Part 3 of 5 · 1 min read
Logical data independence
Why a conceptual-level change should not force every application to change.
Think of a company reorganizing its departments internally, without changing any employee's job title or the forms they fill out every day.
In plain words: logical data independence is the ability to change the conceptual schema without having to rewrite every application built on top of it. The conceptual-level change might be adding a new column or splitting a table. Only applications that actually depend on the changed part need to react. The external level acts as a buffer. As long as a view still produces the same columns it always did, the application behind it never needs to know the conceptual level underneath changed shape.
A worked example. Suppose customers gains a new column, loyalty_tier, for a feature the billing team never asked for. The billing_customers view still returns only customer_id and address, exactly as before. The billing application keeps working unchanged, even though the conceptual schema grew.
The key lines. Nothing about billing_customers needs to change:
-- customers now has a new column, but the view is untouched
SELECT customer_id, address FROM billing_customers;
Logical data independence
Logical data independence is the property that conceptual-level changes do not require changing applications. This holds for any application that works through an external-level view unaffected by the change. Adding a column or restructuring a table are typical examples of such a change.
Example: adding loyalty_tier to customers needs no change to billing_customers or the billing application, since neither ever referenced that column.