Speed of Light

The speed of light in vacuum, written c, is the most heavily loaded constant in modern physics. It fixes the ratio between mass and energy, sets the scale of the electromagnetic field equations, defines the metre, and — in the form of Einstein's second postulate, that c has the same value in every inertial frame regardless of the motion of the source — supplies the foundation on which special relativity is built. Since the 1983 redefinition of the SI metre, c is no longer measured at all: it is defined as exactly 299,792,458 metres per second, and the metre is derived from it.

The literature collected on this wiki treats almost every part of that account as open. The researchers documented here do not agree with one another — some hold that c is constant but relative to a medium or a field rather than to every observer; some that the two-way speed is constant while the one-way speed is not; some that c varies with gravitational potential, with the medium, or with time; some that the constancy observed in experiments is an artefact of the instruments used to measure it; and a few that the light-speed barrier can be exceeded outright. What they share is the conviction that the second postulate was elevated to a principle before the experimental evidence could bear it, and that the 1983 redefinition has since made the question difficult even to ask, because a discrepant measurement now registers as a wrongly calibrated metre rather than as a physical result.

This article states the orthodox position, then organises the objections found in the archive by the idea at stake rather than by author.

The mainstream position

Three claims are usually run together under the heading "constancy of the speed of light", and the papers here often turn on separating them.

The first is that the speed of light in vacuum is independent of the motion of the source. This is the second postulate of Einstein's 1905 paper, and it is what distinguishes an electromagnetic-wave account from a ballistic or emission theory of the Walter Ritz type.

The second, stronger claim is that c is invariant — the same not only for all sources but for all inertial observers, whatever their state of motion. This is what forces the Lorentz transformations, the relativity of simultaneity, time dilation and length contraction.

The third is that c is a universal constant in the further sense of not varying with position, with gravitational potential, with frequency, or with time. General relativity already qualifies this: in Einstein's own later treatment the coordinate velocity of light depends on the gravitational potential, which is how light deflection near the Sun is obtained.

Almost all of the experimental evidence usually cited is evidence for the two-way (out-and-back) speed. Measuring a genuine one-way speed requires two spatially separated clocks, and synchronising those clocks already presupposes an answer about the propagation of the synchronising signal. The circularity is acknowledged in the mainstream literature under the name conventionality of simultaneity; where the researchers on this wiki differ is in refusing to treat it as a harmless convention.

The second postulate and the motion of the source

The earliest line of objection in the archive attacks the evidential basis of the second postulate directly.

Ian McCausland notes in "Binary Stars and the Velocity of Light" (1980) that the binary-star argument — the classical demonstration, due to Willem de Sitter, that light from the approaching and receding components of a spectroscopic binary arrives at the same speed — has been questioned by J. G. Fox and by Parry Moon and Domina Eberle Spencer. Moon and Spencer showed that a particular Riemannian metric reproduces the binary-star observations without the second postulate; McCausland proposes an improved metric which he argues makes their case more convincing.

Joe Alexander Nahhas presses the same observations in the opposite direction in "First Experimental Proof of 'Not' Constant Velocity of Light" (1983), arguing that light-aberration measurements from binary systems are dependent on the spin orientation of the component stars — a dependence which, on his reading, records genuine addition and subtraction of the stars' velocities to that of the light, and so contradicts the constancy postulate.

The most sustained programme in this section of the archive is that of Domina Eberle Spencer and Uma Y. Shama, who over a decade set out and defended the universal time postulate first proposed by Moon and Spencer in 1956 and generalised by Moon, Spencer and Moon in 1990. In "Stellar Aberration and the Postulates on the Velocity of Light" (1996) they argue from James Bradley's 1728 aberration data that Einstein's postulate predicts stellar aberration correctly in coordinate systems in which the star is stationary but fails to predict it at all in the earthbound system in which it is actually observed. "Visualizing the Postulates of the Velocity of Light" (1998) sets the three competing postulates — Einstein's of 1905 and 1907, Ritz's of 1908, and Moon and Spencer's of 1956 — side by side in a single geometric construction, and argues that only the universal time postulate agrees with all experiments analysed, and that only it permits a simple method of synchronising the clocks of two moving observers. "The Interpretation of the Velocity of Light" (1999) gives the physical picture: light travels outward from its source in a Huygens-style spherical wave whose radius grows at c and whose centre remains at the source, reducing to Newton's corpuscular hypothesis for a source in uniform motion and departing from it when the source accelerates. "Comparative Analysis of the Doppler Shift Based on Two Postulates on the Velocity of Light" (2004, with Philip J. Mann) derives the Doppler shift on both Einstein's and the universal time postulate, and "Developments on the Postulate on the Velocity of Light in the Twentieth Century" (2005) reviews the whole sequence from Galileo's first attempt to measure light's speed onward.

Emission theory recurs elsewhere. Nina B. Sotina and Nadia Lvov revisit Ritz in "The Ritz Ballistic Theory & Adjusting the Speed of Light to c near the Earth and Other Celestial Bodies" (2011), arguing that if c is the speed at the output of the source and classical velocity addition holds, the Michelson–Morley result, stellar aberration and a number of related results all fall into place, the astronomical binary-star objection being the single obstacle. Zifeng Li argues in "Particle nature of light and the speed of light" (2010) for a corpuscular account in which reflection, refraction and transmission are absorption and re-emission events, and in which the speed of starlight relative to the Earth approaches the emitted speed only near the Earth.

Clarence L Dulaney returns to Einstein's founding paper itself in "Why Special Relativity?" (2005), asking what motivated it, and arguing that Einstein's treatment of the magnet-and-conductor "asymmetries" was vague and that the equation offered in proof of source-independence was incorrect.

The one-way speed of light

The distinction between the two-way and the one-way speed is the single most productive theme in this archive, and it is where the wiki's contributors have done their most careful work.

Franco Selleri gave the programme its most developed theoretical form. In "Noninvariant One-Way Velocity of Light" (1996) he constructs a set of spacetime transformations between inertial systems — his "inertial transformations", distinct from the Lorentz transformations — from three assumptions: that the two-way velocity of light is c in all inertial systems and all directions; that time dilation occurs with the usual relativistic factor; and that clocks are synchronised "in the way chosen by nature itself", as they are in the Sagnac effect. The resulting theory reproduces the available experimental evidence while the one-way velocity is not invariant. In "Noninvariant One-Way Velocity of Light and Particle Collisions" (1996) he defines energy and momentum consistently with those transformations and shows that they coincide formally with the usual relativistic expressions only in the privileged frame but numerically in all frames — so that the precise data on thresholds for inelastic processes and on particle masses are equally well explained within the new framework. "Noninvariant One-Way Speed of Light and Locally Equivalent Reference Frames" (1997) supplies what he treats as the decisive argument: on a uniformly rotating platform of radius R the velocity of light relative to the rim is necessarily different from c, and it stays different as R is increased with the peripheral velocity held constant — so that in the limit, where any small piece of the rim is arbitrarily well approximated by an inertial frame, the anisotropy survives into the inertial case.

Several contributors have addressed the measurement problem experimentally. Dale Means proposed "A Device to Measure the One-Way Velocity of Light" (1992), noting that while two-way constancy is firmly established the evidence indicates the two transit times may not be equal, and emphasising that his device is portable and so can test the measuring process itself in different environments. Chalmers W Sherwin's "Measurement of the One-Way Speed of Light" (2002) uses the Phipps protocol to establish the absolute phase of a remote clock, and argues that a one-way measurement is possible using only automatic-recording instruments independent of human observers — the analysis resting not on relativity but on a single experimentally supported hypothesis about the rate of a moving clock. John E Carroll's "Measuring a One Way Light Speed" (2008) proposes a method using standard frequency generators, laser pulse generators and oscilloscopes, with pulses sent both from A to B and from B to A. Thomas Geoffrey Franzel's "The Logic of a Newly Designed Optical Experiment May Resolve the One-Way Light Speed Issue" (2012) describes, for a general scientific readership, two configurations of a proposed design. Emil D. Gigov's "Measurement of the One-Way Speed of Light" (2010) proposes an interferometer able to test whether the speed of light in empty space varies during the Doppler effect.

The difficulty of the measurement is itself a subject here. Rodrigo de Abreu and Vasco Guerra argue in "Comment on A One-way Speed of Light Experiment" (2009) that an experiment published in the American Journal of Physics as a one-way measurement is in fact determining the two-way speed; and in "The conceptualization of time and the constancy of the speed of light" (2008) they argue that the null result of Michelson–Morley follows without mathematics from the assumption that all good clocks measure the same time regardless of their construction, so that the "postulate" of two-way constancy is not the independent physical claim it appears to be.

A closely related question is whether the world's timekeeping infrastructure secretly settles the matter. Romano Manaresi argues in "International Atomic Time and the One-Way Speed of Light" (1999), and again with Selleri in "The International Atomic Time and the Velocity of Light" (2004), that it does not: the fact that the synchronisation signals of the International Atomic Time network always arrive "on time", at any hour and in any season, has been read by some authors as showing isotropic one-way propagation even at the Earth's surface, but the proper working of the network in fact imposes no condition on the one-way speed at all.

Ruyong Wang designed a family of crucial experiments around the same point. "From the triangle Sagnac experiment to a practical, crucial experiment of the constancy of the speed of light using atomic clocks on moving objects" (1998) establishes that between two points moving in circular motion the A-to-B and B-to-A travel times differ by the Sagnac interval 2V_DL/c², independently of the radius. "Test of the one-way speed of light and the first-order experiment of Special Relativity using phase-conjugate interferometers" (2003, with Yi Zheng and Aiping Yao) uses a phase-conjugate mirror to reverse the uniform phase shift in a light path, making a first-order test of the isotropy of the one-way speed possible.

GPS, atomic clocks and the Sagnac effect

Because the Global Positioning System must reconcile signals from satellites, from a rotating Earth and from moving receivers, several researchers here treat it as the working laboratory in which the competing accounts of light propagation are already being tested every day.

Ruyong Wang and Ronald R Hatch argue in "Conducting a Crucial Experiment of the Constancy of the Speed of Light Using GPS" (2002) that GPS shows the speed of light to remain c relative to the Earth-Centred Inertial (ECI) non-rotating frame but not relative to an observer or receiver moving within that frame: when a receiver changes its translation speed relative to the ECI frame, the light speed measured relative to the receiver changes.

John R Warfield draws the same conclusion into a positive postulate in "The Speed of Light, the 'Tic Rate' of Atomic Clocks, and the Earth Centered Inertial Frame" (2007): that the Earth-Centred Non-Rotating Inertial Frame is the preferred frame both for the speed of light and for the rate of atomic clocks, and that these two processes are not directly related to each other but only indirectly, through that common frame. His "A Hypothetical Experimental Device that uses the Phenomena of Light Aberration to Demonstrate that the Speed of Light is Not Necessarily Relative to the Observer" (2009) describes a rotating-disc apparatus with a retro-reflective peripheral mirror designed to exhibit aberration under controlled conditions.

Stephan J G Gift has argued from GPS data that the anisotropy is directly measurable. In "Light Speed Invariance is a Remarkable Illusion" (2007) he shows first how an illusion of invariance arises in two-way measurement within a semi-classical absolute-space theory, and then demonstrates what he takes to be a measurable variation in the one-way speed, which he holds invalidates the postulate and confirms a preferred frame. "Doppler Shift Reveals Light Speed Variation BPES" (2010) makes the same case from the Doppler shift, arguing that light-speed variation relative to a moving observer follows classical velocity composition. The paper "Faster West than East: The GPS Invalidates Special Relativity" (2013) held in this archive determines the one-way speed on the rotating Earth by two independent routes — the GPS clock-synchronisation algorithm, and light-speed isotropy in the ECI frame — and concludes that light travels faster west than east relative to the Earth's surface, a result it presents as contradicting the constancy principle while remaining consistent with the Selleri transformations.

Howard C Hayden made the connection to the Sagnac effect early. "On a Recent Mininterpretation of Sagnac's Experiment" (1991) replies to a paper by Dieks and Nienhuis which held that Sagnac's 1913 experiment cannot be explained classically and requires special relativity: Hayden argues that this is both physically and historically erroneous, and that the experiment shows, as do contemporary satellite measurements, that the speed of light is not constant in the sense Einstein used in 1905. He adds that the general-relativistic explanation of Sagnac is refuted by the persistence of a time difference when no area is enclosed. In "Rotating Mossbauer Experiments and the Speed of Light" (1992) he reanalyses Champeney's 1963 rotating Mössbauer experiment, arguing that the supposed exact cancellation between the first-order aether-velocity term and the first-order Lorentz time-dilation term does not survive the experimental fact that clock rates are set by velocity with respect to non-rotating geocentric coordinates.

Charles M Hill approaches timekeeping from the sky rather than the laboratory. "Timekeeping and the Speed of Light - New Insights from Pulsar Observations" (1995) observes that the pulse rates of some millisecond pulsars rival the best atomic clocks while being immune to the solar-system dynamics that impose cyclic variations on Earth-based atomic time; measuring in "pulsar seconds" rather than uncorrected atomic seconds yields, he argues, two distinct measures of the speed of light, each with a physical interpretation, and the comparison indicates that Einstein's definition of time and his principle of relativity are useful but not universal truths.

Light speed and the gravitational potential

A distinct family of papers, associated above all with Petr Beckmann's model and the journal Galilean Electrodynamics, keeps a constant speed of light but changes what it is constant with respect to: not every observer, but the local gravitational field.

Howard C Hayden's "Light Speed as a Function of Gravitational Potential" (1990) takes up Beckmann's claim that light speed is constant with respect to the gravitational field and supplies what Beckmann's model lacked — a specification of how the speed varies with the mass of the locally dominant body and the distance from the point of interest. Hayden derives the dependence from conservation of energy and uses it to predict the deflection of starlight, reporting agreement both with general relativity and, more importantly to him, with measurement.

Charles M Hill examines the same proposal critically in "Does the Local Gravitational Field Govern the Speed of Light?" (1991). He accepts that light propagates at a speed constant with respect to something and takes the identity of that something to be the open question; the gravitational-field model, as Hayden noted, predicts that light velocity combines vectorially with field velocity for an observer moving through the field, and Hill sets out to test that prediction.

Ronald R Hatch sharpens the logic of these derivations in "The Speed of Light, Conservation Laws, and Gravity Probe B" (1996). Two different values for the gravitational dependence of light speed had been derived in Galilean Electrodynamics from conservation of energy; Hatch shows that conservation of energy by itself only fixes a relationship between the gravitational dependence of light speed and the gravitational dependence of mass, so that anyone deriving one has tacitly assumed the other. He uses the Shapiro radar time delay to fix the dependence independently.

Henry P Dart argues in "The Search for Fundamental Units of Measurement" (1987) that the observed variability of time standards with gravitational field strength requires all "universal" standards to vary with field strength, and derives new units of length, mass and time from c, Planck's constant and the gravitational proportionality factor; he questions the validity of general relativity on this basis. Wolfgang Engelhardt's "A Remark on the Constancy of the Velocity of Light" (2001) presses the resulting tension: Einstein postulated constancy in 1905, yet in general relativity conceded that light speed may depend on the gravitational potential, which is how deflection is obtained — while the modern unit system has meanwhile made a discrepant measurement impossible in principle.

Reginald T Cahill applies measured light-speed anisotropy to an outstanding observational puzzle in "Resolving Spacecraft Earth-Flyby Anomalies with Measured Light Speed Anisotropy" (2008). The unexplained speed anomalies found in the Galileo, NEAR, Cassini, Rosetta and MESSENGER Earth flybys are, he argues, not real: they arise from using a relation between observed Doppler shift and spacecraft speed that assumes light speed is isotropic in all frames. Taking the repeatedly measured anisotropy into account resolves them ab initio.

Tom Van Flandern extends the argument past light itself. "On the 'Speed of Gravity'" (1993) argues that the standard answer to the question — that gravity propagates at or near c — is incorrect, and rests on conflating the force of gravity with the hypothetical gravity waves; and "A Complete Relativistic Gravity Model with No Speed-of-Light Limit" (2000) argues that gravity propagates much faster than light, which he takes as the first experimental discrimination between special relativity and Lorentzian relativity in favour of the latter, and notes that GPS already makes full use of Lorentzian relativity's universal simultaneity. Robert L. Kemp's "Conceptual and Critical Analysis of Bill Stubb's Paper - An Assessment of the Gravity Data during the March 9, 1997 Total Solar Eclipse" (2013) reviews an attempt to extract a speed of gravity from eclipse gravimetry.

Aether, medium and anisotropy

For a further group of researchers, the resolution is that light has a medium and that c is its property — see Aether for the wider literature.

David Tombe develops this most explicitly. "The Speed of Light" (2013) presents the aether as a fluid-like substance flowing between positive and negative particles, space being densely packed with electron–positron dipoles in mutual circular orbit whose rotation axes align solenoidally; the speed of light follows from the properties of this sea. In "The Speed of Light varies with Magnetic Flux Density" (2007) he applies Archimedes' principle within Maxwell's sea of molecular vortices and argues that a magnetic field intensity gradient in the steady state is compatible with the double-helix theory of the magnetic field only if the magnetic permeability — conventionally held constant for a given medium — is the variable quantity, from which it follows that the speed of light varies with magnetic flux density.

Karim Amen Khaidarov's "On Electromagnetic Waves Velocity" (2007) gives an aethereal account of electromagnetic-wave propagation and describes the author's own experiment showing variability of wave velocity in the aether. Curtis E Renshaw reviews the whole experimental tradition in "Light Speed and Aether" (2000), analysing the Fizeau-era experiments of 1851–1925 which defined the Fresnel drag coefficient and the Lorentz transformations, and on which Einstein built while discarding the medium itself. Nina B. Sotina asks the question directly in "Speed of Light in 3-Dimensional Euclidean Space" (2013): if one builds an alternative model in ordinary three-dimensional Euclidean space rather than four-dimensional pseudo-Euclidean spacetime, with respect to what frame does light move at c? DeWayne Birkhofer's "A Theoretical Model of the Structure of Space Based on the Speed of Light" (2010) works in the opposite direction, taking the measured light speed as the datum from which to construct the three-dimensional structure of the carrier.

Bob Johnson connects the medium question to the Electric Universe programme in "Evidence for the Anisotropy of the Speed of Light" (2011), arguing that the Beckmann model is consistent both with a charged Earth and with the non-null results of the Michelson–Morley-type experiments, and that those non-null results are better read as an electromagnetically induced directional anisotropy of light speed than as an aether wind.

Roberto A Monti's "Three Major Errors in Relativity and Cosmology" (2000) attacks the standard textbook history head on, contending that the Michelson–Morley–Miller experiments never gave a null result, that the 1919 eclipse never proved general relativity, and that the difference between the Newtonian and Einsteinian light deflection was never confirmed after 1919. Victor Nikolayevich Cochetkov argues the reverse case from within classical mechanics in "Explanation of the Results of the Michelson Experiments Using Classical Mechanics" (2012): applying conservation of momentum and energy to the interferometer yields the conclusion that the path difference between the separated beams is independent of the speed and direction of motion of the luminiferous medium — so that Michelson's results confirm classical mechanics rather than conflicting with it.

Ralph Sansbury takes the most radical position in this group. In "The Speed of Light: Cumulative Instantaneous Forces at a Distance" (2012) he proposes that light is not propagated at all in the usual sense, but is the effect of instantaneous forces at a distance: oscillations of charged particles in a source produce in-phase oscillations first inside atomic nuclei of the receiver and then, after a delay, of its electrons, so that what is measured as a propagation time is a cumulative response delay.

Variable c: in time, with frequency, with the medium

Alan Montgomery, with Lambert Dolphin, argues in "Is the Velocity of Light Constant in Time?" (1993) from a statistical analysis of historical determinations drawn from four sources that the measured value of c has decreased over the past 250 years, and that the probability of a systematic or experimental artefact is low; they add that constants involving atomic phenomena and units of time also appear to be changing, while a third set with no obvious dependence on c does not. In "A Determination and Analysis of Appropriate Values of the Speed of Light to Test the Setterfield Hypothesis" (1994) he tabulates and edits the same data specifically to discriminate between constancy and a decrease of the size claimed by Setterfield and Norman, obtaining a time-dependent weighted regression model with a statistically significant trend which survives analysis by subinterval, distribution, accuracy and precision, and reporting that attempts to identify an experimenter bias were unsuccessful.

Sergey N Arteha's "On Frequency-Dependent Light Speed" (2004) analyses the physical corollaries of a possible dependence of c on frequency and proposes a mechanism for one. Yi-Fang Chang argues in "Contradiction Between the Uncertainty Principle and the Constancy of Light Speed" (2001) that since quantum mechanics implies uncertainty relations for velocity, and light speed is a velocity, c must exhibit statistical fluctuations within small spacetime regions and at high energy — contradicting the constancy assumed in relativity, and indicating to him that relativity and quantum theory can only be unified after both are further developed. Yochanan Fein makes a compatible point in "The Speed of Light, a Fundamental Universal Constant, is an Uncertain Quantity in the Sub-classical Range of Measurements" (1997): the constant is established to high accuracy over classical distances but becomes increasingly uncertain over decreasing scales, precisely where it plays its most fundamental role.

Morton F Spears's "A Reexamination of the Velocity of Light, Dark Mass, and the Accelerated Expansion and Age of the Universe" (2004) derives the velocity of light from the permittivity and permeability of space and argues that, once that dependence is taken into account, the universe is significantly older than generally accepted. Raymond H. Gallucci combines two proposals of Calkins and Renshaw in "The Speed of Light: Constant and Non-Constant" (2013) into an account in which light travels at a constant speed within a given medium when emitted from a stationary source but at a varying speed within that same medium when the source moves; The archive holds two further papers pursuing the cosmological consequences of a variable c: "Can Photon-Particle Interactions be Explained with Varying Light Speed?" (2015), which re-examines Compton scattering and tired-light accounts of the cosmological redshift, and "Gravitational Cosmic Redshift with Variable Light Speed", which asks whether a gravitational redshift survives if the "stretching" of light waves near large masses is not constrained by a constant speed — see Redshift. Paul Karl Hoiland argues more modestly in "Does the Speed of Light Have to be Constant?" (2004), from a problem in Einstein's original closed static model, that the general thrust of special relativity survives without a strictly constant c.

Peter Kohut's "The Nature and Speed of Light" (2010) identifies the speed of light with the escape speed of cosmic expansion. John E Chappell, the founder of the Natural Philosophy Alliance, argued in "Ives-Stillwell, Variable Light Velocity, and Variable Electric Charge, in Terms of a Postulated Theory of Radiation" (1998) that the Ives–Stilwell result of 1938, confirmed only for line-of-flight and perpendicular directions, generalises by elementary trigonometry to all directions, and that in his analysis the gamma factor modifies c but not distance, time or mass.

The illusion of constancy

A separate and distinctive line holds that light speed really is c ± v relative to a moving observer, and that the constant value obtained in every experiment is an artefact of the instruments — because the rods and clocks used to make the measurement are themselves altered by the motion, in a way derivable from Newtonian physics and mass–energy conservation rather than from spacetime geometry.

Paul Marmet states the problem in its sharpest form and then answers it. "The Apparent Constant Velocity of Light" (2000) begins from the question that he holds has never been properly answered: with respect to what does light travel, and how can an observer receding at v from a source receive photons at c rather than at c − v? The standard answer invokes spacetime distortion, which Marmet rejects as incompatible with conventional logic. His own explanation, given also in "Explaining the Illusion of the Constant Velocity of Light" (2000), is that the velocity really is c − v, while the observer's tools always return the number c; the same account, he argues, explains why the two-way measurement gives c while the Sagnac effect exhibits c − v and c + v. He then applies it to the operational case in "GPS and the Illusion of Constant Light Speed" (2003) and "The GPS and the Constant Velocity of Light" (2006), where he argues that GPS measurements give c − v or c + v with v the rotational velocity of the Earth at the cities concerned, and that mass–energy conservation alone requires clocks to run slower in a moving frame — no space contraction or time dilation of the relativistic kind being needed.

Stephan J G Gift's treatment of two-way measurement in "Light Speed Invariance is a Remarkable Illusion" (2007) belongs to the same family, as does Vivian Pope's more philosophical "From Light in Space to Space in Light, the Complete Relativistic Revolution" (2004), which reinterprets the spacetime constant c along lines Pope takes to be Machian, in order to place Einstein's theory in what he regards as its proper philosophical perspective.

c as a defined constant: units and measurement

Because c now defines the metre, several researchers here argue that the question of its constancy has been quietly converted from a physical question into a metrological one.

Wolfgang Engelhardt puts it most pointedly in "A Remark on the Constancy of the Velocity of Light" (2001): the normal metre has in effect been replaced by the caesium second and nine fixed digits, so that a physicist who measured a value differing from the nine legal numbers would simply be held to have used an illegal system of units. Georg Galeczki makes the units the central charge in "Special Relativity's Heel Of Achilles: the Units of Measurements" (2004), arguing that the issue of measurement units was persistently overlooked, that this produced contradictory interpretations of special relativity, and that the contradictions are grounds for rejecting the theory as irrelevant to physics.

Louis Essen — the National Physical Laboratory metrologist who built the first practical caesium clock and later became a well-known critic of relativity — reviewed the measurement problem in "The Velocity of Light" (1952), stressing the importance of the constant for electrical theory and practice, treating the electrical, radio and optical determinations as measurements of one and the same constant, and noting that discrepancies among the results would bear on the theory itself.

Frank H Makinson proposes in "Mathematically Defined Speed of Light" (2005) a definition of the numeric value of c from a physical and a mathematical constant by way of a trigonometric function exploiting electromagnetic relationships, independent of the metre and the second but relatable to them; he extends the geometric treatment of wavelength and frequency in "Extended Geometry for Electrical Engineers" (2009). Leslee A Kulba examines the constants behind the standard expression in "C=(e0m0)^-0.5" (1996), addressing the derivation, arbitrary character and physical status of the permittivity and permeability of free space and their coincidental interconnection with each other and with other constants. James P Siepmann's "The Light Clock: A New Method of Measuring True Time" (1999) proposes to use the objective speed of light over a preset distance as the standard of a non-relative time, on the ground that current interval measurements record a periodic occurrence rather than time itself.

Faster than light

A smaller group holds that the light-speed limit is not merely conventional but violable.

Harold W Milnes's "Faster Than Light?" (1983) reports a theoretical derivation, following directly from Maxwell's equations, together with a simple experiment in which electrical signals greatly exceeded the speed of light, the special conditions involving extremely thin conductors of very low capacitance and inductance. The subsequent priority dispute is recorded in his "The Pappas-Obolensky Affair" (1989). Theodore D Mitsopoulos's "Faster Than Light" (1989) appeared in the Toth-Maatian Review, the quarterly of which Milnes was then editor-in-chief and which was founded as a venue for substantive criticism of leading work across the sciences.

Zifeng Li's "Faster-than-Light Speed and Faster-than-Light Speed Effect Observed" (2011) introduces concepts of light speed and faster-than-light speed and the observational consequences he draws from them. Stephen Deratz argues in "It Is Possible for Objects to Travel Faster than the Speed of Light: Refraction in Moving Mirror Supplies the Evidence" that the limit derives from the velocity-addition formula of the Lorentz transformations, which in turn rests on the FitzGerald–Lorentz contraction introduced to explain a Michelson–Morley null result he regards as erroneous. Phil Bouchard's "Finite Theory of the Universe, Dark Matter Disproof and Faster-Than-Light Speed" (2012) builds a classical-mechanical model in which time dilation is proportional to kinetic energy and to superposed layers of gravitational potential, and which he reports predicts GPS gravitational time dilation, perihelion precession for all planets and gravitational light bending, without dark matter.

Joseph A Rybczyk treats the barrier as the first topic of an expanded relativistic physics in "Breaking the Light Speed Barrier" (2007) and develops the associated kinematics in "The Light Speed Effect" (2009), which correlates light aberration with source-independence and rederives the transverse Doppler effect. Eric Baird's "The Light Speed Barrier: Bending the Rules" (1998) argues that special relativity conceals a mechanism for reducing time-dilation effects between mutually receding objects by relaying signals through a chain of intermediate probes, and that this reveals velocity-dependent curvature within inertial systems. The archive also holds "Explanation of How the Speed of Light May be Exceeded without Violating the Requirement of Infinite Energy" (2016), which appeals to the uncertainty principle applying to a body Lorentz-contracted to very small size.

Criticisms from researchers on this wiki

Beyond the specific physical models above, a substantial group of papers here attacks the postulate itself — on grounds of internal logic, of conservation laws, or of the interpretation of the founding experiments.

Victor Nikolayevich Cochetkov mounts the most systematic conservation-law critique. "Dynamic Paradox of the Special Theory of Relativity" (2010) argues that alongside the familiar paradoxes there is a further dynamic paradox: the conservation of momentum, angular momentum and energy of a closed mechanical system is not compelled in inertial reference systems. "Special Relativity: Depending on the Definition of the Momentum of a Closed System of Bodies from Time" (2010) shows by a concrete example that applying special relativity to a closed mechanical system can make the system's momentum a function of time. "The Special Theory of Relativity and the Law of Conservation of Momentum" (2010) turns the argument around, using conservation of momentum to fix the constants in the two possible coordinate-and-time transformations between inertial systems.

Hartwig Wolfgang Thim's "Einstein's Light Speed Postulate is Illogical" (2010) argues that the isotropy postulate from which the Lorentz transformations are derived is logically defective: when a short pulse is emitted at the instant the origins of a stationary frame K and a moving frame K′ coincide, one spherical wavefront is launched, not two.

Two further papers held in the archive press the point mathematically. "The actual reason behind all the issues with Relativity" (2016) argues by elementary algebra that independent measurements of light speed in different reference frames would contradict the light-speed postulate if the frames could be connected by any velocity-dependent transformation, and identifies this as the source of the theory's paradoxes. "Relativity Is Self-Defeated (1 of 3) — In Terms of Mathematics" (2016) argues that relativity's own derivation forces c = 0, destroying the Lorentz factor whose denominator contains it.

Zifeng Li argues in "Special Relativity Arising from a Misunderstanding of Experimental Results on the Constant Speed of Light" (2008) that what the experiments actually show is that the speed of light relative to its source is constant in vacuum, and that Einstein's generalisation of this to a fixed velocity in any "stationary" system was an interpretive step, not an experimental result. Robert J Heaston examines what would follow if the generalisation were dropped in "The Consequences of Assuming that the Speed of Light is not Constant" (2008), noting that c functions as proportionality constant, necessary component or limiting condition across special and general relativity, mass–energy equivalence, the fine structure constant, the Rydberg number, the Boltzmann constant, the uncertainty principle, the Maxwell equations, the Compton wavelength and the Planck scale — the mortar, in his phrase, of the whole twentieth-century paradigm. His "Einstein's Great Oversight" (1991) argues that the convenience convention c = G = 1 in the field equations concealed that the combination c⁴/G present in those equations is a superforce of about 1.2 × 10⁴⁴ newtons.

Daniel Lee Haulman's "Relative Light Speed" (2010) argues that quantum theory and relativity contradict one another and that if one must be discarded it should be relativity, because it rests on the premise that light speed is not relative. Stanley V. Byers's "Light Speed versus Special Relativity" (2009), released by its author into the public domain, reviews light-speed measurements exhibiting variation relative to the motion of planets, satellites, the solar system and the galaxy. E. S. Fortune's "Relativity of the Velocity of Light" (2012) notes that cosmic inflation and quantum entanglement have both been interpreted as involving velocities in excess of c.

Not every paper here is hostile. Pharis E Williams's "Thermodynamic Basis for the Constancy of the Speed of Light" (1997) argues the opposite case — that the laws of thermodynamics require Einstein's postulate — and is included so that the strongest form of the defence is available alongside the objections.

Related papers in the archive extend the question in other directions: Cynthia Kolb Whitney's "Maxwell's Maximum" (2009) shows that a finite-energy pulse pair governed by Maxwell's coupled field equations always spreads longitudinally during propagation, so that "light in flight" is never the simple signal that special relativity and quantum mechanics model; Richard Oldani's "Magnetostatics at Speed c" (2000) argues that several experiments on the nature of light cannot be explained by either classical or quantum theory; William L Hughes's "A Simple Mathematical Model Which Suggests an Electromagnetic Basis for Inertial Mass" (1995) derives relativistic changes of inertial mass from a charge spinning at the speed of light; Philipp M Kanarev's "The Gravitational Radius of a Black Hole" (2002) rederives the Schwarzschild radius taking the photon's wavelength into account; and Felix F. Gorbatsevich's "Inertia and Gravitation" (2011) treats inertia, the aether and c together. Phil Bouchard has proposed two direct experimental tests of invariance, "Proposal for Bidirectional Light Speed Meter in Motion to Test the Invariance of c" (2013) and "Proposal for Wavelength Meter in Motion to Test the Invariance of Light Speed" (2016), both replacing a static aether with overlapping graviton fields that follow the Earth's rotation. Harry H. Mark's "Speed of Light in Historical Perspective" (2011) sets the whole dispute against its three pivotal measurements — Ole Rømer's observations of Jupiter's satellite, James Bradley's determination from Gamma Draconis, and Airy's water-filled telescope.

From the Autodynamics tradition, David Scott de Hilster's "Carezani Frame Reduction" (2008) presents Ricardo Carezani's finding that one of the frames in the derivation of the Lorentz system in relative motion is mathematically and physically redundant, and that removing it yields equations which do not diverge at c.