the archival tradition, the fundamental problem was the definition of archives. Given this implicit epistemological position, I propose to explore a complementary hypothesis: the fundamental problem on which the very existence of archives depends is the order of archival behavior. The terms "archival behavior" (A.B.) and "research behavior" (R.B.) do not exist in the literature. By exploring the case of competitive knowledge archives in the custodial era, we identified these two terms as functional concepts that allowed us to understand the researcher's relationship to his or her research and archiving practices by linking them to each other. Based on a historical, archival and epistemological experiment carried out at the Jean Piaget Archives, we present a model that articulates the functioning of scientific research with the functioning of the archive for a researcher. This model, through its formalizations, specifies competing knowledge archives versus applied knowledge archives and attempts to define the set of functional meanings endorsed by the first type of archive. It can then define certain archive boundaries, no longer according to their output, but according to their process of creation. Finally, the model can generate some of the principles that have guided archivists' practice for more than a century and a half, such as the principle of provenance or the principle of respect des fonds.