TITLE: An Update on the First Use of the Term "Programming Language"
AUTHOR: Eugene Wallingford
DATE: December 26, 2019  1:55 PM
DESC: 
-----
BODY:
<A HREF="https://twitter.com/cristalopes/status/1208761014180667393">
   This tweet</A>
 and
<A HREF="http://www.cs.uni.edu/~wallingf/blog/archives/monthly/2019-12.html#e2019-12-23T10_23_15.htm">
   this blog entry</A>
 on the first use of the term "programming language" evoked responses
 from readers with some new history and some prior occurrences.

 Doug Moen pointed me to
<A HREF="https://www.fortran.com/FortranForTheIBM704.pdf">
   the 1956 Fortran manual</A>
 from IBM, Chapter 2 of which opens with:

<BLOCKQUOTE><EM>
 Any programming language must provide for expressing numerical
 constants and variable quantities.
</EM></BLOCKQUOTE>

 I was aware of the Fortran manual, which I link to in the notes
 for my compiler course, and its use of the term.  But I had been
 linking to a document dated October 1957, and the file at
 <TT>fortran.com</TT> is dated October 15, 1956.  That beats the
 January 1957 Newell and Shaw paper by a few months.

 As Moen said in his email, "there must be earlier references, but
 it's hard to find original documents that are early enough."

 The oldest candidate I have seen comes from @matt_dz.
<A HREF="https://twitter.com/matt_dz/status/1209196131152015361">
   His tweet</A>
 links to
<A HREF="https://ia801301.us.archive.org/14/items/DTIC_ADA032123/DTIC_ADA032123.pdf">
   this 1976 Stanford tech report</A>,
 "The Early Development of Programming Languages", co-authored by
 Knuth.  On Page 26, it refers to work done by Arthur W. Burks in
 1950:

<BLOCKQUOTE><EM>
 In 1950, Burks sketched a so-called "Intermediate Programming
 Language" which was to be the step one notch above the Internal
 Program Language.
</EM></BLOCKQUOTE>

 Unfortunately, though, this report's only passage from Burke
 refers to the new language as "Intermediate PL", which obscures
 the meaning of the 'P'.  Furthermore, the title of Burke's paper
 uses "program" in the language's name:

<BLOCKQUOTE><EM>
 Arthur W. Burks, "An intermediate program language as an aid in
 program synthesis", Engineering Research Institute, Report for
 Burroughs Adding Machine Company (Ann Arbor, Mich.: Univ. of
 Michigan, 1951), ii+15 pp.
</EM></BLOCKQUOTE>

 The use of "program language" in this title is consistent with
 the terminology in Burks's previous work on an "Internal Program
 Language", to which Knuth et al. also refer.

 Following up on the Stanford tech report, Douglas Moen found the
 book <EM>Mathematical Methods in Program Development</EM>, edited
 by Manfred Broy and Birgit Schieder.  It includes a paper that
 attempts "to identify the first 'programming language', and the
 first use of that term".  Here's a passage from Page 219, via
 Google Books:

<BLOCKQUOTE><EM>
 There is not yet an indication of the term 'programming languages'.
 But around 1950 the term 'program' comes into wide use: 'The logic
 of programming electronic digital computers' (A. W. Burks 1950),
 'Programme organization and initial orders for the EDSAC' (D. J.
 Wheeler 1950), 'A program composition technique' (H. B. Curry 1950),
 'La conception du programme' (Corrado Bohm 1952), and finally
 'Logical or non-mathematical programmes' (C. S. Strachey 1952).
</EM></BLOCKQUOTE>

 And then, on Page 224, it comments specifically on Burks's work:

<BLOCKQUOTE><EM>
 A. W. Burks ('An intermediate program language as an aid in program
 synthesis', 1951) was among the first to use the term program(ming)
 language.
</EM></BLOCKQUOTE>

 The parenthetical in that phrase -- "the first to use the term
 program(ming) language" -- leads me to wonder if Burks may use
 "program language" rather than "programming language" in his 1951
 paper.

 Is it possible that Knuth et al. retrofitted the use of "programming
 language" onto Burks's language?  Their report documents the early
 development of PL ideas, not the history of the term itself.  The
 authors may have used a term that was in common parlance by 1976
 even if Burks had not.  I'd really like to find an original copy of
 Burks's 1951 ERI report to see if he ever uses "programming language"
 when talking about his Intermediate PL.  Maybe after the holiday
 break...

 In any case, the use of <EM>program language</EM> by Burks and others
 circa 1950 seems to be the bridge between use of the terms "program"
 and "language" independently and the use of "programming language"
 that soon became standard.  If Burke and his group never used the
 new term for its intermediate PL, it's likely that someone else did
 between 1951 and release of the 1956 Fortran manual.

 There is so much to learn.  I'm glad that Crista Lopes tweeted her
 question on Sunday and that so many others have contributed to the
 answer!
-----