A corpuscular simulation model for second-order intensity interference phenomena is discussed. It is shown that both the visibility
V=1/2 predicted for two-photon interference experiments with two independent sources and the visibility
V=1 predicted for two-photon interference experiments with a parametric down-conversion source can be explained in terms of a locally causal, modular, adaptive, corpuscular, classical (non-Hamiltonian) dynamical system. Hence, there is no need to invoke quantum theory to explain the so-called nonclassical effects in the interference of signal and idler photons in parametric-down conversion. A revision of the commonly accepted criterion of the nonclassical nature of light is needed.