In the database world, the raw physical data model is at the center of the universe, and queries freely assume intimate details of the data representation (indexes, statistics, metadata). This closed-world assumption and the resulting lack of abstraction have the pleasant effect of allowing the data to outlive the application. On the other hand, this makes it hard to evolve the underlying model independently from the queries over the model.

As the move to the cloud puts pressure on the closed-world assumption of the database, exposing naked data and relying on declarative magic becomes a liability rather than an asset. In the cloud, the roles are reversed, and objects should hide their private data representation, exposing it only via well-defined behavioral interfaces...

The world of database models is noun-based, talking about Customers, Orders, LineItems, etc. Once modelers have designed the data model correctly, they consider their job done.

In the realm of modelers, there is no notion of data abstraction that separates abstract properties of the model from the concrete details of the fully normalized realization in terms of tables with PK/FK (primary-key/foreign-key) relationships. --All Your Database Are Belong to Us, Erik Meijer