Can reverse stealth constructions determine whether a theory permits stealth before computing ghost content?

# Reverse Stealth and Ghost Content in Field Theories

Your question uses specialized terminology from theoretical physics—specifically about "stealth" solutions and ghost fields—that isn't directly addressed in these preprints. Let me unpack what's likely at stake, then explain what the archive *does* cover.

What stealth constructions and ghosts are

In modified gravity theories, a stealth solution is a configuration of the extra fields that hides them from observation—typically by making them identical to their vacuum values while still affecting spacetime geometry in subtle ways. A ghost is a degree of freedom with the wrong sign in its kinetic energy, which causes unitarity violation (probabilities cease to sum to one) and instability in quantum field theory. The tension you're asking about is real: can you check whether a theory *permits* stealth solutions *before* you compute whether those solutions introduce ghosts?

What the archive covers

The preprints provided address related but distinct problems:

Ghost elimination and consistency: Preprints [AG-2024.10-448], [AG-2024.08-594], and [AG-2026.04-415] all tackle how to construct or constrain theories to avoid ghosts. Notably, [AG-2026.04-415] shows that in DHOST theories (which include higher derivatives), gauge symmetry is the "fundamental origin" of ghost elimination—suggesting that symmetry properties determine ghost content before detailed Hamiltonian analysis. [AG-2024.06-500] explicitly constructs "a class of ghost-free theories" in teleparallel gravity using symmetry constraints, which shares the spirit of your question.

However, none of these preprints specifically addresses whether *stealth constructions* can serve as a diagnostic for ghost-freedom, or whether you can check for stealth solutions before computing ghosts.

---

The provided preprints don't directly answer whether reverse stealth constructions can determine ghost-freedom prior to explicit ghost-field computation.