Summing up some of the author's thinking over the years, this article posits that many current models in Translation Studies are out-of-date. One reason is that ultimately they are based on the idea that one translator mediates between and identifiable author and identifiable recipients and renders a 'translation' of an identifiable 'source text'. Furthermore, these models often involve sequentiality, obvious chronological and spatial distances, and they tend to be founded on Indo-European language pairs. They often imply that a translator can be omniscient and consequently make a 'perfect' translation; The present article explores these issues in order to show that hey easily lead to mistakes and flawed thinking. It therefore argues that Translation Studies needs to rethink these issues. The article proceeds to present a framework that allows for today's multiplicity of translation situations. Such a framework is needed because Translation Studies is moving into a globalized world in which translations take place between linguistically unrelated cultures. The article proceeds to a discussion of how such a framework will allow us to deal with source texts, translation, originals, and co-existing tranlsaitons in ways that correspond to translation work - in the past, the present and the future.