Quantum measurement collapse is assumed to be instantaneous. The wavefunction is in a superposition; you measure; it's in an eigenstate. No duration, no process, just a boundary between before and after.
Prlina and Živković propose that collapse could be continuous — a rapid but finite nonunitary process — and that this hypothesis is testable. If measurement timing has inherent uncertainty, individual quantum systems collapse at slightly different moments. Averaging over many such systems produces what looks like standard mixed-state evolution. Strong projective measurements cannot distinguish this from true instantaneous collapse.
But weak measurements can. Weak postselected measurements — where the measurement interaction is intentionally gentle and outcomes are conditioned on a later strong measurement — probe the system without collapsing it fully. The statistical signatures of gradual versus instantaneous collapse differ in the weak measurement regime.
The proposed test uses hydrogen atoms. The temporal texture of collapse — whether it has duration or not — would show up as differences in the post-selected weak values.
The structural finding: strong measurements erase the evidence of their own mechanism. Only a measurement that barely interacts can reveal how interaction works. The gentlest probe sees deepest because it doesn't destroy what it's looking for.