Historical research guide · 0617 · 2026
The classical linear oscillator in classical electrodynamics with classical zero-point radiation
Timothy H. Boyer
In one minute
Timothy Boyer has spent fifty years asking how much of quantum behaviour classical physics can recover if you grant it one thing: that empty space is filled with real random radiation, with an average energy of half a quantum in every mode. Here he applies that programme to the simplest system in physics, a charged mass on a spring. He shows the charge settles into a stable ground state — it radiates, but the vacuum feeds it back exactly what it loses — and that above the ground state it has a ladder of resonant excited states. The step between rungs is Planck’s constant times the oscillator’s own frequency, which is Bohr’s rule arriving from a purely classical calculation. The mechanism turns on a detail usually thrown away: the driving field must be evaluated where the charge actually is, not at the origin, which turns the oscillator into a parametric one with resonances at odd multiples of its natural frequency. Boyer’s closing point is that the field doing all this stays perfectly hidden in the energy books.
Connection
Chapter 2 says the vacuum is a real medium; Boyer’s work is the strongest demonstration of how much of atomic behaviour that assumption alone buys you, and chapter 3 rests on exactly this mechanism — the ground state held in balance against the zero-point field. Chapter 6 gets the warning as well as the encouragement: in equilibrium the exchange is perfectly balanced and therefore invisible, which is precisely the balance any device drawing power from the field has to break.
From the source
The existence of Casimir forces between uncharged conducting surfaces implies, within classical physics, that random radiation must be present even at the zero of temperature — classical electromagnetic zero-point radiation, with average energy per normal mode equal to half Planck’s constant times the frequency, and the same average action variable of half Planck’s constant at every frequency in every inertial frame. Over more than fifty years this classical theory has reproduced Casimir forces, van der Waals forces, oscillator specific heats, diamagnetism, superfluid behaviour, the absence of atomic collapse and the blackbody spectrum, all of which are usually said to require quantum theory.Sec. II C, Classical electrodynamics with classical zero-point radiation
Because classical zero-point radiation is Lorentz invariant, any charged mechanical system that comes to equilibrium in it must be at least approximately Lorentz invariant — and very few potentials meet that criterion. A nonlinear oscillator fails, because its nonlinear term carries a length that combines with the natural frequency to define a velocity other than the speed of light. The one-dimensional linear harmonic oscillator and the Coulomb potential are the important survivors, and even the linear oscillator qualifies only in the large-mass, small-amplitude limit where the oscillation speed is far below the speed of light.Sec. I C, theme 1; Sec. II D
In the large-mass approximation the charged oscillator has a stable ground state in classical zero-point radiation, and the stability survives when higher radiation multipoles are included. Requiring equilibrium for both the dipole and the quadrupole radiation fields fixes the equilibrium spectrum uniquely, up to an overall constant, as the Lorentz-invariant zero-point spectrum — no other random spectrum will do. In that ground state the oscillator scatters the radiation with no time-average net radiation propagating in any direction, so the interaction is completely disguised.Sec. I C, theme 4; Sec. II E; Sec. XI B
The excited states come from a detail the dipole approximation discards. The driving electric field must be evaluated at the charge’s actual position rather than at the origin, and that position dependence turns the equation of motion into a parametric oscillator with an entirely new set of resonances at integer multiples of the natural frequency. For the resonant excited states the dipole radiation the charge emits at its own frequency is balanced by energy taken from zero-point radiation at odd multiples of that frequency, with the oscillator’s action variable equal to the odd multiple of the radiation action variable.Sec. I C, theme 2; Sec. II F; Sec. VIII A 2 and A 3
The Bohr frequency condition appears naturally in this classical context. When the resonance index changes by one unit, the change in the average energy of the charged mechanical oscillator is exactly Planck’s constant times the natural oscillation frequency, and the transition radiates at that same frequency — the frequency of the oscillator in empty space. The change in the oscillator’s energy equals the change in the energy of the driving radiation, so net radiation appears only when the index changes.Sec. IX, Eq. (59)
What the analysis leaves open is stated plainly: the resonant excited states are unstable and decay, the treatment is nonrelativistic and holds only in the large-mass, small-amplitude limit, and the classical description differs from the quantum one in its fluctuations even where the average values agree — the classical energy of a normal mode is a stochastic process, the quantum energy an eigenvalue with no dispersion. The natural extension, flagged through Cole’s work on subharmonic resonances, is the Coulomb potential — the hydrogen atom.Sec. VIII A 1; Sec. X; Sec. XI C; Sec. XII
The way in
https://doi.org/10.1088/2399-6528/ae9fb3How to cite it
Timothy H. Boyer (2026) The classical linear oscillator in classical electrodynamics with classical zero-point radiation doi:10.1088/2399-6528/ae9fb3