Michelson-Morley and the Null Result

Why a substrate is not a classical aether

Interactive: open the Michelson–Morley simulation — glide the apparatus through the substrate and watch light swim at c against the streaming medium; see 1887’s rigid-arm expectation walk the fringes as the stone turns, then zoom into one atom to watch its counter-rotating rim keep c against the wind and pack by 1/\gamma, and put those atoms back into the arms to see the fringes hold still.

The Pristine Bubble: Why You Can't Detect the Substrate from Inside dc1 substrate (has a definite rest frame) Counter-rotating boundary layer EM modons cannot cross p e observer's atom (also a quasiparticle!) modon (light) speed = c everywhere INSIDE: Lorentz-invariant acoustic metric All instruments, clocks, rulers, and light are quasiparticles of the same superfluid He-4 Analogy Phonons in superfluid helium obey ω² = c₁²k² — Lorentzian with c₁ = 238 m/s as "light speed" A phonon MM experiment inside He-4 would also give a null result Why It Works Speed of excitations is set by collective dynamics (equation of state), not individual particle frame The substrate is the metric. You can't step outside it to look. The Key Insight Your rulers contract. Your clocks dilate. Your light bends. All by exactly the amount needed to make the substrate undetectable — because they're all made of it.

The Objection

The first objection any physicist will raise: “You’re proposing a material substrate filling space. Michelson and Morley ruled that out in 1887.”

This objection is well-earned. The classical luminiferous aether was a rigid elastic medium through which light propagated as a disturbance. If such a medium existed and the Earth moved through it, the speed of light should differ in the direction of motion versus perpendicular to it. Michelson and Morley measured this difference to extraordinary precision. They found nothing. The null result was one of the key motivations for special relativity.

The substrate framework must either explain why the null result is expected, or it fails.

Why the Null Result Is Expected

Why Michelson-Morley Gets a Null Result Classical Luminiferous Aether Rigid elastic medium — fixed lattice points AETHER WIND → → → beam splitter c + v c − v (return) √(c² − v²) ✗ PREDICTS FRINGE SHIFT Light speed depends on direction of travel: Δt = (2L/c) · v²/c² At v = 30 km/s (Earth orbital speed): Δt/t ≈ 10⁻⁸ → detectable shift Observer moves through fixed medium → wind is detectable dc1 Superfluid Substrate Coherent collective motion — no fixed positions no detectable wind — collective flow IS the metric modon (photon) speed set by collective dynamics c c ✓ PREDICTS NULL RESULT Modon speed set by equation of state, not frame: ω² = c²k² (acoustic dispersion) Effective metric is Lorentzian by construction: g_μν = (ρ/c)[−(c²−v²), −vⱼ; −vᵢ, δᵢⱼ] Quasiparticles see acoustic geometry, not rest frame vs.

The dc1 substrate is not a classical aether. It is a superfluid — and superfluids have a property that rigid elastic media do not: their collective excitations obey emergent Lorentz invariance even though the superfluid itself has a definite rest frame.

This is not speculative — it is experimentally established. In superfluid helium-4, phonons (sound quasiparticles) propagate at the speed of first sound c_1 \approx 238 m/s. These phonons obey an emergent “Lorentz symmetry” with c_1 playing the role of the speed of light: their dispersion relation is \omega^2 = c_1^2 k^2 at low energies, their effective metric is Minkowskian, and no phonon can be accelerated past c_1 regardless of how much energy you add. A Michelson-Morley experiment performed by phonons on phonon clocks within the superfluid would yield a null result — even though the helium has a definite rest frame and the phonons are moving through a material medium.

The mathematics is rigorous and well-known. Barceló, Liberati, and Visser (2005) showed that any barotropic, irrotational, inviscid fluid produces an acoustic metric that is formally identical to a curved Lorentzian spacetime. Volovik (Chapter 7 of “The Universe in a Helium Droplet”) showed that the emergent Lorentz group for low-energy quasiparticles in He-3 is exact to all orders in the quasiparticle energy, as long as the energy remains far below the superfluid gap.

In the substrate framework, light (modons) and matter (orbital system complexes) are both collective excitations — quasiparticles — of the dc1 superfluid. They propagate through the substrate’s acoustic geometry, which is Lorentzian. The speed of light c is the equilibrium modon propagation speed, set by the substrate’s equation of state. The Michelson-Morley null result follows automatically: it is a measurement of the acoustic metric by acoustic instruments, and the acoustic metric is Lorentz-invariant by construction.

What the Classical Aether Got Wrong

The 19th-century aether failed because it was modeled as an elastic solid — a medium where the constituents have fixed positions and disturbances propagate as displacement waves. Such a medium has a preferred frame, and any observer moving through it can detect that motion by measuring the speed of disturbances in different directions.

A superfluid is different in a fundamental way: its constituents are in coherent collective motion, not fixed positions. The excitations are patterns in the collective flow, not displacements of individual particles. The speed of those patterns is set by the collective dynamics (the equation of state), not by the frame of any individual constituent. This is why the effective metric is Lorentzian — the excitations “see” the collective geometry, not the microscopic rest frame.

The substrate framework makes this precise: the cell vortex cores are in coherent superfluid flow. Modons (photons) are dipole vortex structures in this flow. Their propagation speed is set by the Larichev-Reznik dispersion relation, which depends on the substrate’s density and pressure — collective properties that are frame-independent for low-energy excitations.

The Other Half: No Structure to Betray a Frame

Emergent Lorentz invariance answers why the speed comes back isotropic. It does not, by itself, answer the skeptic’s other question — why a medium this dense leaves no mark in the scattering: why starlight crosses it without dimming, blurring, or splitting into diffraction orders, and why no static structure picks out a rest frame the way a crystal’s lattice vectors would. That second half has a separate answer, developed in full in The Stealth Vacuum, and it is worth stating here because it is what truly retires the aether.

A medium scatters a probe only into momentum transfers \mathbf q that its density actually contains — the scattered intensity tracks the structure factor S(\mathbf q). The classical aether was imagined as an elastic crystal, and a crystal’s S(\mathbf q) is a Bragg comb: sharp spikes at reciprocal-lattice vectors that both diffract light into orders and define a preferred frame. That is precisely the frame Michelson and Morley hunted. The dc1 lattice carries no such comb. It is a domain glass of triangular vortex crystallites — locally ordered, globally disordered-yet-uniform, the disordered hyperuniform texture the bridge equation’s \eta=1 factor already requires for isotropy. A disordered hyperuniform medium has S(\mathbf q)\to0 at small \mathbf q (no long-range scattering, no fog) and no Bragg comb at all (no diffraction grating, no reciprocal lattice to orient against the Earth’s motion). Stealthy hyperuniform materials are, for this reason, transparent even at high density (Leseur, Pierrat & Carminati 2016).

So Michelson and Morley unknowingly tested two independent things — that the medium has no preferred timing and no preferred structure — and a domain-glass superfluid passes both: Lorentz-invariant in its excitations, hyperuniform in its texture. The aether failed not because space is empty, but because the medium that fills it is built, dynamically and structurally, to be unobservable.

Where Lorentz Invariance Could Break Down

Where Lorentz Invariance Breaks Down: The Distinguishing Prediction wavevector k energy ω low k intermediate k → k_Planck Lorentz-invariant regime Planck-scale breakdown breakdown scale ω = ck (exact Lorentz) roton minimum the "dip" — unique to superfluid substrates maxon generic QG correction ω² = c²k² ± k³/M_Planck monotonic — no dip Substrate prediction (phonon → maxon → roton → recovery) Generic quantum gravity correction The roton-minimum signature is qualitatively different from any generic QG prediction. Observable in ultra-high-energy cosmic ray spectra or gamma-ray burst dispersion (Fermi-LAT, CTA). The substrate leaves a fingerprint that a simple Planck-scale cutoff does not.

If the substrate is real, emergent Lorentz invariance should fail at sufficiently high energies — where the quasiparticle description breaks down and the probe resolves the substrate’s granularity. This is the same phenomenon observed in superfluid helium: phonons obey Lorentz symmetry at low k, but at k approaching the roton minimum (k \sim 2 \times 10^{10} m^{-1}), the dispersion relation curves and “Lorentz invariance” fails.

For the dc1 substrate, the breakdown scale is set by the lattice’s own granularity — not the Planck energy but the cell scale \xi \approx 100\;\mum, whose conjugate energy \hbar c/\xi = m_1 c^2 \approx 2 meV (\simTHz) is the substrate’s own “Planck scale” (Roton, Maxon, and the Edge of Spacetime). That sounds catastrophic — visible light sits far above that scale — but the photon is not the substrate’s sound mode: it is a topologically protected modon that propagates at c with no momentum dependence, which is why gamma-ray-burst timing (Fermi-LAT) and fast-radio-burst dispersion see nothing at any energy. What does ride the collective sound branch is the gravitational wave, sitting {\sim}10^{10} below the cell scale — linear to {\sim}10^{-20}, comfortably inside GW170817’s 10^{-15} bound.

Distinguishing prediction: The framework’s Lorentz-invariance edge is therefore not a high-energy cubic correction \omega^2 = c^2 k^2 \pm k^3 / M_\text{Planck} but a low-energy crossover: near \lambda \sim \xi (the $$0.3–10 THz band) the solitonic modon gives way to a delocalized collective mode — the modon→phonon crossover (Photon as Modon). And because the self-organized substrate parks at its marginal point, the realized sound branch is monotonic — the deep helium-style roton dip is an off-critical feature the vacuum does not show (The Marginal Point). The observable window is the terahertz gap, not ultra-high-energy astrophysics.

Separately, the substrate supports slow Tkachenko shear modes at c_T \approx 9 km/s \approx 3 \times 10^{-5}\,c — five orders of magnitude below the speed of light. These are internal lattice oscillations with no counterpart in standard physics (see the canonical mode inventory for the full wave-mode analysis and the sector rule that sorts it). The Tkachenko speed is a zero-parameter prediction of the framework.

The Same Wager in a Different Phase: the Planck–Kleinert Crystal

The elastic-solid aether has a modern, mathematically serious form: Kleinert’s world crystal. This theory proposes that spacetime is a Planck-scale lattice and gravity is the continuum theory of its defects — dislocations as torsion, disclinations as curvature ([R178]). From that Danielewski and Sapa’s quaternion quantum mechanics derives Klein–Gordon, Poisson, and Schrödinger equations from Cauchy’s linear elasticity on that lattice ([R179]). Like this framework, they propose that the vacuum is a real lattice medium and particles are its solitons. It assumes an ideal elastic solid built from Planck masses at the Planck length. Here that lattice is replaced a self-binding superfluid built from 2 meV particles at a 97\;\mum cell.

Both reject the Born reading and make \psi a real field of the medium. With the elastic solid curvature comes from a rescaled deformation potential, compared to here that uses the condensate density and phase through Madelung. Both split the medium’s motion into a scalar compression and a vector twist; their quaternion \sigma=\sigma_0+\hat\phi is the breath-plus-circulation pair of the substrate cell, and its Cauchy–Riemann operator is the one that gives this framework its first-order equation. Both read mass as the medium’s internal cycle at the Compton frequency, spin \tfrac12 as a geometric twist, and gravity as a change in the medium’s wave speed. Where they differ is decidable:

Axis Planck–Kleinert crystal Light Fluid (dc1)
Medium ideal elastic fcc solid self-binding superfluid vortex lattice
Constituent, scale Planck mass, Planck length — assumed dc1 \approx2 meV, 97\;\mum — derived from \rho_\text{DM}, \hbar, c, and a Bessel zero
Wavefunction deformation potential condensate density and phase
Schrödinger from minimizing elastic energy boundary-layer reaction force
Solitons from inserted coupling term G_0\sigma\sigma^* logarithmic equation of state
Spin \tfrac12 quaternion non-commutativity odd boundary parity (same double cover)
Photon transverse shear wave modon
Propagation speeds two: c and \sqrt3\,c (longitudinal) one
Preferred frame present, unaddressed emergent Lorentz invariance; hyperuniform texture
Vacuum energy gravitating Maxwell’s objection left open at \rho_P\sim10^{97} kg/m³ Gibbs–Duhem self-tuning
Outputs Planck-unit identities (\hbar=m_Pcl_P, G=l_P^3/t_P^2m_P) \xi, Higgs VEV, Koide, \alpha, a_0, c_\text{GW}, with stated errors

Three discriminators could be checked:

  1. A second wave speed. A Cauchy solid with Poisson ratio 0.25 carries a longitudinal (compression) wave at \sqrt3\,c alongside the transverse wave at c (their eq. 17). The substrate has one speed for scalar, vector, and tensor modes alike, and GW170817’s |c_\text{GW}/c-1|<6\times10^{-15} is the tensor case of that claim. A crystal must explain why the longitudinal branch is never seen; a superfluid never has it.
  2. A detectable rest frame. An elastic crystal has a Bragg comb and a lattice frame, which is exactly the frame Michelson and Morley hunted and the tests above fail. A superfluid’s excitations ride an acoustic metric that is Lorentzian by construction, and a hyperuniform domain glass has no comb to orient against.
  3. Is the lattice scale an input or an output? The Planck–Kleinert crystal sets its cell to the Planck length by hypothesis; with Planck units as inputs, \hbar=m_Pcl_P and G=l_P^3/(t_P^2m_P) are definitions, and E=hf follows from the same input. The substrate’s cell size is an output of measured constants, and Volovik’s own laboratory analog says that is the expected kind of answer — in superfluid ^4He the effective Planck length comes out at the interatomic distance, a property of the medium located empirically (Emergent Speed of Light).

Summary

Test Classical Aether Substrate Framework
Michelson-Morley null result Fails (predicts fringe shift) Passes (emergent Lorentz invariance from superfluid acoustic metric)
Lorentz invariance at low energy Violated (preferred frame detectable) Exact (quasiparticles see Lorentzian effective metric)
Lorentz invariance at Planck scale N/A Predicts breakdown with roton-minimum spectral signature
Preferred frame detectable? Yes (wind in the aether) No (superfluid flow is the metric, not a background)
Static structure (scattering/diffraction) Bragg comb: diffracts starlight, reciprocal vectors fix a frame None: disordered hyperuniform, no comb, transparent (Stealth Vacuum)

The substrate is not an aether. It is the medium whose collective excitations are spacetime. Michelson and Morley didn’t rule it out — they confirmed that the low-energy physics is exactly what a superfluid acoustic geometry predicts.

This chapter answered the skeptic’s first question — why doesn’t uniform motion through the medium betray a preferred frame? — but not the second: rotation through a medium is detectable, in the Sagnac effect that the same Michelson used to read the Earth’s spin off an interferometer in 1925. The two experiments are the curl-free and rotational readings of one line integral, \oint\vec v\cdot d\vec\ell: zero for translation (null), 2\vec\Omega\cdot\vec A of enclosed circulation for rotation — the irrotational-except-for-circulation signature of a superfluid, and, for matter waves, a fringe count equal to the number of substrate circulation quanta threading the loop. The companion chapter develops it.