This paper will examine several prototypical examples of “representations” to shed light on what this common conception actually entails. It will argue that experientially verifiable representations are extremely rare and that their use as metaphors for what language supposedly does is epistemologically misleading and serves at best discourse communities that seek to prevent their members from realizing their involvement in the phenomenon they talk of. To avoid such undifferentiated uses, the paper suggests to replace “representation” by “re-presentation” (making present again), by “presentation” (creating present realities or illustrations in narratives) or by the coordination of understanding in conversations. This is a shift from correspondence theories of truths via coherence theories of truths to viable conversational practices of living.