The spatial nature of thought
users have to learn how to model when modelers should learn how users think
He gave several examples of where the usual process modeling notations (BPMN, EPC, etc.) leave no room to embed aspects that are important to people in a certain domain.
One of the aspects was space, where he pointed to a research paper on `the spatial nature of thought’. In the experiment reported in this paper, students were asked to model an information system scenario involving a person with a GPS device, which is used to find a bookstore in the area, where the person then performs a transaction. In almost all the sketches the satellite was put “up in the sky”, and components related to each other were positioned in close proximity.
Now, this whole scenario may seem less of a process. After all, we have been taught to separate processes and compositional structures (dynamic vs. static), which is certainly useful from an engineering perspective. But for people to make sense out of process models, space is often important in processes too. Like Michael said in his talk: people in a hospital process modeling project were asking: “But how can i see on which floor this is?”
Coming from a reverse-engineering, a mining perspective, the integration of other aspects (such as space, but also time, or anything that matters) into an automatically created process model is just as relevant. Because if “Designers convert internal representations of their ideas to external representations by drawing on a page”, then the reverse is also true, and automatic creations of an external representation need to resemble the internal representations of the people who look at these models.
Keeping this imperative in mind, there are huge opportunities to enrich models by all kinds of aspects that matter to people, so that they truly create meaning and value.