We study the superposition of a non-Poisson renewal process with the presence of a superimposed Poisson noise. The non-Poisson renewals mark the passage between meta-stable states in system with self-organization. We propose methods to measure the amount of information due to the two in dependent processes in dependently, and we see that a superficial study based on the survival probabilities yield stretched-exponential relaxations. Our method is in fact able to unravel the inverse-power law relaxation of the isolated non-Poisson processes, even when noise is present. We provide examples of this behavior in system of diverse nature, from blinking nano-crystals to weak turbulence. Finally we focus our discussion on events extracted from human electroencephalograms, and we discuss their connection with emerging properties of integrated neural dynamics, i.e. consciousness.