Saw a talk by John C. Knight of the University of Virginia yesterday: The Use of Rigorous Arguments in Engineering. Basically, instead of employing standards to ensure interoperability and reliability of complex systems, he proposes using a (tree-like) structure of justifications that is specific to the project. These two statements remained with me:
An argument without evidence is unfounded.
Evidence without an argument is unexplained.
Although it sounds obvious, we engineers and scientists occasionally forget these principles in the rush to write-up our work.
