1970

Relational Database

Edgar F. Codd (1923–2003)

Storing large amounts of data was one of the early uses for computers, but it wasn’t immediately obvious how the data should be organized. At IBM’s San Jose research laboratory, computer scientist Edgar Codd devised an approach for organizing and arranging data that was more efficient than other models. Instead of grouping together data belonging to the same entity, his approach created large tables of data that had the same conceptual types, with identifying numbers (IDs) defining the relationships between records in different the tables.

For example, an insurance company might have one table of customers, with each customer having a CUSTOMER ID and a name. Then there might be another table of insurance policies, with each having a POLICY ID, a CUSTOMER ID, and a POLICY TYPE ID. A third table might link the POLICY TYPE ID and the details of the policy. In this example, to find the insurance policies for a customer, the computer would first find the CUSTOMER ID, and then find all of the policies that had the same CUSTOMER ID. To get the details of each policy, the system would take the POLICY ID, look it up in the table of policies to get the POLICY TYPE ID, and then search the table of policy types to find the details.

Codd’s groundbreaking research showed that organizing data in this fashion made it more efficient to store, faster to access, and easier to program. Most importantly, he showed that it was possible to create a general-purpose database engine for storing data on the computer’s hard drive, freeing programmers from the task and allowing them to concentrate on their applications. Once the database was developed and deployed, improvements to the underlying software benefitted all of the applications that relied on it. For his work, Codd was awarded the 1981 A.M. Turing Award.

Today the operating systems of both Apple’s iPhone and Google’s Android create a relational database on every smartphone for every app that’s installed, making Codd’s invention one of the dominant ways of storing data.

Edgar Codd created large tables of data that had the same conceptual types.