Résumé : The Mach-Zehnder Interferometer (MZI) is chosen to illustrate the long-standing wave particle duality problem. Why is which-way (welcher weg) information incompatible with wave interference? How to explain Wheeler's delayed choice experiment? Most crucially, how can the photon divide at the first beam splitter and yet terminate on either arm with its undiminished energy? The position advanced is that the photon has two identities, one supporting particle features and the other wave features. There is photon kinetic energy that never splits (on half-silvered mirrors) or diffracts (in pinholes or slits). Then there are photon probability waves that do diffract and can reinforce or cancel. Photon kinetic energy is oscillatory; its cycles require/occupy time. E = mc2 suggests that kinetic energy is physically real as occurrence in time just as rest mass is physically real as existence in space; both are quantized and both occupy/require a dimension for their occurrence or existence. Photon kinetic energy (KE) thus resides in time, but is still present/available for interactions (events) in space; rest mass (e.g., your desk) resides in space but is still present/available for interactions (events) in time. While photon probability waves progress in space and diffract there, photon KE resides in time and never diffracts in space; at reception it always arrives whole and imitates particle impact without being a particle. Photon probability waves are real; they diffract in space. Acknowledging that the photon has two identities (residing energy and progressing probability), explains photon dual nature. And wave-particle duality is central to quantum mechanics. Understanding it leads to new insights into entanglement, nonlocality and the measurement problem. Supporting video: https://youtu.be/A1Wabkr0YFE