Peter Naur

Peter Naur (25 October 1928 – 3 January 2016) was born in Frederiksberg, Denmark. He is remembered as the editor of the ALGOL 60 report, a title that undersells the job he actually did: casting the argument of an international committee into one coherent document.

From astronomy

He took a master's in astronomy at the University of Copenhagen in 1949, and spent 1950–51 at Cambridge, where he used EDSAC to compute the orbits of comets and asteroids. His encounter with computers was a by-product of that work. In 1957 he received a doctorate for a thesis on the minor planet 51 Nemausa and the fundamental system of declinations. A trained astronomer writing the foundations of computer science was not unusual for the period; there were no other trained computer scientists.

Regnecentralen and ALGOL

From 1959 to 1969 he worked at Regnecentralen, Denmark's first computer company, lecturing at the same time at the Technical University of Denmark and the Niels Bohr Institute.

After the 1958 Zurich conference produced the preliminary ALGOL report, an implementation conference met in Copenhagen and Regnecentralen began publishing the ALGOL Bulletin, with Naur as editor. Holding the forum put him at the centre of the international design argument, and in November 1959 he was chosen for the European language-design group.

Ahead of the January 1960 Paris conference, Naur worked up a completely new draft report from the preliminary report and the preparatory meetings' recommendations. The conference adopted it as the basis for its work and then negotiated item by item. The resulting document describes itself as the union of the committee's concepts and the intersection of its agreements.

In it, he reworked the syntax notation John Backus had built for ALGOL 58 and wrote the whole language out in it. In 1964 Donald Knuth proposed the change of name under which it has been known ever since: Backus-Naur form, BNF.

The University of Copenhagen also credits Naur with pushing for the inclusion of recursion in the language. In April 1962 some of the original authors reconvened to produce the Revised Report.

Datalogi

In 1969 he took Denmark's first professorship in computing, which he held until 1998, and from which the University of Copenhagen's department of computer science (DIKU) became autonomous in 1970.

Naur disliked the phrase computer science: naming a discipline after a machine struck him as a category error. His alternative, datalogi — the science of the nature and use of data — actually took hold in Denmark and Sweden, which is why the department still carries that name.

What a program is

His later interests moved away from formal language definition and towards programming as a human activity.

The 1985 essay "Programming as Theory Building" treats a program not as a deliverable but as a theory held by the people who built it. Code and documentation are incomplete projections of that theory; consequently a program dies when the team possessing the theory is dissolved, however intact the source may be. It remains one of the most frequently cited arguments in software engineering. Computing: A Human Activity (1992) collects that line of thought.

He ended up arguing that artificial intelligence is impossible — that computers cannot reproduce human decision-making. The person who showed most precisely what could be formalized spent his later career on what could not.

The Turing Award

Naur received the ACM A.M. Turing Award in 2005. The citation, as quoted by the University of Copenhagen, reads: "For fundamental contributions to programming language design and the definition of Algol 60, to compiler design, and to the art and practice of computer programming."

He died in Herlev, Denmark, on 3 January 2016.