In this biography
Quick facts
- Born
- 3 December 1924
- Died
- 17 March 2007
- Birthplace
- Philadelphia, Pennsylvania, United States
- Country
- United States
- Nationality
- Not independently established
- Known for
- Leading the original FORTRAN team, syntax notation and functional programming research
- Earned degrees
- BS mathematics, 1949; MA mathematics, 1950, Columbia
- IBM career
- 1950–1991
- FORTRAN release
- April 1957 initial distribution
- Turing Award
- 1977 award year
Why this life matters
Backus helped make high-level programming credible through a compiler designed for efficient execution, while continually questioning the limitations of conventional programming.
Timeline
Key moments
Born in Philadelphia
Born on 3 December.
Columbia mathematics
Completed bachelor’s and master’s degrees.
IBM programmer
Began work associated with the Selective Sequence Electronic Calculator.
FORTRAN development
Led the developing numerical-language and compiler project.
Original system distributed
An April letter announced that the IBM 704 system was ready for distribution.
IBM Fellow
Became an IBM Fellow.
Turing Award
Recognised for high-level systems and formal language specification.
Retirement
Retired from IBM.
Death
Died on 17 March in Ashland, Oregon.
A circuitous education led to mathematics
John Warner Backus was born in Philadelphia on 3 December 1924. His early university and military studies included chemistry, engineering and medicine, but his completed university qualifications were in mathematics: a BS from Columbia in 1949 and an MA in 1950. He did not earn the doctorate sometimes casually implied by his later scientific stature. Columbia’s computing history records that the GI Bill supported his studies. The Library of Congress’s biographical chronology distinguishes attendance at earlier institutions from those completed degrees, a useful correction to a career story often reduced to an anecdote about disliking conventional study.
Programming began with a large machine and detailed work
Backus joined IBM in 1950 and worked as a programmer on the Selective Sequence Electronic Calculator. He later led development of Speedcoding for the IBM 701. Such systems addressed a real burden: expressing numerical procedures in the instructions and storage conventions of a particular computer. They also had costs of their own. In his later FORTRAN history, Backus described the scepticism that automatic programming would produce sufficiently efficient results. His account acknowledges earlier systems, including the algebraic compiler of J. Halcombe Laning and Neal Zierler. FORTRAN therefore belongs within an existing history of programming systems.
FORTRAN was a team’s language and compiler
Backus led the group that developed FORTRAN for the IBM 704. The original 1957 paper credits thirteen authors, including Lois M. Haibt, Sheldon Best, Harlan Herrick, Robert Nelson, Irving Ziller, Peter Sheridan, Richard Goldberg, David Sayre and others. Its divisions identify contributions such as Haibt’s flow analysis and Best’s register work. This allocation is stronger evidence than calling Backus the sole inventor. The project connected concise mathematical notation with a translator expected to produce efficient machine programs. Its aim included reducing preparation and debugging effort without imposing an unacceptable execution penalty.
Sources for this section: [5]
The release record exposes practical constraints
The October 1956 reference manual describes separate source and object programs and targets numerical computation, while acknowledging weaker support for some logical tasks. Backus’s 24 April 1957 letter announced that the initial system was ready for distribution to SHARE members owning IBM 704s. It specified memory and peripheral requirements, instructions for protecting master tapes and ways to report errors. That letter also acknowledged that undiscovered system defects remained. A useful language needed a working distribution, documentation, libraries, corrections and feedback. It did not become reliable merely because the translation was automatic.
Syntax notation and a new challenge after FORTRAN
Backus participated in the ALGOL 58 and ALGOL 60 work. His technique for describing syntax was improved and used by Peter Naur in the ALGOL report, giving rise to the familiar Backus–Naur Form name. A grammar describes which constructions a language permits; it is not by itself a proof that every permitted program computes the intended answer. Backus became an IBM Fellow in 1963. From the late 1960s he pursued applicative and functional approaches, with collaborators including Dines Bjørner and Paul McJones, seeking more direct ways to build and reason about whole computations.
The later programme was a proposal to examine
The 1977 Turing Award recognised Backus’s high-level programming systems and formal language specification. His associated lecture argued for a functional style and an algebra of programs. McJones’s research archive records both the ideas and the difficulties of turning them into complete practical systems, including critical responses from other researchers. Backus’s later work should therefore be described as an influential research programme, not a completed solution to every software problem. He retired from IBM in 1991 and died in Ashland, Oregon, on 17 March 2007. Across these phases, the recurring question was how to reduce the human effort required to tell a computer what to do.
Quick answers
Frequently asked questions about John Backus
- Was Backus the only creator of FORTRAN?
- No. He led the IBM project; the original paper names thirteen authors and assigns particular compiler sections to team members.
- Was FORTRAN the first attempt at high-level programming?
- No. Backus’s own history discusses earlier automatic programming systems and algebraic compilers. The original FORTRAN project’s efficient compilation was a central part of its achievement.
Keep exploring
Lives connected to this story
- John Cocke
Explore another approach to coordinating computer architecture and compiled programs.
- Donald Knuth
Compare the compiler problem with the different task of organising programs for readers.
- Dorothy Vaughan
Explore a separate career in scientific computing and the later application of FORTRAN at NASA; compiler creation and use are distinct contributions.
- John von Neumann
Explore the earlier problem of storing instructions and organising a computer, distinct from the later compiler project.
Sources and further reading
Explore the evidence
These references support the biography and offer further reading. Official records, first-person accounts and independent reporting provide different perspectives; image records document attribution and licensing.
- 1. Columbia University Computing HistoryJohn Backus: degrees, IBM career and correspondence
- 2. Library of Congress Manuscript DivisionJohn W. Backus Papers: biographical note and collection description
Institutional indexed biographical note and collection description read. Direct finding-aid request returned 403; no manuscript boxes or unpublished correspondence were examined.
- 3. IBMJohn Backus: programming systems and research career
- 4. IEEE Computer SocietyJohn Warner Backus: biography and publication record
- 5. J. W. Backus and twelve coauthors; Computer History Museum Software Preservation GroupThe FORTRAN Automatic Coding System (1957), pp. 188–198, original reprint
Complete twelve-page reprint acquired. Author list, opening rationale, language examples, six-section organisation, and sections 4–5 read; selected pages visually checked. No historical compiler was executed.
- 6. IBM Applied Science Division and Programming Research Department; Computer History Museum scanProgrammer’s Reference Manual: FORTRAN for the IBM 704 EDPM, 15 October 1956
Complete 54-page scan acquired; title dated 15 October 1956, opening scope, card format, FREQUENCY and source/object-machine sections read. This advance manual predicted a late-1956 availability; the actual distribution claim uses the April 1957 letter.
- 7. IBM; Computer History Museum Software Preservation GroupJohn W. Backus to John Greenstadt, 24 April 1957: initial FORTRAN distribution
- 8. John Backus; Computer History Museum Software Preservation GroupThe History of FORTRAN I, II, and III, History of Programming Languages (1981), pp. 25–45
Complete 21-page 1981 chapter acquired; relevant opening economics, prior systems, project chronology and author’s correction of his earlier recollection read. Retrospective testimony is identified, not treated as a contemporaneous log.
- 9. Computer History Museum Software Preservation GroupHistory of John Backus’s functional programming project, Paul McJones
McJones’s first-person archive account read for collaboration, 1977 award and research limitations; detailed FP implementation listings were not run.
