Traveling (or propagating) wave are oscillating electrical voltages observed sequentially or simultaneously at different brain locations, relating them functionally. They are observed at various scales, all over the brain, in various frequency bands and following varied stimuli (or in their absence), making them an ubiquitous phenomenon. Ermentrout and Kleinfeld [2001] proposed three possible mechanisms: 1) A unique oscillator activating cortical areas with different time delays (resulting in a spurious wave), 2) A chain of successively activated oscillators, 3) Coupled local oscillators generating wave propagation due to phase delay. In scalp EEG, volume conduction considerably complicates the problem, and allows to consider a fourth “mechanism”, in which only two sources can generate phase delays at many electrodes. We show that phase differences in the α band appear along the anteroposterior axis during the ECR condition, and this pattern is stable across recordings taken years apart. A simplistic model of two sources is enough to model phase differences incompatible with transmission delays. Care must therefore be brought to the apriori misconception that PLI index is not sensitive to volume conduction.