Edgar F. Codd

Verified Timeline

Lore & Background

Edgar Frank Codd was born in Fortuneswell, on the Isle of Portland in Dorset, England. After attending Poole Grammar School, he studied mathematics and chemistry at Exeter College, Oxford, before serving as a pilot in the RAF Coastal Command during the Second World War, flying Sunderlands. In 1948, he moved to New York to work for IBM as a mathematical programmer, first on the Selective Sequence Electronic (SSEC) project and later on the IBM 701 and 702. In 1953, dismayed by Senator Joseph McCarthy, Codd moved to Ottawa, Ontario, Canada, returning to the US in 1957 to work for IBM. From 1961 to 1965 he pursued his doctorate in computer science at the University of Michigan in Ann Arbor, advised by John Henry Holland, on self-replication in cellular automata. Two years later, he moved to San Jose, California, to work at IBM's San Jose Research Laboratory, where he continued until the 1980s. He was appointed IBM Fellow in 1976.

Reader's Guide

Codd's most significant contribution was the relational model for database management, which he published in his 1970 paper 'A Relational Model of Data for Large Shared Data Banks' after an internal IBM paper one year earlier. Initially, IBM refused to implement the model to preserve revenue from IMS/DB, a hierarchical database. Codd then showed IBM customers the potential of the implementation, and they pressured IBM. IBM included a System R subproject in its Future Systems project but put developers not thoroughly familiar with Codd's ideas in charge, isolating the team from Codd. They created SEQUEL, a non-relational language, which was copied by Larry Ellison in 1979 for Oracle Database. Codd continued to develop the model, sometimes with Christopher J. Date, and the Boyce–Codd normal form is named after him. Codd's theorem equates the expressive power of relational algebra and relational calculus. As the relational model became fashionable, Codd fought to prevent misuse of the term, publishing his 12 rules to define a relational database, which made his position at IBM difficult, leading him to leave and form a consulting company. He also coined the term Online analytical processing (OLAP) and wrote the 'twelve laws of online analytical processing', though controversy erupted over undisclosed sponsorship by Arbor Software.

Did You Know?

• Codd served as a pilot in the RAF Coastal Command during the Second World War, flying Sunderland aircraft.
• His 1965 PhD thesis on self-replication in cellular automata showed that a set of eight states was sufficient for universal computation and construction.
• Codd's design for a self-replicating computer was not actually implemented until 2010.
• He coined the term 'Online analytical processing' (OLAP) and wrote the 'twelve laws of online analytical processing'.
• In 2004, SIGMOD renamed its highest prize the SIGMOD Edgar F. Codd Innovations Award in his honour.

Early Life and the Road to IBM

Edgar Frank Codd was born on 19 August 1923 in Fortuneswell, on the Isle of Portland in Dorset, England. After attending Poole Grammar School, he enrolled at Exeter College, Oxford, to study mathematics and chemistry. His academic trajectory was interrupted by the Second World War, during which he served as a pilot in the RAF Coastal Command, flying Sunderland aircraft. After the war ended, Codd crossed the Atlantic in 1948, settling in New York to work as a mathematical programmer at IBM. He contributed to the Selective Sequence Electronic Computer project and later helped develop the IBM 701 and 702 mainframes. In 1953, unsettled by the political climate surrounding Senator Joseph McCarthy, Codd relocated to Ottawa, Ontario, Canada. His standing at the company grew steadily, culminating in his appointment as IBM Fellow in 1976.

The Relational Model and IBM's Resistance

In 1970, Codd published his landmark paper, "A Relational Model of Data for Large Shared Data Banks," which laid the theoretical groundwork for what would become the dominant paradigm in database technology. The ideas had first circulated internally at IBM a year earlier, but the company proved remarkably reluctant to act on them. IBM's leadership was unwilling to abandon IMS/DB, a hierarchical database product generating significant revenue in the 1970s, and viewed the relational model as a threat to that income. Codd, frustrated by the inaction, demonstrated the model's potential directly to IBM's customers, who then pressured the company to move forward. IBM eventually folded a System R subproject into its Future Systems initiative, yet assigned developers who lacked deep familiarity with Codd's framework and kept the team isolated from him. They bypassed Codd's own Alpha query language and instead built SEQUEL, a non-relational system. Despite this, SEQUEL proved so dramatically superior to existing alternatives that in 1979 Larry Ellison drew on pre-launch conference papers to create his Oracle Database, which actually reached the market before IBM's own SQL/DS. The name SEQUEL had to be changed to SQL because of its proprietary status.

Legacy, Standards, and the Fight for Purity

Codd's influence extended far beyond his original 1970 paper. His theorem, proven within that seminal work, established that relational algebra and relational calculus possess equivalent expressive power—a result that became a cornerstone of database theory. The Boyce-Codd normal form, a key stage in database normalization, bears his name. As the relational model surged in popularity during the early 1980s, Codd waged a sometimes bitter campaign against vendors who slapped a relational veneer onto older hierarchical technology. To draw a clear line, he published his twelve rules defining what genuinely constituted a relational database management system. This stance made his position at IBM increasingly untenable, and he ultimately left to co-found a consulting firm with Christopher J. Date, his long-time collaborator. Codd also coined the term "Online Analytical Processing" (OLAP) and authored the twelve laws governing it, though that paper was later withdrawn by Computerworld after it emerged that it had been sponsored by Arbor Software—later absorbed by Hyperion and then Oracle—without the conflict of interest being disclosed. In 2004, SIGMOD renamed its top prize the Edgar F. Codd Innovations Award in his honor.

Beyond the Databases: Cellular Automata and Final Years

Before his database work made him famous, Codd had already produced a remarkable doctoral contribution. His 1965 PhD thesis at the University of Michigan, advised by John Henry Holland, explored self-replication in cellular automata, extending the earlier theoretical work of John von Neumann. Codd demonstrated that a set of just eight states was sufficient for universal computation and construction in such systems. Strikingly, his design for a self-replicating computer was not actually implemented until 2010, decades after his death. Codd's personal life carried its own tensions. In the 1990s, his health steadily deteriorated, and he was forced to cease his professional work entirely. Edgar Frank Codd passed away on 18 April 2003 at his home in Williams Island, Florida, from heart failure, at the age of seventy-nine. He had been inducted as a Fellow of the Association for Computing Machinery in 1994, and his 1981 receipt of the ACM Turing Award had already cemented his place among the most influential figures in computing history.