Citation is Medium-Independent

From: Stevan Harnad <>
Date: Wed, 11 Aug 1999 20:32:18 +0100

On Wed, 11 Aug 1999, forsdyke wrote:

> sh> There is no reason whatsoever why AUTHORS should sit back and wait to
> sh> self-archive until journals and their editors first see fit to plan a
> To "self archive" you need
> (i) a secure archive,

LANL is that; so is CogPrints. So will E-biomed be.

> (ii) a citation key which will allow you and others to cite and
> access.

I don't know what this means. The citation "key" for self-archived
refereed-journal papers is just the usual journal citation, and the
access is the archive. For unrefereed preprints the access is the
archive and the preprint citation details plus its archive code.

> The first has long been in existence. The whole internet is one big
> archive. If you have a document, you have long been free to "post" it
> at numerous sites, which themselves form archives (e.g.
> bionet.journals.note). Here it can be dated, and by sending it to
> multiple archives you make it less likely that it will be deleted and
> lost for all time.

Yes, self-archiving has long been available, but archives often stop
being maintained. That is the virtue of collective, high-profile
archives like LANL and E-biomed.

(Nor is a lot of the Net/Web either reliable or secure.
But, more relevant, most of what is "posted" and even "archived" on
the Net is nonarchival -- neither intended nor worth saving or citing,
any more than throwaway conversation is.)

> The problem comes when you want to cite it. Clearly you do not want to
> have to cite all the sites where it is deposited (and an editor of a
> paper journal would not permit it).

This is a non-problem. How would you cite it in paper? Do the same;
leave out the page numbers, and add the archive URL.

> Furthermore, there is no unanimous agreement on how, say,
> bionet.journals.note should be cited. I have suggested following the
> existing way journals are cited in the paper literature.

Most of what is posted on Usenet and Listserv lists is not worth citing
or saving. It certainly is not the "archival literature." If it IS
worth citing and saving, it should be deposited in an archival serial
of some sort (whether refereed or unrefereed, established or new, paper
or online). And then there is again no problem or mystery: Use the
serial's citation format, plus the archive URL.

But let us not conflate the fact that (a) something has been posted and
"archived" somewhere on the Net (which, with a few exceptions -- easily
remedied by also archiving it somewhere more formal -- is just a
vanity-press, auto-promotional matter, if not a Global Graffiti Board
for Trivial Pursuit), with (b) the kind of formal publication I would think
this discussion was concerned with.

The non-problem arises out of this conflation. For a refereed journal
already has a "citation key" -- the formal, continuous, ARCHIVAL serial
that it is -- and it just takes a trivial modification to adapt that
"key" to the online medium. But an arbitrary posting to a list is just
that! (It's rather as if we started to worry about the "citation key"
for notes passed to one another during class! Not everything is
"archival," even if it can now be "archived," and has an ongoing
subject "thread.")

> Thus:
> Harnad, S. (1999) Bionet.journal. note 0811, 1253. The forgotten
> importance of editors.
> In this case, the "volume" is the month and day, and the page is the
> hour and min, of the initial deposit into the site. Surely, someone
> setting up newsgroup software could arrange this in the heading of
> electronic contributions?

(I can say what I'm about to say, because it's in the context of the
putative immortalization of my own opus, above): If I had written that
as a graffito on a lavatory wall, that too would have been "archived"
(as long as it was not erased), with time, place, date, title, author.
Is it a "problem," and do we need a "citation key," in order to inform
people how to refer to that graffito, and how to locate the lavatory
archiving it?

If there was no lav-wall problem before, there isn't one now, for
skywritten graffiti either. If my words really are immortal, they
should be placed in a formal archival serial (whether paper or online
-- the importance is the formal continuity, which is why they are called
serials), and that should in turn be reliably archived, and cited the
usual way, along with its URL.

By way of example; I too own a Usenet Newsgroup:
sci.psychology.journals.psycoloquy -- but that also happens to be a
refereed archival journal of 10 years standing now. It has its own
citation format, and it has more reliable and accessible archival sites
than its Usenet (and Listserv incarnations,

There is another way that postings worthy of citation can carve a
permanent niche in collective memory: They can be preserved by
individuals as "personal communications" (= letters, whether paper or
online), and they can then either be kept alive by the oral (and now
online) tradition, or they can find their way into yet another
permanent entity, a book (whether paper or online).

That more or less exhausts the conventional possibilities, nor do I
really see any pressing need for new ones. It would be nice if all
skywritings could readily be "grepped" back at will, but it's not such a
great loss if they can't be, and only the ones worth archiving formally
can (and are).

So I think you have to distinguish that part of your agenda which is
(ceteris paribus) formally the same as wishing to index graffiti, from
the part that is concerned with establishing more archival organs
online. In neither case are any fundamental problems involved; only
conflating the two makes it seem as if there were a problem.

Stevan Harnad
Professor of Cognitive Science
Department of Electronics and phone: +44 2380 592-582
Computer Science fax: +44 2380 592-865
University of Southampton
Highfield, Southampton
Received on Wed Feb 10 1999 - 19:17:43 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:45:35 GMT