One of the most interesting thoughts from Andrew Sabl's book Hume's Politics is that what elevates an institutional rule to constitutional status is NOT that it was established in some golden-age founding moment, but rather that it was reasserted successfully many times over history, because it had proved useful.
That what makes something a part of the constitution is recurrent reassertion inverts the usual relationship between constitutional principle and founding moment. Hume, in studying the history of England, finds that Magna Charta and other constitutional principles like hereditary monarchy were always feeble at their introduction and violated many, many times. What strengthened them was leaders and people returning to them to solve problems, and that process eventually made them conventional.
@kevinjelliott
Interesting observation. But now I wonder if the rules were used to solve similar problems that came up repeatedly (since the core of human experience is very durable), or if they were repeatedly reinterpreted to solve new problems (because society is ever changing)?
@tobychev Here at least, it seems to be the former. Similar problems arise in government--who should be the ruler after the current one?; what limits are there on what rulers may do?--and conventions arise to answer them.
That said, things do change and so do conventions for answering those Qs. So when power shifts from monarchs to parliaments, the succession question changes because heredity doesn't answer it anymore. Now it becomes how long does a parliament's mandate last.