It has become a truism to state that mathematical knowledge or "meaning" and the norms and practices of (school) mathematical communities are negotiated and socially constructed before individuals construct (internalize) them for themselves. In this study I suggest that the social in existing conceptualizations is based on a trivial notion that fails to capture the social as the mathematics of mathematics. An episode from a second-grade mathematics curriculum is used to articulate a strong version of the social and the transactional order as origin of anything that counts as mathematical.