Context. We present a reanalysis of the distribution of galaxies in the $\log(\langle I\rangle_e)-\log(R_e)$ plane under a new theoretical perspective. Aims. Using the data of the WINGS database and those of the Illustris simulation we will demonstrate that the origin of the observed distribution in this parameter space can be understood only by accepting a new interpretation of the $\log(L)$-$\log(\sigma)$ relation Methods. We simulate the distribution of galaxies in the $\log(\langle I\rangle_e)-\log(R_e)$ plane starting from the new $L=L_0'\sigma^\beta$ relation proposed by D'Onofrio et al. (2020) and we discuss the physical mechanisms that are hidden in this empirical law. Results. The artificial distribution obtained assuming that beta spans either positive and negative values and that $L'_0$ changes with $\beta$, is perfectly superposed to the observational data, once it is postulated that the Zone of Exclusion (ZoE) is the limit of virialized and quenched objects. Conclusions. We have demonstrated that the distribution of galaxies in the $\log(\langle I\rangle_e)-\log(R_e)$ plane is not linked to the peculiar light profiles of the galaxies of different luminosity, but originate from the mass assembly history of galaxies, made of merging, star formation events, star evolution and quenching of the stellar population.