Search Mailing List Archives

Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[liberationtech] Who first created/coined email?

Yosem Companys companys at
Fri Mar 2 10:44:45 PST 2012

> I can't find clarity about who first created a mechanism we can call
> email on a timesharing system.

Who first made a timeshared messaging system is one thing.  Who first
coined the word "email" for electronic mail is another.  Both are
worth looking into.

If only the academics of the time had had a system for communicating
ideas with each other, that had left a paper trail.  Hey wait, they
did!  It was called papers and conferences and proceedings and
journals.  If only the public had easy access to that stuff now, we
could merely search it for the first occurrence of "email" and get an
interesting data point.  Unfortunately, all that research
correspondence is locked up behind paywalls, if it has been digitized
at all.  Maybe one of you who has a paywall account (e.g. "ACM Digital
Library" card or a similar IEEE thing) could do such a search?

What does the O.E.D. say?  They're behind a paywall, too.

The Usenet archives at show a single mention of
EMAIL in 1981, but it refers to a specific program for email, and it
originated in an ARPAnet mailing list that was gatewayed to Usenet.  A
Human-Nets posting from 19 May 1981 by ROODE at SRI-KL (David Roode)
complains about CompuServe, saying:

 It truly does seem to me that their attitude embodies the idea that
 their potential customers will be ignorant of any better way of doing
 things than those they choose.  Their EMAIL program is the most
 cumbersome messaging system I have seen, and its improved palatability
 for the naive user over MM, MSG, RMAIL and the like is not clear to

(CompuServe ran on PDP-10s; what was the origin of the EMAIL program
that they offered the public?)

The first mention of email as a service (in the Google Groups archive)
was more than a year later, on 15 October 1982, by my old friend
Robert Elton Maas <REM at MIT-MC>, talking about PCNET, which is a
project that I worked on with him (he wrote the 8080 CP/M
implementation, I wrote the Apple II one, and we got them to exchange
non-internet email over dialup modems, with each other and with a
Commodore PET, using a nice packet protocol; we demoed it at the West
Coast Computer Faire).  His message says in part:

 A single network design should support all speeds of equipment
 rather than requiring different speeds to be on different and
 disjoint networks (at the least, gateways for the major services,
 email, ftp, telnet) should exist even where differing equipment
 requires differing low-level protocols.

(The topic of that whole Human-Nets digest issue was "WorldNet", a
hard-to-visualize dream about tying all sorts of computers together

By 1983 there are a hundred references to email, in many newsgroups,
including net.sources,, net.general, net.bicycle,, and net.sf-lovers.  Here's a notable one:

 Newsgroups: net.mail
 From: diam... at cwruecmp.UUCP (John Diamant)
 Date: Thu, 10-Nov-83 00:28:06 EST
 Local: Thurs, Nov 10 1983 12:28 am
 Subject: Re: speaking of needed words

 In response to the need for a word to distinguish between a letter and
 electronic mail:  this word exists.  It is EMAIL (for electronic mail,
 of course).

 John Diamant                            Usenet: ...decvax!cwruecmp!diamant
 Case Western Reserve University         CSNet:  diamant at Case
 Cleveland, Ohio                         ARPA:   diamant.Case at Rand-Relay

There's an interesting history of ARPANET email here:

It was written by Ian R. Hardy of UC Berkeley in 1996 as a history
thesis.  It uses the word "email" throughout, but doesn't go into the
history of the word.  But it does mention the paucity of official
references to electronic mail (as a network service) in early official
ARPANET reports and such, quoting Ray Tomlinson and Frank Heart.
Since email wasn't an official government project or priority, it
wasn't mentioned til "about 1976", even though all the researchers
were using it and it was 75% of the network traffic.

RFC 808 of 1 March 1982, "Summary of Computer Mail Services Meeting"
by Jon Postel documents a meeting held on 10 January 1979 at BBN.  Its
appendix contains a list of all known electronic mail systems on the
ARPAnet as of January 1979 (compiled by Dave Farber!) and their author
and operating system.  The majority of these names use the word "mail"
(e.g. RDMAIL, MAILER, Read-mail) but none use "email", nor does the
RFC itself.

Among the RFCs, the first RFC to include "email" or "e-mail" was RFC
977 of February 1986, for NNTP.  The word occurred in an example
Usenet posting, apparently a real one snatched from a spool directory
by the RFC author, from newsgroup net.unix-wizards, dated 25 Sep 85, and
containing the concluding paragraph "Please reply by E-mail.  Thanks
in advance."  The next was RFC 1048 of February 1988, "BOOTP Vendor
Extensions", which provided contact information for registering
additional data fields at a postal address, "or by E-mail as:
JKREYNOLDS at ISI.EDU".  The next was RFC 1060 of March 1990, one of the
"Assigned Numbers" series, which marked a change in the RFC format.
It included an "Authors' Addresses" section, which contained "Email:
JKREY at ISI.EDU" and "Email: POSTEL at ISI.EDU".  The previous Assigned
Numbers, RFC 1010 of May 1987, did not have an "Authors' Addresses"
section, and though it included a table of protocol designers and
their companies and email addresses, no column heading was used to
describe the email addresses.

The first RFC to include the word "email" or "e-mail" in a sentence
was RFC 1093 of February 1989, "The NSFNET Routing Architecture",
whose introduction said, "Thanks also to: Milo Medin (NASA), John Moy
(Proteon) and Greg Satz (Cisco) for discussing this document by email
and/or phone."  They started coming thick and fast then, and by
December 1989, RFC 1137's page headings called the document's topic
"E-Mail Address and Quoted Strings".  By RFC 1336 in May 1992, Steve
Crocker is quoted as saying:

          E-mail dominates the Internet, and it's likely to remain the
          dominant use of the Internet in the future.  Nonetheless, I
          expect to see an exciting array of other applications which
          become heavily used and cause a change in the perception of
          the Internet as primarily a "mail system."

This doesn't tell us the precise origin of the word, but it does
indicate that in 1981-2 very few people on the ARPAnet and Usenet were
using it, and that in 1983 its usage broke out on the Usenet.  Perhaps
further inquiries should be made on the Usenet and Unix mail side of
history.  And does anyone have conveniently searchable archives of
early ARPAnet mailing lists, BITNET, CSNET, etc?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the liberationtech mailing list