Distinguishing effective models from microscopic theories

Consider dilute nonrelativistic bound charges driven weakly enough that their restoring force can be linearized. Treat their displacement as a damped harmonic oscillator and neglect magnetic forces, spatial dispersion and collective local-field corrections. Use complex amplitudes with time dependence exp of minus i omega t. The damping term then enters the response denominator with a negative imaginary sign. Polarization is charge density times displacement, yielding the susceptibility that multiplies the macroscopic electric field.

This model connects charge density, charge, mass and restoring frequency to an observable response. Real solids contain multiple resonances, interactions and quantum oscillator strengths; a single classical oscillator is not a derivation of all those details. Its usefulness here is that its internal variable can be eliminated exactly within the linear model. The resulting frequency-dependent constitutive law is more informative than a constant, but both descriptions have assumptions that can be tested.

Set damping to zero for a finite comparison well below resonance. The static susceptibility is the squared plasma frequency divided by the squared restoring frequency. Expanding the exact rational response in frequency gives that constant followed by a quadratic correction. In the time domain this expansion corresponds to retaining successive temporal derivatives in a constitutive relation; keeping only the constant is justified by slow variation relative to the internal oscillation, rather than by an assertion that the charges do not move.

The geometric series supplies an explicit remainder when the frequency ratio has magnitude below one. For the constant approximation alone, its absolute error relative to the exact undamped susceptibility equals the squared frequency ratio. This is an error in susceptibility, not automatically the same relative error in refractive index or a measured optical phase. A different observable requires propagating the response through its measurement formula. Close to resonance the low-frequency hierarchy fails, and damping or other omitted physics becomes essential.

Different charge densities and restoring frequencies can yield the same static ratio. Static data therefore leave an identification ambiguity even within this restricted oscillator family. Frequency-dependent measurements can constrain the next coefficient, while independent microscopic information may constrain density or oscillator strengths. This is a concrete reason that a successful effective law does not uniquely reveal its microscopic constituents. Conversely, lacking unique microscopic identification does not make the low-frequency prediction useless.

A responsible research claim states which level was derived and which parameters were measured. One may derive a constitutive approximation from an assumed oscillator equation, empirically fit its coefficients, and propose a new microscopic origin for those coefficients. These are three different achievements. A proposal becomes more discriminating when it predicts a response outside the matching data while remaining inside its controlled domain. The practice comparison constructs such a frequency measurement without pretending that either toy material is a complete quantum theory.