Abstract. The hypothesis of quantum self-interference is not directly observable, but has at least three necessary implications. First, a quantum entity must have no less than two open paths. Second, the size of the interval between any two consecutive quanta must be irrelevant. Third, which-path information must not be available to any observer. All of these predictions have been tested and found to be false. A similar demonstration is provided for the hypothesis of quantum erasure. In contrast, if quanta are treated as real particles, acting as sources of real waves, then all types of interference can be explained with a single causal mechanism, without logical or experimental inconsistencies.Keywords: Self-interference, Quantum Erasure, Photon, Wave-Particle Dualism. PACS: 03.65.Ta, 14.70.Bh, 42.50.Xa. In the early days of quantum mechanics, there were no easy ways to test its interpretations. More often than not, scientists had to rely on thought experiments. In our days, on the other hand, it is hard to come up with a quantum property that is not already testable. Granted, some features of quantum systems are undetectable by definition. However, unobservable causes are only relevant if they produce observable outcomes. Therefore, any hypothesis about quantum properties can be verified by checking its necessary implications against actual experimental data. For example, the concept of self-interference leads to three major predictions. Firstly, a particle must cover more than one path if it is to interfere with itself. This might happen through non-local propagation, or local splittingrecombination. Either way, it must be essential to have two open paths (or more) for every single photon, whenever self-interference is at work. Secondly, the presence or absence of other quanta must be of no consequence. Large numbers of coherent photons are still necessary for the detection of fringes, but the size of the interval between any two consecutive quanta must be irrelevant. They could be light-years away from each other, as far as selfinterference is concerned. Thirdly, it is also known that interference vanishes whenever direct path verification is attempted. This is frequently cited as proof that path distinguishability destroys self-interference. If so, any indication about actual photonic trajectories, even if obtained by non-invasive means, should be incompatible with the observation of fringes in the same experiment. All of these implications have been tested extensively, often with surprising results. An overview of representative findings is presented below.The first prediction of self-interference, regarding the necessity of multiple paths for individual quanta, was tested in a very creative way by Sillitto and Wykes [1]. They used an electric shutter to switch on and off the paths of a Young interferometer. Specifically, they made sure that the two paths were never open at the same time, yet both of them were switched several times before a photon had time to reach the detector. The rate of emission wa...