Can a single particle be at two places at the same time?

Einstein, Podolsky and Rosen's (EPR) seminal paper on quantumcorrelations of entangled states, and Bell's subsequent discovery thatquantum theory is incompatible with any locally realistic hiddenvariable theory, have spurred substantial discussion and manyexperiments on the nature of quantum nonlocality and entanglement. Mostexperimental tests follow Bohm and Aharonov's reformulation of EPR'sgedanken experiment. Hence, the experiments involve two space-likeseparated entangled particles, typically two polarization entangled photons.
In 1990, Tan, Walls, and Collet proposed a test of nonlocality involvingonly one particle, a single photon. Their proposal initiated a debate ifit is indeed possible for a single particle to have measurable (andcorrelated) properties at two space-like separated locations? The twomain argument against this possibility were: 1) that detection of theparticle at one location would immediately nullify any possibility ofrecording any property associated with the particle (except for itsabsence) at any other location, and 2) that detection of the particlewould typically involve the simultaneous detection of particles fromauxiliary states (e.g., a local oscillator), compromising the notion ofsingle particle nonlocality.
We have performed an experiment where the properties of a single photon,split between two spatial modes, are measured. We measure the relativephase between the superposition between zero and one photon at twolocations, by co-propagating a local oscillator in the same spatialmode, but with an orthogonal polarization. The obtained correlationviolates classical physics and clearly support the quantum mechanicalpredictions.