2011 was a pivotal year. It was the year when the originators of the first formal normative frameworks of Responsible (Research) and Innovation (RRI), lost contact with their own history. Our paper aims to recover this history. It does so in two ways. First it develops the concept of de-facto responsible research and innovation, referring to the ongoing creative and political process through which new projects encapsulating particular normative visions and prescribed actions of ‘responsibility’ encounter the existing de-facto institutionalised performation of it, both ‘bottom-up’ and historically pre-existing. History, and the a-priori existence of a wide variety of context-shaped responsibility concepts and projects are central to this process. Second, it creates a typology of de-facto responsible innovation ‘narratives’ comprising for each narrative, a differentiating idea, normative philosophy and discourse; a distinctive socio-technical agencement of actors and material devices; and the policies, programmes and initiatives designed by political actors as the pragmatic organising architecture to promote and support each. Our aim here is to bring some analytical order to the multiplicity of histories, visions, content and situations of research and innovation in which actor collectives participate, contest and steer responsibility claims, processes and outcomes. So-doing, we propose six ‘Grand Narratives’ of de-facto responsible research and innovation. The six narratives are presented as discrete ‘ideal types’, recognising that in their empirical form they inter-mingle and overlap. Indeed it is this dialectical unfolding process of inter-mingling and differentiation which fascinates us. We wonder whether the outcome will be the creation of an integrated meta-narrative; whether a dominant policy design will emerge; whether we will witness the maintenance of a highly variegated landscape as actors experimentally pursue local normative preferences and problem framings of responsibility; or whether a limited number of organising variants (here we propose six) will remain internally coherent but normatively differentiated, and theoretically and practically separated each from the others