Does decorrelating from mass miss signatures where new physics *is* the mass shift itself?

This is a sharp question that touches on a real methodological tension in beyond–Standard Model searches.

The short answer: Yes, decorrelating from mass can miss signatures where the mass shift itself is the signal — but the preprints here show that physicists are actively designing experiments *around* that blind spot.

The Decorrelation Problem

When experimentalists search for new physics using atomic clocks, precision measurements, or collider data, they often subtract off the known mass of particles to isolate anomalous *shifts*. The logic is sound: if you measure a frequency change, you want to know whether it's "just" a mass change (boring) or something more exotic (interesting). But as you're pointing out, this assumes mass shifts are a *nuisance*, not the signal itself.

Where New Physics *Is* the Mass Shift

Several of these papers directly address exactly this scenario:

Ultralight dark matter and equivalence-principle violation [AG-2025.04-573] explores theories where "new fundamental physics causes apparent variations in particle mass ratios in space and time" — often *not* uniform across particles. This is a mass shift as the signal. The key insight: if an electron–proton mass ratio oscillates over time, that's not "just a mass change"; it's evidence of new light fields coupling differently to different particles. By comparing *ratios* of masses (via atomic clock frequency comparisons), experimenters sidestep the decorrelation problem entirely.

Dimension-8 operators in dark matter models [AG-2024.04-1361] show how shift-symmetric ultralight fields generate observable effects in fundamental constant variations and Lorentz violation — again, mass-like shifts that carry new physics information because they violate symmetries (Lorentz invariance, equivalence principle) rather than simply rescaling everything uniformly.

Why This Matters

The signal-versus-nuisance distinction hinges on *uniformity*. A universal mass shift (everything gets heavier by 1%) is hard to detect and arguably unphysical. But if the electron mass shifts while the proton mass doesn't, or if masses oscillate in time, that's *falsifiable and model-specific*. The new physics is encoded in the *pattern* of shifts, not the overall magnitude.

The practical upshot [AG-2025.04-573]: atomic clock networks can achieve sensitivity to electron–proton mass ratio variations as small as parts in 10¹⁸ — far better than any single mass measurement — precisely because they're built to detect *relative* shifts that decorrelation would otherwise erase.

These papers don't dodge your concern; they reframe it as a feature: the experiments are designed to *preserve* mass-shift signatures where new physics breaks symmetries.