This article tries to offer a different perspective on the issue of what it means for some physical structure to be a representation. In the first sections, it will be shown how and why this issue is still far from settled. This will be done by emphasizing the—what I will call—metaphysically promiscuous character of representation. For although representations are typically assumed to be some sort of physical objects or structures, on closer inspection, the notion of representation is used in such a variety of ways that its fundamental metaphysical status is far from obvious. Proceeding from these observations, it will be argued that, even though “representation” pre-theoretically indeed often picks out objects, their representational status is best not understood in terms of their physical properties or their causal-functional profile. It will be argued that, what it means for some physical structure to be—as a matter of fact—a representation, only first becomes intelligible in relation to certain socio-normative practices in which the cognitive capacity to relate to something as something it is not is prescriptively called upon. Moreover, an answer to the oft-heard question of what makes something (i.e., some physical object or structure) a representation is readily available, provided we take into account certain cognitive abilities, as well as a socio-normative context in which these abilities are normatively regulated. It will be concluded that at the fundamental metaphysical level, the phenomenon of representation is best understood as a triadic relation which involves, but does not reduce to, certain physical objects or structures. Finally, this socio-normative account of representation will be compared with two dominant notions of representation within cognitive science: symbolic representation and S-representation.