[SIGCIS-Members] First usage of the word "subroutine"?

[Thomas Haigh]

Hello everyone,

 

I'm wondering if anyone has pinned down the origin of the word "subroutine."
OED currently has as the first two example usages:

 

c1946   H. H. Goldstine & J. Von Neumann in J. Von Neumann Coll. Wks. (1961)
V. 25   Both..machines are controlled by instructions punched into several
tapes and they can be ordered to switch from one to the other as desired.
They are usually referred to as 'master routine' and 'sub-routine' tapes.
[The ellipsis hide reference to the Harvard Mark 1 and to the Bell Labs
relay machines].

 

1948   H. H. Goldstine & J. Von Neumann in J. Von Neumann Coll. Wks. (1963)
V. 217   We call the coded sequence of a problem a routine, and one which is
formed for the purpose of possible substitution into other routines a
subroutine.

 

In my work on ENIAC with Mark Priestley and Crispin Rope we have located an
earlier (1945) usage in the "AMP Report" authored by members of the ENIAC
team. "It is possible to have the main routine divided into sub-routines, in
which case one stepper is used to feed another stepper, thus allowing the
proper sub-routine to be chosen in the course of a regular routine."  J. P.
Eckert, J. W. Mauchly, H. H. Goldstine, and J. G. Brainerd, Description of
the ENIAC and Comments on Electronic Digital Machines. AMP Report 171.2R.
Distributed by the Applied Mathematics Panel, National Defense Research
Committee, November  30. Philadelphia, PA: Moore School of Electrical
Engineering, 1945.

 

I'm also aware that the term has been applied in later usage to useful
blocks of code saved and copied into programs as needed on the Harvard Mark
1, e.g. in K. W. Beyer, Grace Hopper and the Invention of the Information
Age. Cambridge, MA: MIT Press, 2009 pages 95-106. That fits in the general
sense with the 1946 von Neumann quote, although the Harvard Mark 1 did not
in 1945 have the ability to switch to a second tape and was instead confined
to running like a player piano over a single series of instructions. I did
not spot any quotation of the actual word "subroutine" from 1945 or 1946
sources in Beyer's book. His citations in that section all come from much
later oral histories and reminiscences, not from 1940s documents. (Also
Beyer seems unaware of the refinement of the idea by von Neumann and his
colleagues in the 1948 installment of the "Planning and Coding" report,
instead considering only the transfer of ideas to the EDSAC team in the UK).

 

So, my questions are:

 

1)       Does anyone have a documented usage prior to November 1945 of the
word "subroutine" or the hyphenated "sub-routine"?

2)       Are there primary sources that will tell us whether the use of
"routine" to describe a program was standard terminology at Harvard, at Bell
Labs, or at both? I'd expect the coinage of "subroutine" to come naturally
at a site in which programs were called "routines" and less so elsewhere.

 

Best wishes,

 

Tom

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sigcis.org/pipermail/members-sigcis.org/attachments/20131111/f9146b1a/attachment-0001.htm>