Abstract
This thesis develops a conceptual framework - based on the work of Erving Goffman - to understand the practical exercise of police suspicions. In so doing, it generates provisional insights about how the police decide to stop and search members of the public, and assesses the role of broader influences on officer decision-making. The thesis draws primarily on extensive observations of police patrols, and face-to-face interviews with operational officers. These data were gathered as part of a mixed-methods study carried out for the Home Office on the reforms to stop and search introduced following the Stephen Lawrence Inquiry. It is argued that police suspicions are generated through two related processes, which are also to be found in everyday social interactions. On the one hand, social information about a person's identity is 'signalled' to other people, which results in them adjusting their conduct towards that person. Understanding this communicative process in the context of policing helps to identify the physical, categorical, and behavioural signals which are ascribed meaning, and prompt suspicions. On the other hand, the police are also seen to use interpretative schema - through a tacit cognitive process of 'framing' - to locate social information and define it as suspicious. Importantly, because signals are patterned temporally and geographically, and frames are influenced by the organisational and social context, there is a need to develop a systemic understanding of police decision-making. In examining these processes, 'methodical suspicion' and 'categorical suspicion' are found to be pervasive in police work. The thesis shows that suspicions tend to fall on the socially marginal - particularly young men, those from ethnic minority backgrounds, and people who are 'known' to the police - and concludes that routine police decisions and encounters are central to the production, and reproduction, of social order.