[en] Describing an assembly of an infinite number of nucleons in interaction via a two-body potential as a non-relativistic many body problem in the first place, we envisage corrections to this picture due to suppressed degress of freedom at the level of the two-body potential. At variance with relativistic many-body theory, the solution of the non-relativistic problem with a two-body potential only is sufficiently under control at present so that evaluating corrections in this framework is of particular interest. These corrections come primarily from additional three-body forces either due to finite density effects (Pauli blocking of fermions) or are of genuine origin: relativistic dynamical processes and effects from the intrinsic structure of the nucleon. Recalling the successful treatment of electromagnetic interactions in nuclei in terms of meson-exchange currents, we establish novel consistency requirements between the initial two-body force and the well identified residual three-body force. We show further that the nucleon-antinucleon pair term required in the analysis of meson-exchange currents has a genuine three-body counterpart resulting from time ordered diagrams containing a single Z-branch. Its contribution to the energy per particle is repulsive and varies with a high power of the density. Thereby we obtain the important saturating effect present in relativistic mean field approaches. We envisage next the role of the first radial nucleon resonance N-star (1/2, 1/2) (Roper resonance) in inducing a specific three-body force. The meson-nucleon-Roper coupling constants and form factors are evaluated in a relativistic quark model. Gathering all self-consistent corrections to the binding energy per particle of infinitely many nucleons, we find that the final equation of state is solely governed by the density dependence of medium corrections to the free σ-meson mass. We discuss a first attempt to extract this density dependence from an empirical equation of state