In this note we put forward briefly some arguments which convince us that the "microfoundation problem" is ill-posed (at least, as it has been studied traditionally). First of all, that the only valid microfoundation is the walrasian one is no longer a tenable position. Secondly, even in this respect, the most usual hypothesesis accepted by the proponents of a rigorous microfoundation procedure is the representative agent, which is in contrast with their very viewpoint. Thirdly, we know of no analysis in which all macroeconomic propositions are deduced from microeconomic theorems (if this were possible, one cannot see how the distinction would be meaningful). Finally, we accept and motivate the idea that macroeconomics treats economic properties which are "emergent" with respect to the analysis of individual behaviour: but this is a completely different story from the one told in the usual micro-macro debate.