Swarm intelligence requires an ability to process information at the level of systems, rather than as individual data points alone. This chapter considers an approach to processing swarms of information in which ‘noise’ is identified by situating it within system-level structures that are drawn from narrative. Narrative is an information-structuring process that characterises anomalies by forming and reforming system-level structures around it. It uses a range of perspectives and ontologies as stepping stones, linking them directly and indirectly to identify the behaviour and influence of agents that are unexpected. We considered how this identification process operates in two examples: the fluid flows around a whisker drone and a convoy of troops suddenly attacked by a farmer. In the second example, we explore how Back’s synthetic language technique could be amplified using narrative-like structures to produce emergent synthetic narratives. The goal is to conceptualise an underlying structure for the organisation of information by which swarm intelligence could construct a common operating picture. A desired outcome is for a swarm of agents to be more responsive, be able to react appropriately to on-the-fly situations in the real world and be more collectively intelligent. We consider how this might be made possible by combining approaches that connect information at various levels of a system, by a meta process of connecting the heterogeneous research ontologies in our group.