Your example of Alan Kay isn't a very good one though, because we can objectively test their opinions on say, the best way to implement `malloc` by having them do it and testing the results against the computer. There is no such equivalent for literature.