Automatic programming: properties and performance of FORTRAN systems I and II
Backus, John W.
Session 2, Paper 3, Proceedings Symposium on the Mechanisation of Thought Processes, Teddington, Middlesex, England, The National Physical Laboratory, Her Majesty's Stationary Office (HMSO), pp. 232-255. "Written at a distance of one year after the delivery of the first FORTRAN processor for the 704, this paper is significant in its presentation of FORTRAN as an 'automatic programming system' in the environment of a symposium of the mechanisation of thought processes. Other attendees at the meeting included Jan Garwick (Norway), John McCarthy (USA), Grace Murray Hopper (USA) and Christopher Strachey (GB), each of whom commented on the presentation by Backus. Obviously Garwick was much more interested in telling the audience of developments by Ole-Johan Dahl while McCarthy (the author of LISP) praised FORTRAN for its ability to express '... quite lengthy algebraic expressions ...' and the implementation of separate compilation of subroutines (presumably in FORTRAN II). Hopper states that '... there is a lack of understanding of the systemizsation [sic] of FORTRAN' and asks Backus to emphasize that Fortran does more than just the 'housekeeping' for the programmer." [Lee 1984]