Bell-type experiments confront us with a threefold demand: any satisfactory account must preserve Bell-inequality violation, operational no-signalling, and consistency with relativistic causal structure. Standard discussions often treat this situation as if it required a physically meaningful transit time for a collapse-like influence between spacelike separated measurement events. In this paper, I argue that this presupposition is false. Working within Sequential Time Theory (STT), I formulate admissibility conditions for temporal variables in Bell configurations and prove that no state-independent and foliation-independent calibration map exists that could define a physical collapse-transit time between spacelike separated local record-completion events. The quantity required to define a collapse speed is therefore not extremely small; it is non-constructible. To establish this result, I reconstruct a Bell experiment as a partially ordered record structure consisting of source events, local record events at each wing, and later comparison events. Local measurements are represented by compatible instruments on admissible hypersurface descriptions. I then show that Bell-inequality violation and operational no-signalling remain fully intact without any admissible collapse-transit variable. What is eliminated is not the empirical Bell phenomenon, but only the pseudo-variable that would redescribe a correlation structure as though it were a directed transport process between the two wings. The resulting claim is narrower than a complete relativistic theory of quantum measurement, but stronger than a merely interpretive remark. The paper yields a theorem-level boundary condition on future measurement theory: any successful extension may model local detector dynamics, irreversible amplification, delayed recording, or moving-frame configurations, but it must not reintroduce a state-independent and foliation-independent collapse-transit time between spacelike separated local record-completion events. In STT terms, Bell nonlocality does not force superluminal propagation; it forces the abandonment of an inadmissible temporal variable.
Teruhito Kojima (Wed,) studied this question.