The driven damped harmonic oscillator is taught in every physics course. Two resonance frequencies are standard: the amplitude resonance (where displacement peaks) and the velocity resonance (where velocity peaks, coinciding with the natural frequency). Every textbook derives both. The problem is considered solved — has been since the 19th century.
Note (2026-11-15; read against the paper's full text): The paper gives two readings of energy resonance and the essay presents one. For the time-averaged steady-state energy it derives an exact resonance frequency, approximated by the mean of the amplitude and velocity resonances (the essay's third frequency). But for the instantaneous total energy it argues that the amplitude resonance frequency is itself the energy resonance, because it gives the largest peak values; so the abstract's 'we argue' makes amplitude resonance and energy resonance coincide in that sense. The paper's velocity resonance is also described as the resonance of input power and power loss (at the natural frequency), not of kinetic energy rate. The essay's count of three resonances for three quantities, and its closing question 'complete with respect to which quantity', are the essay's own framing.
Lelas and Poljak (arXiv:2501.04797) point out a gap: the energy resonance — the driving frequency that maximizes the time-averaged steady-state energy — is different from both. It has an exact expression, and that expression is excellently approximated by the arithmetic mean of the amplitude and velocity resonance frequencies.
Three different frequencies maximize three different quantities in the same system. Amplitude resonance is below the natural frequency (shifted down by damping). Velocity resonance equals the natural frequency exactly. Energy resonance sits between them — approximately at their average.
The structural insight is about what “resonance” means. The word implies a single phenomenon — driving at the system's preferred frequency. But in a damped system, different observable quantities prefer different frequencies. The system doesn't have one resonance; it has a family of resonances parameterized by what you're measuring. The amplitude resonance maximizes displacement. The velocity resonance maximizes kinetic energy rate. The energy resonance maximizes total stored energy. Same oscillator, same driving, different answers to “what frequency is best?”
The 19th-century solution is complete. But “complete” left out the question nobody thought to ask: complete with respect to which quantity?