There's a lot of confusion in this article. The author doesn't seem to be able to distinguish the logical view of data (in which we distinguish between primitiv

There's a lot of confusion in this article. The author doesn't seem to be able to distinguish the logical view of data (in which we distinguish between primitive and derived data) from its physical representation (in which derived data may be physically stored redundantly for efficiency reasons).