Subjects and lessons to explore.

Law

Databases & SQL · Foundations

Relational Databases

9 min read 5 practice questions
9 key termsSourcesReport an error
Want it in plain words first? Jump to Eli explains — the same idea, no jargon.
On this page 9 sections
  1. In 30 seconds
  2. Why this matters
  3. The college version
  4. Eli explains
  5. Worked example
  6. Key takeaway
  7. Quick check
  8. Study tools
  9. Sources & references

In 30 seconds

A is a database that stores data in tables linked by shared values, the working definition used by IBM and Oracle. Computer scientist Edgar F. Codd proposed the in 1970, and it became the standard way to organize business data. Tables hold data in rows and columns, and related tables connect through common values such as a customer number rather than by copying data. Most business software runs on relational databases, and is the standard language for talking to them. The model is a choice, not magic.

Why this matters

Relational databases run the systems behind everyday life: orders, bookings, accounts, and records. Knowing what one is, and why the table-and-shared-value design became the default, lets you see how a shop's order form connects to its customer file, why businesses avoid storing the same address in fifty places, and what people mean when they say a system runs on a database or uses SQL. It is also the foundation every later topic in this subject builds on: tables, rows, columns, keys, relationships, and SQL all assume this one idea. Learning it now makes the rest of the subject concrete, and it gives you a fair way to judge database claims you will hear for the rest of your career.

The college version

A working definition

A relational database is a database that stores data in tables linked by common values. IBM, which makes some of the most widely used database software, describes a relational database as one that organizes data into rows and columns that form tables, with the data points related to each other. Oracle, another major database company, says a relational database stores and provides access to data points that are related to one another. Put the two descriptions together and the working definition is simple: the data lives in tables, and the tables are connected by values they share. Nothing about this definition requires size or glamour: a corner shop's customer file is stored relationally if it is kept this way, and so is a national bank's account system.

A short history: Codd, 1970

Before the relational model, every application typically stored its data in its own unique structure, so the same information was arranged differently in each program that used it. In 1970, computer scientist Edgar F. Codd published a paper titled 'A Relational Model of Data for Large Shared Data Banks' that proposed a standard way to represent and query data: organize it into tables, originally called relations, and let tables be related through the values they hold in common. The idea took hold over the following decade, and relational systems became the industry's main way of storing structured data. The historical fact matters because it shows the model is a designed answer to a real problem, not a law of computing.

Tables: the core idea

At the center of the relational model sits the : data organized in rows and columns, the same shape as a well-kept spreadsheet. Each holds one entry, one customer or one order, and each holds one kind of fact, such as the customer's city or the order's date. W3Schools, whose SQL tutorial has taught a generation of beginners, puts it plainly: the data in a relational database is stored in objects called tables, and a table is a collection of related entries consisting of columns and rows. What makes tables more than tidy lists is that they are the unit of connection; the model's original term for a table was 'relation,' and it is from that word that relational databases get their name. The mechanics of tables, rows, and columns each have their own topics; here the point is that this grid is the model's building block.

Relations between tables

The key idea of a relational database is how tables connect: through shared values, not by copying whole datasets into one another. Consider a bookstore with two tables. The customers table holds one row per customer, with a customer number, a name, and a city. The orders table holds one row per order, with an order number, the customer number of whoever placed it, the date, and the total. The two tables are linked because both contain the customer number. When the shop needs to know who placed order 501, it matches that order's customer number against the customers table, and the database pulls the two tables together on the common value. Notice what the order row does not contain: the customer's name and address are not copied into every order. The order stores only the number, and the link does the rest.

Why relational won

Three strengths explain why the relational model became the default for business data. Flexibility: because data sits in tables that can be linked as needed, new questions can be answered without rebuilding the data, and Oracle credits tables as an intuitive, efficient, and flexible way to store structured information. Consistency: each fact is stored in one place, so a value changed once is correct everywhere it is referenced, and Oracle singles out data consistency as a hallmark of the relational model. Avoiding duplication: storing each fact once instead of copying it into every record that mentions it reduces redundancy; the disciplined process for deciding which facts belong in which table is called normalization, a topic of its own. Together, these three strengths let one database serve many applications without the same information drifting apart.

SQL, NoSQL, and an honest framing

Relational databases have a standard language: SQL, short for Structured Query Language, which W3Schools describes as a standard language for accessing and manipulating databases. People and programs use SQL to ask questions of the tables, add new rows, and update facts. A tiny taste: to see every customer's name and city, a query might read SELECT name, city FROM customers;. That is the whole language lesson here, because sql-basics owns the syntax. Relational is the default for most business data, but it is not the only model: the main alternative family is non-relational, or , databases that store data in forms other than tables, often chosen for unstructured data such as text, images, and video. Microsoft Learn notes that relational systems are a common way to store and manage transactional and analytical data, which is why most order, account, and record-keeping systems use them. None of this makes the model magic; it is a designed model with strengths and limits, and knowing which model fits which job is part of database literacy.

Eli, the EliExplains learning guide

Eli explains

The same idea, in plain words

Explain it like I’m 10

A relational database is a database that stores its data in tables, like spreadsheets with rows and columns. The clever part is how tables connect: instead of copying a customer's address into every order, an order row stores just the customer's number, and that shared number links the two tables. Edgar F. Codd proposed this design in 1970, and it became the default for business data because it is flexible, keeps facts consistent, and avoids writing the same information in many places. SQL is the standard language used to talk to these databases.

Picture it like this

Think of a school office that keeps a student card box and a loan box. Each card holds one student's name and address; each loan slip holds one book's title and the student's card number. To find who borrowed a book, the office matches card numbers, and a student who moves changes the address in one place, the card, not on every loan slip. A relational database works the same way: the tables are the card boxes, and the shared numbers are what the office matches.

Where the picture stops working

The card box makes tables sound small and human-managed. Real relational databases hold billions of rows, are queried by software rather than clerks, and are guarded by rules about which values may be shared and what happens when data changes. The matching idea is the same; the scale and the machinery are not.

Worked example

Cedar Harbor Books keeps two tables. The customers table holds one row per customer: customer 1001 is Amara Chen of Portland, customer 1002 is Leo Ortiz of Salem. The orders table holds one row per order: order 501 was placed by customer 1001 on March 14 for $42.50. The order row stores only the customer number, not Amara's name or address. When a clerk asks who placed order 501, the system finds customer 1001 in the customers table by matching the shared value. When Amara moves, her address changes once, in the customers table, and every order she has ever placed automatically shows the new address because no order ever copied the old one. That one design choice, linking by shared values instead of copying data, is what makes the relational model's consistency and low duplication possible.

Key takeaway

A relational database stores data in tables and links those tables by shared values, a design proposed by Edgar F. Codd in 1970 that became the default for business data because it is flexible, consistent, and avoids duplication. SQL is the standard language for working with it.

Quick check

3 questions here, of 5 in this lesson’s practice set. Answers stay hidden until you check.

Question 1 of 3foundational

Which statement best defines a relational database?

Choose an answer, then check it.
Question 2 of 3intermediate

Edgar F. Codd is credited with proposing the relational model in a 1970 paper. What was the model's central idea?

Choose an answer, then check it.
Question 3 of 3intermediate

Riverside Bikes keeps a customers table and an orders table. An order row stores the customer's number but not the customer's full address. Why is that the better design?

Choose an answer, then check it.
Practice all 5

Keep learning

Ready to build on this? Continue to the next lesson.

Practice this lesson
Study tools & related lessonsYou’ll learn to · Common mistakes · Easily confused · Key vocabulary · Related

You’ll learn to

  • Define a relational database as a database that stores data in tables linked by common values, using the working descriptions of IBM and Oracle.
  • State the historical origin of the relational model: proposed by Edgar F. Codd in his 1970 paper.
  • Explain how tables hold data in rows and columns and how related tables are linked by shared values rather than by copying whole datasets.
  • Name the three main strengths behind the relational model's dominance: flexibility, consistency, and avoiding duplication.
  • Contrast relational databases with the non-relational (NoSQL) family in one line, and identify SQL as the standard language for relational databases.

Common mistakes

  • Thinking tables are linked by copying data into each other.

    Tables are linked by shared values, such as a customer number. Copying a customer's address into every order creates duplication, and the copies drift apart when the address changes.

  • Believing the relational model has existed as long as computers have.

    Codd proposed the relational model in 1970. Earlier systems stored data in per-application structures, and the table model won acceptance over the following decade.

  • Assuming every database is relational.

    Relational specifically means data in tables linked by shared values. The non-relational (NoSQL) family stores data in other forms, such as documents or key-value pairs.

  • Thinking SQL is a feature of one company's product.

    SQL is a standard language for accessing and manipulating databases, used across relational systems from many vendors. Its syntax is the subject of the sql-basics topic.

  • Treating the relational model as the only correct way to store data.

    It is the default for most business data, but it is a model with strengths and limits. Some unstructured-data workloads fit the non-relational family better.

Easily confused

Relational database vs. Non-relational (NoSQL) database

A relational database stores data in tables linked by shared values; a NoSQL database stores data in other forms, such as documents or key-value pairs, and is often chosen for unstructured data.

Storing a shared value vs. Copying the whole dataset

A shared value such as a customer number links two tables and stays current when the fact changes once; copying whole datasets into each table duplicates facts and lets them drift apart.

Relational model vs. Per-application file structures

The relational model, proposed by Codd in 1970, gives all applications one standard table-based way to represent and query data; earlier per-application structures arranged the same information differently in every program.

Key vocabulary

relational database
A database that stores data in tables, with related tables linked by shared values such as a customer number.
relational model
The way of organizing data, proposed by Edgar F. Codd in 1970, in which data is kept in tables that can be related through common values.
table
A grid of rows and columns holding one kind of related data, such as customers or orders; the core structure of a relational database.
row
A single entry in a table, describing one customer, one order, or one other item.
column
One named kind of fact in a table, such as the city of a customer or the date of an order.
shared value
A piece of data, such as a customer number, that appears in two tables and is what links them together.
SQL
The standard language for storing, retrieving, and changing data in relational databases.
NoSQL
A family of non-relational databases that store data in forms other than tables, used for many unstructured-data workloads.
RDBMS
Software that stores, manages, and retrieves data in relational databases.

Sources & references

  1. What is a relational database? — IBM
  2. What Is a Relational Database? — Oracle
  3. SQL Introduction — W3Schools
  4. Explore fundamental relational data concepts (Microsoft Learn training module) — Microsoft Learn

EliExplains lessons are original prose written from the open, credible references above. See Copyright & Licensing.

Researched 2026-08-21

Educational content only. It is not medical, legal or professional advice. Found an error? Tell us.