I found a very interesting example for the students to model using UML or another diagrammatic, object-oriented language: the simulation of a complete consulting company, part of an assessment centre.
Another: a description of the 50s listing the “duties” of a good wife. This is a class description for “housewife”, which makes us laugh today. The goal could be to first model this class after the original description, and then alter it to make it more real … the model contains attributes and relationships, things the objects of that class know and do.
The students delighted me in the 1st lecture on BPM with a few good, deep questions of the sort “what are – if any – the limits of process modeling?” I am now glad that I stepped on Robert Rosen a few weeks ago who looked at the relationship of machine and nature using mathematical category theory. His results are not undisputed – this might partly be because he is quite a polemic, too. Accuses his fellow biologists of essentially using the machine metaphor in their scientific explorations, which he considers dangerous.
Thinking about the student’s question made me think about the nature of language – I read up on Chomsky and was interested to hear that his Universal Grammar is considered more relevant to building compilers than modern science of language.