(no subject)

[timelets]

I now think the difference b/w radical and incremental innovations lies in their respective domains. Their codomain is the same. In short, radical innovation (e.g. a paradigm shift) is a functor, while incremental innovation (e.g. "normal science") is an endofunctor.
===
One of the implicit assumptions behind Bob Coecke's approach is that functions compose, i.e. domains and codomains match. This is true for natural processes. In artificial processes discovering/creating matching domains/codomains is a separate task.