Re: Paper not accepted by a journal - still a pre-print?

From: Stevan Harnad <>
Date: Tue, 6 Aug 2002 19:42:38 +0100

On Tue, 6 Aug 2002, Eberhard R. Hilf wrote:

> Of course: 'preprint' and 'unrefereed'.

That's fine by way of reply to the query, which asked whether a paper
rejected by a journal should be self-archived as "refereed" or

> Refereed means in an archive, that the paper has passed.

This is incorrect. This is not what "refereed" means, and here, as a
reply to the above query, it can only cause confusion. "Refereed" does
not mean in an archive; and it is unclear what it would/should mean to
"pass" in an archive.

"Refereed" means having successfully passed peer review by a refereed
journal, a refereed conference proceedings, or some other established
and recognized form of peer review.

> Otherwise it is 'sent back' and the author can do what he wants.
> As long as a journal has not accepted it.

Unclear: Sent by by whom, from what, for what reason?

If you mean a paper is unrefereed until/unless it has been accepted by
a refereed journal, that is correct. But then please make it clear that
what you mean here is that it has been submitted to and refereed by (and
"sent back by") a journal (etc.), not an archive. The poster's question
was about whether to archive a journal-rejected article in an archive
as "refereed" or "unrefereed." (The rationale for the question had
presumably been that in a sense it HAS been refereed, only it has
failed to be accepted.) Hence the clear answer to the poster is: if the
journal has rejected it, it is not "refereed," and should be archived
as unrefereed.

The poster was obviously not asking about how to tag a paper in an
archive if the ARCHIVE has rejected the paper!

> If a journal has accepted it, and their policy is not to have referees
> at all, then it is 'journal article' and 'unrefereed'. I know of no
> such journal.


> P.S.: But be aware: in an e-archive you can have many more subtle and
> precise levels of certification. And they are an advantage and make
> the e-print so much more powerful than just the refereed/unrefereed,
> saying nothing about the quality of the act.

It is important to point out that the above is not a fact, but merely a
speculation by Ebs about a hypothetical future. The fact is that it is
journal refereeing -- and especially the established quality level and
standards (and impact factor) of the particular journal that has accepted
the paper in question -- that provides the only official certification
at the present time. Nor is there yet any evidence whatsoever of
"more subtle and precise levels of certification." Unless Ebs can
cite references indicating exactly what certifiers he is referring to
(and what the evidence it that they are "more subtle and precise" than
standard peer review, it is important that he make it clear that he is
merely speculating at this time.

> So, create a field -- "certification" -- and give it a list of
> possiblities, say c0 - c7.
> For example:
> c=0 author thinks paper should be archived

This seems trivial. Would an author self-archive a paper that he did not
think should be archived? The AUTHORNAME tag seems to cover this.

> c=1 author is a professional by attached homepage showing his PhD in the
> field or his prof. position in a profess. institution of the field.

This is not the "certification" tag but the DEPARTMENT/INSTITUTION tag
(and URL) (and perhaps an optional DEGREES tag). Again, re-interpreting
this as "certification" is confusing and trivializing certification.

> c=2 a technical check has been made (formats, metadata, etc. by the
> archive)

This seems superfluous. It is not "certification" of the content. It is
merely an archive's own routine checking of metadata. Appearing in that
archive is de facto attestation to the fact that the paper has met that
archive's criteria (whatever those happen to be).

> c=3 a library expert has read the paper

So what? The relevant experts are specialists in the field of the paper,
not librarians.

> c=4 a loose screening has been done by an external expert on that
> field (topical screening)

What does this mean?

> c=5 a thorough blind refereeing has been done by a real expert.

Until further notice, to be able to say this with the kind of authority
needed so that people can trust it, and can know what to do with it, it
has to be said by a refereed journal, with known quality standards. Hence
"c-5" is conventional peer review, the only form of certification so far.
This is covered by the REFEREED and JOURNAL-NAME tags.

> c=6 paper has been annotated, commented by other professionals openly.
> and so forth.

This is peer commentary, not peer review. A valuable supplement to peer
review, but not a substitute for it, and covered by the COMMENTARY and

    Peer Review, Peer Commentary, and Eprint Archive Policy

> Since you as an archive will store the paper right from the beginning,
> you just keep changing the value of c and keep the paper.

These are not changing values of a certification "c" but additional,

> Nothing is rejected, but the user is told what status the paper has.

> Further reading:
> E.R.Hilf and H.-J.Waetjen:
> Scientific Refereeing in a Distributed World

Institutional archives are neither for accepting/rejecting nor for
certifying. They are for archiving their own research output. They can
have internal criteria for what may be archived therein (e.g., only research,
only research by researchers affiliated with the institution, and
any further internal criteria the institution may wish to impose on its
archive), but this has nothing to do with certification, peer review,
or other established external indicators of quality-level (chief of which
is JOURNAL-NAME at the present time).

Stevan Harnad

NOTE: A complete archive of the ongoing discussion of providing free
access to the refereed journal literature online is available at the
American Scientist September Forum (98 & 99 & 00 & 01):

Discussion can be posted to:

See also the Budapest Open Access Initiative:

and the Free Online Scholarship Movement:
Received on Tue Aug 06 2002 - 19:42:38 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:46:36 GMT