Sociologists may be tempted to think that largely covert intra-personal processes, such as the manipulation of the meaning of stimuli, while necessary to ensure a sufficiency of agency, fall outside the scope of a discipline focused on ‘social action’. But Thomas and Thomas’ theorem, to the effect that, ‘If men (sic) define situations as real, they are real in their consequences’, suggests otherwise, as too does Weber’s definition of action as ‘behaviour to which the … individual attaches a subjective meaning’. It follows that covert action must necessarily be of interest to the sociologist, although the parameters of the unit act still need to be identified. The difficulties of defining action in terms of intention can be overcome by focusing on attention, such that what an actor is ‘doing’ is simply that which they are ‘attending to’. This not only helps distinguish action from mere habit whilst also clearly distinguishing the unit act, but also has methodological advantages. It follows that what someone is ‘attending to’ is also necessarily ‘meaningful’, signifying that action is not just a matter of ‘meaning’ but also of emotion, effort, and imagination. It is for this reason that understanding action requires a dynamic behavioural analysis.