The author discusses the existing theoretical approaches of semiotically informed research in HCI, what is useful and the limitations. He proposes a radical rethink to this approach through a re-evaluation of important semiotic concepts and applied semiotic methods. Using a semiotic model of interaction he explores this concept through several studies that help to develop his argument. He concludes that this semiotics of interaction is more appropriate than other versions because it focuses on the characteristics of interactive media as they are experienced and the way in which users make sense of them rather than thinking about interface design or usability issues.