Infrastructures for transport, telecommunication, or provision of energy or drinking water all have in common that they are complex networks of social and technological components. The functioning of such networks depends to a large extent on the behavior of the social components.
Infrastructures for transport, telecommunication, or provision of energy or drinking water all have in common that they are complex networks of social and technological components. The functioning of such networks depends to a large extent on the behavior of the social components. This explains why so many scientists are studying how people – as individuals, as groups, as organizations – use, manage, and develop infrastructures. As they seek answers to very different questions, these scientists often look upon people and their (collective) behavior in very different ways. The models that economists use to predict investment choices by energy companies are very different from the models that psychologists use to explain why people are afraid of hydrogen as car fuel. Are these models in essence based on the same assumptions? Can they be used in combination?
These and similar questions were central to one of the smaller projects within the Next Generation Infrastructures research program: Harnessing Multi-Actor System Complexity. By means of literature research, conceptual analysis of scientific articles, and interviews and workshops with scientists from various disciplines, TU Delft researchers Pieter Bots and PJ Beers tried to elicit and categorize the ways in which human behavior is conceptualized and represented across disciplines. Their ultimate goal: to facilitate interdisciplinary research. For such categorization would allow researchers in the field of infrastructures to more rapidly find “common ground” while working on different, though related, problems. Thus, developing a so-called “ontology” – a shared terminology that crisply and consistently defines the meaning of concepts – for describing individual and collective human behavior became a key effort within the project.
As it turned out, most researchers take a “rational actor” perspective: they assume that the behavior of people follows from their decisions, and that these decisions follow from a deliberation that is “rational” in the sense that it follows logically from what people want (their goals) and what they know (the available information). However, this common perspective still leaves researchers tremendous freedom in the way they describe the deliberation process. Economists tend to aggregate the decisions of thousands of individuals into elegant mathematical equations that can then predict quantitative variables such as prices. Political scientists tend to frame decision making behavior as strategic games that stakeholders play. This means that they must not only describe the goals and information of people, but also the “rules of the game”. To complicate things further, people learn from their interactions, meaning that they may change their views. And for such learning behavior, scientists have – again – thought up different theories. Even then, there may be good reasons for going beyond the “rational actor” perspective. Psychological factors, for example: a person may lack the courage to perform a rational action. Moreover, behavior often does not follow from decisions, but from habits and routines that are embedded in the social-cultural context.
In addition to eliciting concepts by perusing scientific articles, Bots and Beers also experimented with processes in which researchers themselves collaboratively develop concept definitions. They found that such joint efforts are very difficult to organize, for apparently quite rational reasons: they often ask of individual scientists to diverge from the terminology of their own discipline, whereas the pay-off of developing a shared language occurs only when working as a group to solve a real problem. It seems that joint ontology development will succeed only for problems interesting enough to bind many scientists together over many years.