Approaches to pragmatic reasoning in artificial intelligence traditionally have been based upon alternating sequences of instantaneous states and operators. This formulation has encountered difficulty in our attempts to represent activity in real-world task environments. In this paper, we offer an alternative framework based upon notions of circumstance and process, each of which holds or is in progress over an interval of time. Events mark the beginning and ending of each interval; instants represent constraints on event times. We illustrate our approach through definition of a process corresponding to a simple PUSH operator. We briefly discuss aspects of the resultant outlook on planning and pragmatic reasoning.