I would like to hear a SQL philosopher/historian explain how this situation came about. It seems such a violation of language orthogonality - lying in wait to create bad data.