Follow-up Q20 (Ives): Is empty space like a pond?

Date: 2026-10-08. Role: STEP F follow-up investigator. Web time: about 2 minutes by the timer (start 1791505788, end 1791505902, 114 s); Chrome for searching and browsing, curl for saving Wikipedia/Wikisource wikitext to research/pages/ (not published). The question was answerable from standard, well-documented experiments, so the 20-minute cap was not needed.

The question (owner's words, lightly polished)

Is empty space like a pond? Two equal waves combined in phase give the sum of both at that point. The premise: a pond is special because we can see the water; if we could only see the waves it might be more like empty space, and crossing beams that do not seem to combine would point to something special about space, perhaps a preferred direction. What do experiments show: crossing coherent beams and their fringes; how Wiener's 1890 standing-wave experiment and Ives's own thin-film work bear on it; tests for a preferred direction (Michelson-Morley, Hughes-Drever, modern cavity tests); what a wave-only observer could still infer about a medium (dispersion, loss, frame dependence); and whether space is a medium with properties (epsilon0, mu0, displacement current, the quantum vacuum, Einstein's 1920 Leiden ether) but no rest frame.

What was searched

What the project's own files say

• Ives 1951 Rumford lecture, "Adventures with Standing Light Waves" (Proc. Am. Acad. Arts Sci. 81), p.11-12 [read: _text]: Young and Fresnel showed "two beams of light, in constant phase relation, could be superposed to produce regions of darkness"; the extreme case is two beams in the same line in opposite directions, giving "layers in continuous agitation, between which are regions of no activity". Wiener (1890) put a very thin photographic film at a small angle above a mirror: "At the nodes there should be no photographic action, at the loops, strong action. This was in fact what he found". Ives: corpuscles "would have to be imagined as passing out of existence and being reborn, as they went to and from the mirror. Waves of course present no comparable difficulty."
• Same lecture, p.30 [read: _text]: of the Ives-Fry repetition of Wiener's experiment, "Electrons are emitted with energy hv, but they follow minutely the orders of the standing wave pattern"; and the verification "considerably transcends ... the original simple experiment of Wiener".
• Ives and Fry, JOSA 23:73 (1933), p.73 [read: _text]: in Wiener's and in Drude and Nernst's (fluorescent probe) experiments "the action of the light was found to be associated with the electric vector"; the alkali film "plays the same role as the probe surface used by Wiener, Drude and Nernst, except that it is in contact with the reflecting surface"; the aim was to see "whether the photoelectric current followed the very considerable alterations in the energy density which computations indicate". Discussion (p.11): the results are "a further and more detailed confirmation of the point in classical wave theory investigated by Wiener", now quantitative.
• Primer 4.8 and claims.md (thin-film work) [read]: photoemission from thin alkali films was found proportional to the standing-wave intensity in the film (NAS memoir pp.21-22); selective/normal ratios 15-25 (ApJ 60:209, 1924).
• Primer 4.9 and arbiter 7d (ripple tank, Ives and Stilwell, JOSA 31:14, 1941; 32:25, 1942) [read]: capillary ripples on mercury, wave speed 19.20 cm/s, sources moving at 11.28 cm/s (v/c = 0.585); moving patterns match the stationary ones after FitzGerald and Larmor-Lorentz scaling; one ratio measured 0.852 against the ideal 0.658, which the authors put down to dispersion of capillary waves.

What was found on the web

1. Wiener 1890. "Wiener had demonstrated, in the context of the electromagnetic theory, that the electric field is the 'active ingredient' in light waves" [web: https://skullsinthestars.com/2008/05/04/classic-science-paper-otto-wieners-experiment-1890/]. The page explains that at a mirror the electric field has a node and the magnetic field a maximum at the surface, and the blackening followed the electric-field antinodes. Photonics Spectra dictionary (snippet): the nodes found "correspond to the nodes of the electric vector" [web: https://www.photonics.com, search snippet]. Stetson (JOSA 1968, snippet): Wiener's results "establish that the photographic process is sensitive to only the E vector and not to the H vector" [web: https://opg.optica.org, search snippet]. A 2024 arXiv paper (J. Chen, "The fundamental factor of optical interference") questions the "only the electric field" reading [web: arxiv.org, snippet only, not read]; recorded as a dissenting lead.
2. Independent beams do combine. Magyar and Mandel, "Interference fringes produced by superposition of two independent maser light beams", Nature 198:255-256 (1963): fringes photographed from two independent pulsed ruby lasers [web: citation in Kin 2021 arXiv and the UBA PDF, search snippets]; Pfleegor and Mandel (1967) did it with two independent single-mode lasers [web: ui.adsabs.harvard.edu snippet]. Doug Marett (conspiracyoflight.com, 2010; the same site that hosts the Ives papers): two independent HeNe lasers on one photodetector give beat signals that are "interference patterns changing at the difference frequency"; the beats come from "crossed modes of the same polarization", the orthogonally polarized pair only shows up through a polarizer at 45 degrees, and "block one beam path - on doing so, all beat frequencies disappear" [web: http://www.conspiracyoflight.com/LaserInterference/LaserInterference.html]. Marett presents it as refuting Dirac's "interference between two different photons never occurs"; the mainstream reading is that Dirac's sentence is about single-photon amplitudes and that classical-field interference of independent lasers is expected (Mandel's own interpretation) [memory: medium].
3. Fresnel-Arago laws. "Two orthogonal, coherent linearly polarized waves cannot interfere (because there is no interference cross term)" [web: https://scienceworld.wolfram.com, snippet; https://en.wikipedia.org/wiki/Fresnel–Arago_laws, snippet]. Rath, Kanseri et al. (Am. J. Phys. 2008) verified the laws in a Mach-Zehnder interferometer: "light waves with orthogonal polarization do not interfere" [web: https://web.iitd.ac.in/~bkanseri/AJP2008, snippet].
4. Preferred direction for light. "The current precision, by which an anisotropy of the speed of light can be excluded, is at the 10^-17 level", in relation to the Solar System's ~368 km/s motion relative to the cosmic microwave background (CMB) [web: https://en.wikipedia.org/wiki/Modern_searches_for_Lorentz_violation, citing Herrmann et al., Phys. Rev. D 80:105011 (2009) and Eisele et al., PRL 103:090401 (2009)]. "As of 2015, optical and microwave resonator experiments have improved this limit to Delta c/c ≈ 10^-18" (Nagel et al. 2015, rotating orthogonal microwave resonators); Brillet and Hall (1979) were at about 5 x 10^-9 in their table [web: https://en.wikipedia.org/wiki/Michelson–Morley_experiment].
5. Preferred direction for matter (Hughes-Drever). Hughes et al. (PRL 4:342, 1960) and Drever (1961): "spectroscopic tests of the isotropy of mass and space"; "The sharp, unsplit NMR line of this isotope of lithium is evidence for the isotropy of mass and space"; they "test the isotropy of the interactions of matter itself, that is, of protons, neutrons, and electrons". Later clock-comparison limits in the table run from 10^-25 to 10^-30 GeV; the Lorentz-violation page gives comagnetometer limits at the 10^-33 GeV level [web: https://en.wikipedia.org/wiki/Hughes–Drever_experiment; Modern_searches page].
6. Dispersion of the vacuum. Fermi-LAT (Nature 462:331, 2009, GRB 090510) and later burst studies: "no energy dependence and thus no observable Lorentz violation occurs in the photon sector even beyond the Planck energy". In 2007 MAGIC "claimed a possible energy dependence of the speed of photons from the galaxy Markarian 501"; the page says it "was superseded by the substantially more precise measurements of the Fermi-LAT group" [web: Modern_searches page]. Cosmological birefringence searches (Wardle et al. 1997, "Observational Evidence against Birefringence over Cosmological Distances") also found none [web: same page, reference list].
7. Light does scatter light, weakly. Light-by-light scattering "has first been seen in 2016 by the ATLAS collaboration ... and was then confirmed by the CMS collaboration", in ultraperipheral lead-lead collisions [web: https://en.wikipedia.org/wiki/Two-photon_physics].
8. epsilon0 and mu0. Since the 2019 SI redefinition, "the electron charge became a numerically defined quantity, making epsilon0 and mu0 measured quantities, neither of them exact, but related by the equation epsilon0 mu0 c^2 = 1"; epsilon0 ≈ 8.8541878 x 10^-12 F/m [web: https://en.wikipedia.org/wiki/Vacuum_permittivity].
9. Einstein, Leiden 1920 [web: https://en.wikisource.org/wiki/Ether_and_the_Theory_of_Relativity]: "according to the general theory of relativity space is endowed with physical qualities; in this sense, therefore, there exists an ether"; "space without ether is unthinkable; for in such space there not only would be no propagation of light, but also no possibility of existence for standards of space and time"; "But this ether may not be thought of as endowed with the quality characteristic of ponderable media, as consisting of parts which may be tracked through time. The idea of motion may not be applied to it."

Reasoning: from the evidence to the answer

1. Do crossing beams combine? Yes, wherever they overlap, by the same rule as pond ripples: the fields add (two equal in-phase waves of amplitude A give 2A, so four times the intensity at a bright fringe, zero at a dark one, twice the single-beam intensity on average, so energy is conserved). After the crossing each beam goes on unchanged; small pond ripples do the same. Why it often looks as if beams "do not combine": - the fringes are fine. Two beams of wavelength lambda crossing at angle theta give fringe spacing lambda / (2 sin(theta/2)). Worked example [calc]: helium-neon light, lambda = 633 nm, theta = 1 degree: 0.633 um / (2 x 0.008727) = 36 um, too fine for the eye on a wall; at 30 degrees it is 1.2 um. - independent sources drift in phase, so the pattern moves too fast to see (Magyar-Mandel caught it with short pulses; Marett saw it as beats at a photodetector). - orthogonal polarizations give no fringes at all (Fresnel-Arago). This is a property of the light (transverse vector waves), not of space. A pond's surface waves are effectively one-component, so they always interfere; light has two independent polarization components. So the observation does not point to a special property of space or to a preferred direction. In this respect space and a pond behave alike: both are linear (at ordinary amplitudes) superposing media for their waves.

2. Wiener and Ives's thin films. Wiener (1890) showed that an incident and a reflected beam superpose into a standing wave with nodes and loops half a wavelength apart, and that the photographic action follows the electric field (node at the mirror). Ives and Fry (1933) made it quantitative with a photoelectric probe, and Ives's thin-film work showed photoemission proportional to the standing-wave intensity inside a film. Both are direct, quantitative proof that light waves in space add exactly as superposed waves in a medium do, including the "sum of both" in phase and cancellation out of phase. Ives used it against photons (Rumford lecture); quantum theory predicts the same detection pattern (arbiter, claim 7a), so it does not decide that question, but it does settle the pond question: light superposes.

3. Preferred direction. Michelson-Morley (1887) and its resonator successors bound any direction dependence of light's two-way speed to about 1 part in 10^17 (2009) and 10^18 (2015), although Earth moves at ~368 km/s relative to the CMB. Hughes-Drever experiments show matter itself (nuclear energy levels) has no direction dependence to far smaller levels. Note what these tests can and cannot do: they rule out a pond-like medium whose waves travel at a fixed speed relative to the medium as seen by any observer (a Galilean ether), but they cannot separate Ives's Lorentz ether from special relativity, because Ives's ether predicts the same null (claims 2-3; "completely equivalent in substance", Robertson). There is a preferred frame in the contents of the universe (the CMB rest frame); no experiment so far finds one in the laws of light or matter.

4. What a wave-only observer can infer about a medium. The pond reveals itself even to someone who sees only the ripples: - dispersion: ripple speed depends on wavelength (this is exactly what spoiled Ives and Stilwell's mercury ripple-tank invariance: 0.852 measured against 0.658 ideal); - loss: ripples die out with distance; - frame dependence: a ripple-watcher who moves through the water measures different speeds upstream and downstream, unless (as in Ives's ripple tank) his rulers and clocks are themselves ripple patterns; - nonlinearity: large waves steepen, break and scatter each other; - a rest frame you can feel: drifting with the water changes things. For the vacuum: no dispersion found (Fermi-LAT, beyond the Planck scale for a linear energy dependence; one earlier MAGIC hint superseded); no loss (light from galaxies billions of years old arrives with no vacuum absorption [memory: high]); no frame dependence to 10^-17 to 10^-18; but there is a tiny nonlinearity (light-by-light scattering, ATLAS 2016, CMS), and the vacuum has fixed response constants (epsilon0, mu0, impedance mu0 c ≈ 376.7 ohm [calc from epsilon0 mu0 c^2 = 1]). So a wave-only observer would infer a medium with properties, and would find that this "medium" has none of the properties that single out a rest frame.

5. A medium with properties but no rest frame. The record supports this description, with competing readings that make the same predictions: - Einstein 1920 (Leiden): space "is endowed with physical qualities; in this sense ... there exists an ether", but "the idea of motion may not be applied to it". - Maxwell's vacuum: epsilon0 and mu0 set c = 1/sqrt(epsilon0 mu0) and the impedance of free space; the displacement current epsilon0 dE/dt is what lets an electromagnetic wave sustain itself in empty space with no moving charge [memory: high]. The constants are the same in every inertial frame, which is why they define no rest frame. - Quantum vacuum: light-by-light scattering (observed), and the Casimir effect and vacuum polarization [memory: high; not web-checked in this run], are properties of a vacuum state that is the same for all inertial observers in the standard theory [memory: high]. - Ives's Lorentz ether (and Ruderfer's reading of it): a real medium with a rest frame that is in principle undetectable by rods, clocks and light. It predicts the same results; it differs from Einstein 1920 in asserting a rest frame that no measurement reveals (Ives: the ether's "elusiveness by any measurements of the sort we have been considering", 1937, claims.md claim 2). - The question's "perhaps a preferred direction": not found at the precision of current experiments; it remains a live research field (Standard-Model Extension searches) because many quantum-gravity ideas allow tiny violations [web: Modern_searches page, framing].

The intuition in the question that "if we could only see the waves, a pond might look like empty space" is close to Lorentz's and Ives's own picture: Ives and Stilwell's ripple-tank creatures, built of ripples, could not detect their motion through the mercury. The difference the experiments show is that a real pond gives itself away through dispersion, loss and a rest frame you can drift with; the vacuum, so far, gives itself away only through its constants and its tiny quantum nonlinearity, none of which picks out a frame.

Open

• Whether Lorentz invariance fails below 10^-18 (light) or the current clock-comparison bounds; the Standard-Model Extension data tables (Kostelecky and Russell, "Data Tables for Lorentz and CPT Violation", arXiv:0801.0287, updated yearly) [memory: high] are the record to watch.
• Direct laboratory detection of vacuum birefringence in a strong magnetic field (PVLAS reported limits, not a detection, as of its last results [memory: medium]); evidence from neutron-star polarization (2016) is reported but debated [memory: medium].
• Wiener's oblique-incidence variant (45 degrees, polarization in the plane of incidence, no fringes) is cited from memory [memory: medium]; no web source was found in this run. Wiener's 1890 paper (Ann. Phys. 276:203) would settle it.
• Chen (arXiv 2024) disputes that only the electric field matters in optical interference; snippet only, not read.

Sources

• Ives, "Adventures with Standing Light Waves", Proc. Am. Acad. Arts Sci. 81:1-32 (1951), pp.11-12, 30 [read: _text]
• Ives and Fry, "Standing Light Waves; Repetition of an Experiment by Wiener, Using a Photoelectric Probe Surface", JOSA 23:73 (1933) [read: _text]
• Ives and Stilwell, JOSA 31:14 (1941), 32:25 (1942), via primer 4.9 and arbiter 7d [read]
• https://skullsinthestars.com/2008/05/04/classic-science-paper-otto-wieners-experiment-1890/
• http://www.conspiracyoflight.com/LaserInterference/LaserInterference.html (Marett 2010)
• Magyar and Mandel, Nature 198:255 (1963); Pfleegor and Mandel (1967) [web: ADS snippet; Phys. Rev. 159:1084 from memory: high]
• https://en.wikipedia.org/wiki/Fresnel–Arago_laws ; Rath, Kanseri et al., Am. J. Phys. (2008) [web: snippets]
• https://en.wikipedia.org/wiki/Modern_searches_for_Lorentz_violation (Herrmann 2009; Eisele 2009; Fermi-LAT 2009; MAGIC 2007)
• https://en.wikipedia.org/wiki/Michelson–Morley_experiment (Nagel 2015)
• https://en.wikipedia.org/wiki/Hughes–Drever_experiment
• https://en.wikipedia.org/wiki/Two-photon_physics (ATLAS 2016, CMS)
• https://en.wikipedia.org/wiki/Vacuum_permittivity
• https://en.wikisource.org/wiki/Ether_and_the_Theory_of_Relativity (Einstein 1920)