In some ways calculus and Newtonian physics are reductive theories: they explain phenomena in terms of aggregates of (infinitesimal) units. On the other hand, even if one were to believe that Schwarzschild metrics somehow correspond to point-sources in General Relativity, the non-linearity of the Einstein's equations makes it hard to see that a concept of a point-source could lend itself to superposition. In this talk I present some surprising results regarding ``macroscopic” relativistic objects seen as aggregates of idealized relativistic point-sources.