(contract, duty of care, possession and ownership), quasi-normative notions (damage, statements, fault, interest and expectations) and legal concepts (rights and duties) is the key to analysing factual problems. An exam question in the University of Cambridge’s Contract and Tort Paper II once invited candidates to discuss, in relation to an action for damages, the following statement: ‘Before the law can be applied to the facts, the facts must be categorised, but before the facts can be categorised, the law must be applied.’ Modern developments in (constructivist) epistemology (theory of scientific knowledge) and cognitive science have given us a major insight into answering this question. Science constructs its own models, which act as both the science and the object of science and thus, when it comes to law and legal science, what is required is the building of a model within the facts which will act at one and the same time as the means of understanding both the facts and the law (see above, p 170; and Petev (1999) 43 APD 27). In saying that all law is about persons, things and actions, the Romans, perhaps unconsciously, recognised this (see above, pp 27–31). At any rate, they certainly provided the tools and the insights for categorising the facts and applying the law.