete collapse model and a preferred-frame consequence (chs. 8–9). There is also a derivation of the free Schrödinger equation from spacetime-translation invariance plus a relativistic energy–momentum relation, with linearity assumed (ch. 5). And there are critical surveys of ψ-epistemic views, ψ-ontology theorems, nomologism, Albert’s wave-function realism, Everett, Bohm, Penrose’s gravity-induced collapse and CSL.

Key insight

Projective measurements give only a probabilistic link between ψ and experience. Protective measurements give a definite one: a single protected system shifts the pointer by exactly ⟨A⟩. If one is a realist about classical mechanics because its measurements have definite results, consistency requires realism about ψ. The quantity measured, ρ(x) = |ψ(x)|² times a charge, is a real charge density, but it cannot be a field, because its parts do not repel. So it must be the time-average of a point charge that is at one place at each instant and visits every place in every infinitesimal interval. That is possible only if the motion is discontinuous and random.

Assumptions

• Protective-measurement premise. In the Zeno or adiabatic limit the protected branch has amplitude → 1, and the pointer shifts by ∫g(t)dt·⟨A⟩ = ⟨A⟩ with no entanglement (eqs. 1.9–1.10, 4.2–4.4). The protection procedure carries no information about ψ (pp. 36–37; contested, see corrections).
• Ontological-models second assumption. The measuring device’s behaviour is fixed by the system’s ontic state together with the device’s properties. For protective measurement, the ontic state determines the definite result (pp. 32–33).
• Criterion of reality (§4.4, pp. 41–42). If a measurement yields a definite pointer value, and the pointer-shift rate during the measurement is also determined by that value, the value reflects a property of the system.
• Linearity of free evolution is a stated presupposition of the ch. 5 derivation (pp. 46, 51).
• Field self-interaction premise. Any two co-existing parts of a charged or massive distribution interact (pp. 62–63, 168).
• Uniform instantaneous condition. The instantaneous condition fixing position change is “always deterministic or indeterministic for any state of motion”. Gao calls this “seemingly reasonable” and leaves it for future work (p. 67).
• Psychophysical supervenience. A mental state supervenes on a whole physical state, not on a part of it (fn. 7, p. 100). Mental states are not instantaneous (p. 108).
• Discrete time. Instants have duration t_P, motivated by minimal-length arguments (Garay 1995) and by the claim that randomness cannot survive integration in continuous time (pp. 120–122).
• Conservation of energy at ensemble level for collapse (§8.4.2), and a “meta-law” that nature lets its laws manifest (p. 138).

Key results

• Protective measurement (§1.3). With Zeno protection over [0,τ] at N → ∞ times, |t=τ⟩ = |ψ⟩|φ(x₀ + ⟨A⟩)⟩ (eq. 1.10). By Ehrenfest, any non-disturbing measurement shifts the pointer by ⟨ψ|A|ψ⟩ (eqs. 1.12–1.13). Taking A as a normalised projector on a region Vₙ and B = (ħ/2mi)(A∇+∇A), the results give ρ and j, and so ψ up to a global phase (eqs. 1.14–1.16).
• ψ-ontology (§4.2). Spin example: for |0⟩ and |+⟩, each protected by a magnetic field along its own spin direction, a protective measurement of P₀ = |0⟩⟨0| gives 1 and 1/2 respectively (p. 33). If both states were compatible with one λ, the results would coincide. For realistic protective measurements, a positive overlap probability p leads to a contradiction as conditions approach the ideal (p. 35, fn. 7).
• §4.3. The ψ-ontology theorems fail if the ontic state can change discontinuously during the measurement. The protective argument survives, because the pointer-shift rate at each instant is g(t)⟨A⟩ (eq. 4.1) and the protected ψ is constant.
• Schrödinger equation (ch. 5). Time-translation invariance plus linearity give dH/dt = 0. Space-translation invariance gives [p̂,H] = 0, hence solutions e^{i(px−Et)}. The four-vector (E/c, −p) gives E² = p²c² + m²c⁴, hence H = p̂²/2m and eq. (5.14), with ħ absorbed into the units of m.
• Charge density (§6.2). A protective measurement of charge in region V gives Q∫_V|ψ|²dv (eq. 6.4). So “a charged quantum system such as an electron is a charge cloud”, confirming Schrödinger’s charge-density hypothesis (p. 61).
• Effective, not continuous (§6.3). A continuous distribution would self-interact, which contradicts the superposition principle and hydrogen spectra. Continuous ergodic motion fails for momentum eigenstates, nodes (infinite speed) and two separated boxes. So the motion is discontinuous, with presence density = |ψ|².
• Configuration space (§7.2). Under Galilean transformations each coordinate triple transforms as a 3-space point. N mass parameters are each tied to one triple. A position of configuration space cannot host two distinct point masses. So an N-body ψ describes N entities in 3-space, not one entity in 3N-space.
• Entanglement by time-division (§7.2.3). Treating an entangled two-body state as both configurations co-existing at one instant would give a product state. So the configurations alternate on dense, discontinuous instant sets, and entangled particles jump “in a precisely simultaneous way”.
• Measure-theoretic description (§7.3.2). The position density ρ(x,t) = lim M_{Δx,Δt}/(ΔxΔt) (eq. 7.4) and the flux density j satisfy continuity (eq. 7.5). ψ = √ρ exp(im∫v dx/ħ) (eq. 7.8). |ψ|² is read as a single-case propensity (p. 88).
• Spin and the Kochen–Specker theorem (§7.3.4). RDM applies to position, momentum and energy. The spin of a free system is definite along one direction and does not undergo RDM. Gao takes the finite-observables option for evading Kochen–Specker, appealing to Occam’s razor (p. 91).
• Measurement problem (§8.1). Gao reformulates it as the incompatibility of (A1) mental state supervenes on ψ and (A2) linear dynamics. Against Everett: the weak (many-minds-like) form violates supervenience; a NOT-like unitary that swaps branches leaves ψ unchanged but changes each mind’s content (p. 101). Against Bohm: supervenience on relative particle positions is inconsistent with the Born rule when branches are spatial translations of each other (pp. 103–104). Bohmian particles are “more hidden” than ψ (p. 116).
• Collapse model (§8.4). Stays in branch Eᵢ at each t_P shift the probabilities as ΔPᵢ = k(1−Pᵢ). Ensemble energy conservation forces ΔPᵢ/(1−Pᵢ) = k for all i (eq. 8.19). Off-diagonals decay by (1−k²) per step (eq. 8.20), so τ_c ≈ t_P/k². With k ≈ ΔE t_P/ħ this gives τ_c ≈ ħE_P/(ΔE)² (eq. 8.26), where E_P = h/t_P as printed. The equation is scale-invariant (eq. 8.30). Energy is discretised by the cosmic event horizon: E₁ = hc/4R_U ≈ 10⁻³³ eV for photons, ≈ 10⁻⁷² eV for electrons (p. 130).
• Numbers (pp. 131–137). Atomic photon, ΔE ~ 10⁻⁶ eV: τ_c ~ 10²⁵ s. SQUID (Friedman et al. 2000), ΔE ≈ 8.6×10⁻⁶ eV: τ_c ~ 10²³ s. ¹⁸⁰Ta isomer, 75 keV: about 20 min. Avalanche photodiode, ΔE ≈ 2.5×10¹¹ eV: τ_c ≈ 1.25×10⁻¹⁰ s. Geiger counter, ΔE ≈ 10⁹ eV: 10⁻⁵ s. Dust particle (10⁻⁵ cm) in air: about 10⁻⁴ s. One neuron, ΔE ≈ 10⁴ eV: 10⁵ s. 10⁷ neurons: about 10⁻⁹ s, so perceptions are definite.
• Penrose and CSL (§8.5). Penrose’s unstable-particle analogy is too weak, and the Schrödinger–Newton self-gravity implies an electrostatic self-interaction that is ruled out. CSL’s smeared mass density contradicts protective measurements, and its noise field lacks a legitimate source.
• Relativity (ch. 9). Under standard synchrony, RDM and collapse appear distorted in every frame but one: two particles at one instant, correlation reversal in a fraction 2|ab|² of the time, non-normalised states. None of this is observable. Absolute simultaneity can be restored conventionally: x′ = (x−vt)/√(1−v²/c²), t′ = √(1−v²/c²)·t (eqs. 9.16–9.17). Either way there is a preferred frame, and Gao argues it is detectable through τ_c (eq. 9.22), expected to be the CMB frame. QFT’s ontology is still RDM particles, now created and annihilated (§9.4).

Claims

id	claim	strength	support
C1	Protective measurement yields ⟨A⟩ on a single system with certainty in the ideal limit	strong	standard derivation (Aharonov–Anandan–Vaidman), eqs. 1.9–1.13
C2	Within the extended ontological-models framework plus preparation non-contextuality, ψ is ontic	moderate	argument by contradiction, pp. 33–35
C3	ψ is ontic with “no auxiliary assumptions”	weak	pp. 35–37. It rests on the premise that the protection procedure has no information about ψ, which fn. 10 (p. 37) partly concedes is false for Zeno and adiabatic protection
C4	The weaker criterion of reality establishes ψ-reality without assuming ontic states	weak	§4.4. The criterion is proposed, not independently justified. The argument is from parity with classical mechanics
C5	The free Schrödinger equation follows from spacetime-translation invariance and relativistic invariance	moderate	derivation, ch. 5. It assumes linearity and a scalar analytic ψ(x,t), and absorbs ħ into m. Gao flags linearity as essential
C6	An electron has a real charge density	ψ	²Q
C7	The charge density is effective (a moving point particle), not a continuous field	weak–moderate	argument from absent self-interaction, pp. 62–64. It depends on the field self-interaction premise
C8	Ergodic motion generating	ψ	² must be discontinuous
C9	The motion is random	weak	p. 67. It depends on an assumption Gao explicitly leaves unargued
C10	An N-body ψ describes N particles in 3-space, and entangled particles jump synchronously	moderate	Galilean transformation, mass parameters, time-division argument, §7.2
C11	Discontinuous motion is mathematically “more natural” because continuous functions have measure zero	weak	pp. 84–85. There is no canonical probability measure on “all functions”, so the argument is heuristic
C12	The weak form of Everett violates psychophysical supervenience	weak–moderate	argument from the branch-swapping unitary, p. 101
C13	Supervenience of mind on Bohmian particle positions is inconsistent with the Born rule	weak	argument from translated branches, pp. 103–104. Bohmians would reply that positions within the full configuration, not relative configurations, fix the outcome
C14	The Born probabilities originate in RDM	weak	a non-coincidence argument, pp. 108–109. Gao calls it an “assumption”
C15	An energy-conserving collapse model with τ_c ≈ ħE_P/(ΔE)² gives definite outcomes, localisation and definite perceptions	moderate (derivation) / weak (physical basis)	eqs. 8.14–8.29 are derived from stated rules. The “deeper basis” (§8.4.5) is admitted to be “very speculative” (p. 139)
C16	A ¹⁸⁰Ta ground/isomer superposition collapses in about 20 min	moderate (as a model prediction)	eq. 8.26 at a 75 keV gap. Untested
C17	A preferred Lorentz frame exists and is detectable through seasonal variation of collapse times (Δτ_c ≈ 4×10⁻⁴ τ_c)	weak	eqs. 9.18–9.22. It assumes the collapse formula holds in one frame and ΔE transforms as for E ≈ pc
C18	Schrödinger–Newton and CSL’s smeared density are ruled out by the absent self-interaction and by protective measurements	weak	inherits C6–C7

Method

Gao works by conceptual argument plus elementary quantum-mechanical derivation (ch. 1, ch. 4, ch. 5), and builds a phenomenological model (§8.4). The same strategy runs through the book: take protective measurement as a definite link between ψ and experience; apply one “universal” rule connecting experience to reality to both classical and quantum mechanics; then ask which ontology (field, configuration-space object, or particle) can host the properties so revealed. The collapse model is set up as a discrete Markov process on the branch probabilities. Its update rule is fixed by requiring ensemble-level energy conservation and dimensional analysis in t_P.

Concepts

• Protective measurement. A measurement during which the state is kept unchanged, by the Zeno effect or by adiabatic weak coupling to a nondegenerate energy eigenstate, so that one system yields ⟨A⟩ with certainty.
• Random discontinuous motion (RDM). At each (discrete) instant a particle is at one position, chosen at random with density |ψ|². Over any interval it visits all of supp ψ, and the motion has no trajectory or classical velocity.
• Position density / flux density. ρ(x,t) is the Lebesgue-measure density of the set of instants at which the particle is at x (eq. 7.4). j(x,t) = ρv satisfies continuity. Together they fix ψ up to a phase.
• Effective charge distribution. The time-average over an infinitesimal interval of a point charge’s RDM. It equals |ψ|²Q and does not interact with itself.
• Weaker criterion of reality. A definite pointer value whose shift rate is also fixed by that value marks a real property (§4.4).
• Weak vs strong Everett. Strong: several physical observers after measurement. Weak: one physical observer with several minds. Gao classes Wallace (2012) as the weak form (fn. 6, p. 100).
• Energy-conserved collapse. Collapse toward energy eigenstates, driven by ΔE summed over subsystems, conserving the energy distribution at ensemble level.
• Preferred Lorentz frame. The frame in which collapse is simultaneous and the one-way speed of light is isotropic (§9.2).

Connections

• ψ-ontology. It extends and competes with PBR (LIT-062). Gao presents it as free of PBR’s preparation independence. Hance et al. (LIT-090) on whether ψ-ontic and ψ-epistemic exhaust the options bears on Gao’s parity argument. The principal published rebuttal, Combes, Ferrie, Leifer & Pusey (2015, arXiv:1509.08893), is cited and answered in one footnote. It is not held in the record.
• Wave-function ontologies. Chen (LIT-204) lists this book among the book-length treatments. Allori’s functionalism (LIT-200) and Belot’s case study (LIT-100; Gao cites Belot 2012 at fn. 5, p. 30) are the neighbouring positions. Poltorak (LIT-164) works within the configuration-space wave-function realism that ch. 7.1 rejects. Carroll’s Hilbert-space fundamentalism (LIT-123) is the opposite pole: for Carroll the universal state is fundamental, for Gao ψ is a property of 3-space particles.
• Dorato (LIT-176). Both reject Esfeld et al.’s holistic dispositionalism, by parallel effective-wave-function arguments (Gao pp. 29–30; Dorato pp. 14–15). Dorato locates powers in events. Gao’s propensity reading of |ψ|² (p. 88) is itself dispositional.
• Rovelli (LIT-177). RQM is the antithesis on both counts: no ontic ψ, and no preferred frame. Rovelli (1994) is among the critics of protective measurement that Gao lists (p. 33).
• Relativity and simultaneity. Ch. 9 connects to the conventionality-of-simultaneity literature (Reichenbach, Grünbaum, Janis). The record holds none of it.

Bearing on the record

• It is the record’s fullest source for the protective-measurement argument and the RDM ontology, and a secondary source for the Penrose and CSL critiques. Any record account asserting “protective measurement establishes ψ-reality” should cite it together with Combes et al. (2015) and state the protection-procedure premise.
• Metaphysics: Gao defends a particle ontology in 3-space, discontinuous random motion, single-case propensities, discrete time, and absolute simultaneity or a preferred frame. He presupposes a “universal approach” to ontology: the same experience-to-reality rule for every theory (Epilogue, fn. 11).
• No ML-practice content. Nothing here bears on the Anthology of the SOTA.

Limitations

• The load-bearing premise of ch. 4 (the protection procedure does not “know” ψ) is contested, and the book’s reply is brief. Protection by Zeno effect or adiabatic potential already restricts ψ to a discrete eigenbasis that the experimenter chose.
• The field-self-interaction premise decides between particle and field ontologies in chs. 6 and 9 and in the Epilogue. It is asserted as “a fundamental assumption in physics” (p. 168), not argued.
• The argument that the motion is random (C9) rests on an assumption explicitly deferred (p. 67). The measure-theoretic “naturalness” argument (C11) is not rigorous.
• The Bohm and Everett critiques in §8.1 turn on specific readings of psychophysical supervenience that proponents reject.
• The collapse model’s basis (tendencies to “stay”, a “meta-law”) is admitted to be speculative (p. 139). Its E_P is printed inconsistently: E_P = h/t_P on p. 127, while eq. (9.18) uses ħ²/t_P, a factor-2π difference absorbed into “order unity”.
• Ch. 9 and several footnotes (possible superluminal signalling via conscious observers, fn. 9 p. 103 and p. 160; dark energy from discrete spacetime, fn. 36 p. 130) are speculative and lean heavily on the author’s own earlier papers.
• This reading is of the arXiv draft. The CUP text may have been revised.

Open questions

• Does any protective-measurement argument survive once the protection is required to be independent of ψ, i.e. once the experimenter is not allowed to pick the eigenbasis that contains ψ? What would settle it is a protocol that measures ⟨A⟩ on an unknown ψ drawn from a continuous set.
• Can the energy-conserving collapse model be tested on nuclear-isomer superpositions (¹⁸⁰Ta, τ_c ~ 20 min), and does it survive current collapse-model bounds (e.g. X-ray spontaneous-emission tests)? The book gives no comparison.
• Is RDM consistent with relativistic QFT beyond the heuristic in §9.4? In particular, how does it handle the absence of local number operators, which Gao sets aside as irrelevant?
• Is the seasonal collapse-time variation (Δτ_c ≈ 4×10⁻⁴ τ_c) measurable once environmental decoherence is excluded (fn. 9, p. 162)?

Corrections to the seeded skim

• Length: this is a monograph over 150 pp. (194 PDF pp.), and it was read in full, references included.
• The dossier’s summary, “protective measurements, combined with the ontological-models assumption”, undersells ch. 4. Gao gives three routes. (a) Within an extended ontological-models framework, using an auxiliary preparation non-contextuality assumption (pp. 33–35). (b) Two versions he says need “no auxiliary assumptions” (p. 35). They rest on the premise that a protective measurement’s result depends only on the measured observable and the ontic state. (c) A new “weaker criterion of reality” outside the framework (§4.4, pp. 41–43): a definite result whose pointer-shift rate is also fixed by that value reflects a property of the system. All three turn on one premise, that the protection procedure contains “no information about the measured wave function” (pp. 36–37). Gao’s own fn. 10 (p. 37) undercuts it. Zeno protection needs an observable O of which ψ is a nondegenerate eigenstate, and adiabatic protection needs ψ to be a nondegenerate eigenstate of H, so the procedure selects ψ out of a discrete set. This is the Combes, Ferrie, Leifer & Pusey (2015) objection. Gao answers it in fn. 13 (p. 40) only by asserting that their models “implicitly assume … the protection procedure knows the measured wave function”. The dossier should state this as the argument’s load-bearing premise.
• The dossier says RDM “requires a stochastic, nonlinear collapse term”. The requirement is conditional. RDM on its own is Bell’s “Everett (?)” theory (§7.4.4, §8.2). Gao grants that it reproduces quantum predictions, but rejects it because instantaneous mental states are “implausible” (p. 108). Collapse follows only from a further assumption, argued as a non-coincidence, that the Born probabilities originate in RDM (§8.3). Everett is then excluded as purely unitary, and Bohm cannot make the probabilities come wholly from RDM (§8.3.2).
• The dossier omits the collapse model’s specifics and its testable predictions. Preferred basis: energy eigenstates. Energy is conserved at ensemble level. Time is discrete at t_P. The collapse law is eq. (8.28), and the many-body ΔE is the sum of absolute subsystem uncertainties, not the total-energy uncertainty (eq. 8.29), which is how Gao evades Pearle’s (2004) objection to energy-driven models. Predictions: a superposition of ¹⁸⁰Ta’s ground state and its 75 keV metastable isomer should collapse in about 20 min (p. 131). A detectable preferred frame gives τ_c ≈ (1+v/c)⁻² ħ²/(t_P ΔE²) (eq. 9.22), so collapse times should vary seasonally by Δτ_c ≈ 4×10⁻⁴ τ_c (p. 162).
• The Epilogue’s claim that particle and field ontologies “may … result in different predictions” (pp. 168–169) is not a pending test. The observation it relies on, no self-interaction between two wavepackets of one electron, is already made. What does the work is a premise Gao calls “a fundamental assumption in physics”: any two parts of a charged field interact, while a particle at one instant does not interact with itself at another. Wallace-style spacetime state realism and Ghirardi’s mass density deny that premise.
• Pages: the Epilogue runs pp. 165–171 (the dossier gives 165–168), and the references pp. 172–186.
• Ch. 3’s critique targets both Dürr–Goldstein–Zanghì nomologism and Esfeld et al.’s (2013) Bohmian dispositionalism (p. 29; fn. 5, p. 30, invoking Belot 2012). Gao cites Dorato (2015, “Laws of nature and the reality of the wave function”) in fn. 4, p. 29. So it bears on Dorato (LIT-176) and Belot (LIT-100) as well as Allori.
• metaphysics tag: justified. The book defends theses about what exists: particles, not fields or configuration-space objects. It holds that motion is essentially discontinuous and random, that ψ is a single-case propensity property, that time is discrete (Planck time), and that simultaneity is absolute or there is a preferred frame. It also takes a stand on psychophysical supervenience (§8.1).
• A consciousness tag is arguable (§8.1’s mind-level formulation of the measurement problem, the “vividness” conjecture on p. 105, the neuron collapse estimate on p. 137) but not recommended. That material serves the measurement problem and does not present a theory of experience.