What is database normalization and what are the normal forms?
Quick answer
Normalization is organising tables to reduce redundancy and update anomalies: 1NF means atomic values, 2NF removes partial dependencies on part of a composite key, and 3NF removes dependencies between non-key columns.
In 1NF every column holds a single value (no comma-separated lists) and each row is unique. In 2NF the table is in 1NF and every non-key column depends on the whole primary key, which matters for composite keys. In 3NF the table is in 2NF and non-key columns depend only on the key, not on other non-key columns (a city and zip_code pair stored with an order is an example of what to move out).
Normalisation avoids storing the same fact in many places, so updates and deletes cannot leave inconsistent data. The trade-off is more joins. Reporting databases and read-heavy systems often denormalise on purpose, accepting some redundancy for faster reads.
Key points
- 1NF: atomic values. 2NF: no partial dependency. 3NF: no transitive dependency
- Goal: remove insert, update and delete anomalies
- Denormalise deliberately for read performance