Denormalization as Calculated Trade-off¶

Strategic redundancy can be justified when the cost of inconsistency is lower than the cost of joins, but requires explicit maintenance strategies to prevent data drift.

Tension Addressed¶

Normalization vs. Query Performance

Recognizable As¶

Pre-computed aggregates, duplicated fields for common queries, embedded copies of frequently-accessed related data, caching layers

Canonical Sources¶

Kleppmann, "Designing Data-Intensive Applications" (2017); Date, "An Introduction to Database Systems"

Transfers To¶

Caching strategies (copy vs. reference), API response design (embedded vs. linked resources), report generation (materialized vs. computed)

Denormalization is not the absence of normalization but a conscious decision to violate it. The key discipline is tracking which data is authoritative and which is derived.