Directly observing momentum transfer in twin-slit "which-way" experiments

Abstract

Is the destruction of interference by a which-way measurement due to a random momentum transfer Weierstrass pgreater-than or equivalent toplanck constant over two pi/s, with s the slit separation? The weak-valued probability distribution Pwv(Weierstrass p), which is directly observable, provides a subtle answer. Pwv(Weierstrass p) cannot have support on the interval [-planck constant over two pi/s,planck constant over two pi/s]. Nevertheless, its moments can be identically zero.