Is SQL exactly relational algebra?

No. SQL is based on relational algebra and calculus but differs in key ways: it uses multiset (bag) semantics so duplicates can appear, adds grouping and aggregation, supports nulls with three-valued logic, and includes procedural and data-definition features the pure algebra lacks.

Why learn relational algebra if SQL is what is used in practice?

Relational algebra is the formal model behind query optimization: database systems translate SQL into algebra-like expressions and apply equivalence rules to find efficient plans. Understanding the algebra explains why certain query rewrites are valid and helps reason about query equivalence and cost.

SQL and Relational Algebra

Relational algebra is a procedural language of operators over relations that gives the relational model its formal semantics, while SQL is the standard declarative language that database systems compile into algebra-like execution plans.

Definition

The relational algebra is a set of operators that take relations as input and produce relations as output; SQL (Structured Query Language) is the ISO-standard declarative language for defining, querying, and updating relational databases, whose query semantics correspond closely to relational algebra extended with grouping, aggregation, and bag semantics.

Scope

This topic covers the operators of the relational algebra — selection, projection, Cartesian product, join, union, difference, and renaming — and their use as a formal basis for queries and query optimization. It covers the core of SQL: SELECT-FROM-WHERE queries, joins, grouping and aggregation, subqueries, set operations, and data-definition and data-manipulation statements, together with SQL's multiset semantics and three-valued logic for nulls. It excludes how queries are physically executed and optimized (covered in query processing and optimization).

Core questions

What are the primitive and derived operators of the relational algebra?
How does a SQL SELECT-FROM-WHERE query map onto relational-algebra operations?
How do grouping and aggregation extend the basic algebra?
How do nulls and three-valued logic affect query results?
Why does SQL use multiset (bag) rather than set semantics?

Key concepts

selection, projection, join
union, difference, Cartesian product
SELECT-FROM-WHERE
grouping and aggregation
subqueries and correlation
set and multiset operations
three-valued logic and nulls
data definition and manipulation language

Key theories

Relational algebra operators: Selection, projection, union, set difference, Cartesian product, and renaming form a complete set from which join, intersection, and division are derived; algebraic identities among them justify query-rewriting optimizations.
Declarative querying and relational completeness: SQL lets users state what data they want rather than how to fetch it; because it can express every relational-algebra query (and more), it is relationally complete, and the system is free to choose an efficient execution strategy.
Three-valued logic for nulls: SQL evaluates conditions over true, false, and unknown to handle null values; this affects comparisons, WHERE filtering, and aggregation, and is a common source of subtle query errors.

Clinical relevance

SQL is one of the most widely used languages in computing: it is the interface to virtually every relational database powering business applications, analytics, and web services, and relational algebra is the theoretical tool query optimizers use to rewrite and plan those SQL queries efficiently.

History

Codd's 1970 paper introduced the relational algebra and calculus. At IBM's System R project Chamberlin and Boyce designed SEQUEL (1974), later renamed SQL, as a more accessible declarative language. SQL was standardized by ANSI and ISO beginning in 1986 and has been extended through successive standards while remaining grounded in the relational algebra.

Key figures

Edgar F. Codd
Donald D. Chamberlin
Raymond F. Boyce

Seminal works

codd1970
chamberlin1974
garciamolina2008

Frequently asked questions

Is SQL exactly relational algebra?: No. SQL is based on relational algebra and calculus but differs in key ways: it uses multiset (bag) semantics so duplicates can appear, adds grouping and aggregation, supports nulls with three-valued logic, and includes procedural and data-definition features the pure algebra lacks.
Why learn relational algebra if SQL is what is used in practice?: Relational algebra is the formal model behind query optimization: database systems translate SQL into algebra-like expressions and apply equivalence rules to find efficient plans. Understanding the algebra explains why certain query rewrites are valid and helps reason about query equivalence and cost.