TITLE: The First Use of the Term "Programming Language"?
AUTHOR: Eugene Wallingford
DATE: December 23, 2019 10:23 AM
DESC: 
-----
BODY:
 Yesterday, Crista Lopes
<A HREF="https://twitter.com/cristalopes/status/1208761014180667393">
   asked a history question</A>
 on Twitter:

<BLOCKQUOTE><EM>
 Hey, CS History Twitter:  I just read Iverson's preface of his
 1962 book carefully, and suddenly this occurred to me:  did he
 coin the term "programming language"?  Was that the first time
 a programming language was called "programming language"?
</EM></BLOCKQUOTE>

 In a follow-up, she noted that  McCarthy's <EM>CACM</EM> paper
 on LISP from roughly the same time called Lisp a 'programming
 system'", not a programming language.

 I had a vague recollection from my grad school days that Newell
 and Simon might have used the term.  I looked up
<A HREF="https://en.wikipedia.org/wiki/Information_Processing_Language">
   IPL</A>,
 the Information Processing Language they created in the mid-1950s
 with Shaw.  IPL pioneered the notion of list processing, though
 at the level of assembly language.  I first learned of it while
 devouring Newell and Simon's early work on AI and reading every
 thing I could find about programs such as the General Problem
 Solver and Logic Theorist.

 That wikipedia page has a link to 
<A HREF="http://bitsavers.org/pdf/rand/ipl/">
   this unexpected cache of documents</A>
 on IPL from Newell, Simon, and Shaw's days at Rand.  The oldest
 of these is a January 1957 paper,
<A HREF="http://bitsavers.org/pdf/rand/ipl/P-954_Programming_The_Logic_Theory_Machine_Jan57.pdf">
   Programming the Logic Theory Machine</A>,
 by Newell and Shaw that was presented at the Western Joint
 Computer Conference (WJCC) the next month.  It details their
 efforts to build computer systems to perform symbolic reasoning,
 as well as the language they used to code their programs.

 There it is on Page 5: a section titled "Requirements for the
 Programming Language".  They even define what they mean by
 <EM>programming language</EM>:

<BLOCKQUOTE><EM>
 We can transform these statements about the general nature of
 the program of LT into a set of requirements for a programming
 language.  By a programming language we mean a set of symbols
 and conventions that allows a programmer to specify to the
 computer what processes he wants carried out.
</EM></BLOCKQUOTE>

 Other than the gendered language, that definition works pretty
 well even today.

 The fact that Newell and Shaw defined "programming language"
 in this paper indicates that the term probably was not in
 widespread use at the time.  The WJCC was
<A HREF="https://en.wikipedia.org/wiki/Joint_Computer_Conference">
   a major computing conference</A>
 of the day.  The researchers and engineers who attended it
 would likely be familiar with common jargon of the industry.

 Reading papers about IPL is an education across a range of
 ideas in computing.  Researchers at the dawn of computing had
 to contend with -- and invent -- concepts at multiple levels
 of abstraction and figure out how to implement the on machines
 with limited size and speed.  What a treat these papers are.

 I love to read original papers from the beginning of our
 discipline, and I love to learn about the history of words.  A
 few of my students do, too.  One student stopped in after the
 last day of my compilers course this semester to thank me for
 telling stories about the history of compilers occasionally.
 Next semester, I teach our Programming Languages and Paradigms
 course again, and this little story might add a touch of color
 to our first days together.

 All this said, I am neither a historian of computer science nor
 a lexicographer.  If you know of an earlier occurrence of the
 term "programming language" than Newell and Shaw's from January
 1957, I would love to hear from you
<A HREF="mailto:wallingf@cs.uni.edu">
   by email</A>
 or
<A HREF="https://twitter.com/wallingf/status/1208764620397334528">
   on Twitter</A>.
-----