The description of lithic assemblages focuses carefully on describing how functional aims are organized. These aims are often considered as a system of binary opposites : weapon/ tool, domestic/ hunting. Gradually, the idea that the partition between the production of weapons and collective tools reflects a form of social partition took root (Bon, 2009). We call this hypothesis double segmentation. It entails a direct correspondence between these two modes of organization. This partition cannot be demonstrated for the transformation or the use of tools, but it can be considered for the production of blanks. We will thus test this hypothesis here. It is based on the pretext that the Upper Palaeolithic involves “ a clearer individuation of the social actors” in comparison with the Middle Palaeolithic (Valentin and Bon, 2012, p. 180). These authors draw on the idea that objects participate in stabilizing intersubjective relations, following the work of S. Strum and B. Latour (1987/ 2006). Several integration methods exist for assessing Upper Palaeolithic functional aims. A methodological framework was thus established to incorporate these different methods, based on the theoretical organization of the products in the operational chain and the operational sequence. This first grid shows that neither the partition of production activities or their imbrication allows us to directly link both parts of the hypothesis. A second grid was thus established. It lists the operational configurations established for the Chatelperronian and the Laborian in the southwest of France to define how these two records of activities are linked. This is the first overall synthesis of the organization of production at the scale of the Upper Palaeolithic in France. It brings to light a wide diversity of solutions and shows that the separation between weapons and tools is not the dominant configuration at this level of analysis. However, we cannot infer that there is no social partition of activities during this long period. As we cannot link both sides of the hypothesis, this appears to imply that it is false. It is thus appropriate to explore the current paradigms leading to this proposal. Several pitfalls have been observed. The first concerns the binary system of categorization : weapons versus common tools, which is a limiting factor (Tartar et al., 2006). In certain cases, this does not necessarily involve the production of two different categories of blanks, but of an assemblage that is differentiated after production (Perlès, 1991), which then becomes “ a weapon by destination”. It is thus the use of the blank that defines its role rather than the pre-existence of distinct categories. This appears to be true for several categories of objects, such as the Chatelperron blank (Pelegrin,