Analysing and Diagramming Complex Heterogeneous Activities

We increasingly live in heterogeneous ever-changing webs of activities where human actions are intertwined with events created by automatic machines. In order to make such webs understandable to its human participants, their structure should be represented by displays emphasizing their action aspect. The paper suggests thematic roles as a semantics for actions, argues that a selection of well-known diagramming techniques can be defined within this theory, and uses the theory to discuss new issues related to process control and mobile technology. 1 Purpose and overview The topic of this paper is complex activities, i.e. purposeful activities comprising many interrelated actions, cooperating actors, and automatic systems. In Section 2, I present a semantics for analysing such activities and the way they change; Section 3 argues that a selection of already existing diagramming techniques can be interpreted by means of this semantics; and Section 4 describes modelling problems caused by inhomogeneous webs of activities performed by humans and machines, and discusses how they can be treated within the framework. The basic motivating assumption in the following is that technology, ranging from administrative information technology to control technology of complex machinery, is related to actions in two ways: 1 As emphasised in Lind (1990, 1994) and Rasmussen, Pejtersen & Goodstein (1994), complex machinery cannot be understood without taking into account the goals, means and ends of the engineer. Although at some level, machines work according to purely physical forces, these forces have been harnessed by the engineer to perform purposeful actions. In this sense, machines can be said to be crystallised human actions and operations. 2 As emphasised by Goldkuhl & Ågerfalk (2002), Liu (2000), Bækgaard (2001), and many, many others, information systems are media for people to plan, execute, monitor, and evaluate business activities. If this is true, then the displays offered to the operators of the machinery or the users of the information systems ought to represent the workings of the machinery in terms of the actions it crystallizes or mediates. This again makes it relevant to devise diagramming methods that displays actions in an illuminating and correct way. For example, in their recent study of air safety, Revilla & Sanchez-Alarcos 2003 emphasises the importance of knowing how the system works in opposition to understanding what it does – understanding the logical model in opposition to the operational model. In addition to this, a more radical view claims that organizations, standards, and technology are increasingly becoming unstable entities that cannot be designed to last for a prolonged