Re: Self-Archiving vs. Self-Publishing FAQ

From: Stevan Harnad <>
Date: Tue, 1 Feb 2000 14:30:10 +0000

Bernard Naylor's corrections about the legal definition of publication
are no doubt correct, but I think they miss the points at issue.

When academics ask about public archiving vs. publication, what they
have in mind is not the technical or legal definition of "publication,"
but something closer to what the word has meant all along for the
academic, in the "Publish-or-Perish" (PoP)sense.

It has never been sufficient (in the PoP sense) merely to put vanity
press publications on one's CV. Promotion/tenure review committees
weigh publications in the "peer-review" sense of the word: refereed
journal articles or monographs accepted by distinguished, known for
their quality-control standards, not merely texts that have somehow
contrived to be publicized.

So what was at issue in the query to which I was replying (and that was
why I referred to "preconceptions") was, I believe, PoP publication,
with CV value.

There is the further issue of priority for ideas and findings, which is
semi-independent of this: If I am the first to (correctly) prove
Fermat's Last Theorem, it matters less to me when or whether it appears
in a refereed journal than it does that I should get the credit for
having done it first. For this form of priority, presenting the
findings orally at a learned conference, or sending multiple dated
copies to peers would be the critical factor. (But I do so at my risk,
because my unrefereed proof could also be wrong, and then the only
priority I have is in making a fool of myself.)

These are, I think, what academics really have in mind when they
wonder about the relation between public online self-archiving and
classical PoP publication.

Now some quote/commentary on Bernard's corrections:

On Tue, 1 Feb 2000, Bernard Naylor wrote:

> > ap> I have some problems understanding your message.
> > ap> I have to know if this site is an archiving one
> > ap> or a publishing one via Internet.
> >
>sh> There are too many preconceptions in your question. There is not an
>sh> "either/or" relation between archiving and publishing (and publishing
>sh> has more than one meaning).
>sh> (1) Everything that appears on the Internet (or on paper,
>sh> for that matter, even once), counts legally as "publication."
> This is not true in the domain of paper. If I write to one
> individual claiming that the earth is flat, that is not a publication.

Correct, but also not at issue here, as it would not have been listed
in an academic's CV as a publication either.

> If I write a short paper purporting to prove that the earth is flat and
> circulate it to a number of my friends on a restricted basis, that is
> not a publication. In both of those instances, I do create a document
> in which there is copyright but I do not create a publication.

Again correct, but again not at issue. See point about Fermat and
priority. (I believe, by the way, that one has copyright, hence
intellectual ownership, even if one sets one's text to paper once --
but I'm not sure, and nothing at all depends on whether that is or is
not the case.)

> If I write such a paper and circulate it to my friends and make it
> clear that it can be circulated as widely as anyone may desire, then I
> have probably published it. At least, I suspect the courts might think
> so.

But as far as its PoP CV value is concerned, this case is no better
than the preceding one, in which priority and intellectual ownership
have been asserted by making one's words public, but nothing more.

> On the Internet, however, the distinction between "conversation" (in
> which I assert something to a chosen, limited number of other parties)
> and "publication" (in which I volunteer my assertion for scrutiny by
> the whole world) is much more difficult to define and I think the
> courts might have a field day. Perhaps they already have.

I think unrefereed papers and discussion on the Internet fall in
exactly the same category as the above: Good for establishing priority
and intellectual ownership, for publicizing and eliciting feedback, but
no PoP CV value eo ipso (except if the ideas and findings themselves
prove to be new, correct and important, which is of course a greater
value than mere PoP value).

> sh> But that, I
> sh> assume, is NOT what you mean by publication here. You mean:
> sh>
> sh> (2) Journal publication (or what used to be meant by journal
> sh> publication: acceptance by and appearance in a refereed journal).
> The question of whether something has been refereed or not has nothing
> to do with whether it is published. If I publish (that is, offer to the
> public at large) an unrefereed pamphlet on 1 January, setting out a
> proof that the earth is flat, and someone publishes a refereed article
> in a scholarly journal six months later setting out the same proof, I
> would have strong grounds (at the least) to claim "prior discovery".

Again, you are conflating (i) publicizing, (ii) priority and (iii) PoP
value (publication): Ideas and findings that prove to be new, correct
and important have an intrinsic academic value that transcends PoP
value, a fortiori. For these, whether or not they appear in refereed
journals is superfluous (they could just as well be announced on TV).

But most of what scholars and scientists do does not pass this high
threshold. Hence it is only peer review that takes the measure of its
value. Without that quality-control certification, they are simply
self-publicizing; caveat emptor.

  Harnad, S. (1998) The invisible hand of peer review. Nature [online]
  (5 Nov. 1998)

> Hence I think the jury has to be "still out" on the status of an
> article lying in an electronic repository and therefore available to
> the whole world - but not refereed. Suppose somebody writes a highly
> controversial article (for example, proving that the earth is flat) and
> finds that no refereed journal will publish it, I think that if such an
> article is lodged in an electronic repository on a recorded date, a
> case for "prior discovery" would quite likely stand up because I think
> that reasonable people might conclude that, refereed or not, the proof
> has been published. I therefore think that the relationship between the
> unrefereed article in the electronic repository and the identical
> article, subsequently refereed, in a paper or electronic journal is
> quite ambiguous and still needs definition. My guess is that, whatever
> we might wish to assert, the first occasion on which the article is
> offered on unrestricted circulation to the world at large will be
> deemed to be the moment of publication. I doubt very much whether the
> question of whether it has been refereed or not at some stage will be
> found particularly significant in the determination of the matter.

I do not believe that academics who are wondering about the status of
a paper they self-archive online in an Open Archive like CogPrints are
concerned about questions like these. They want to know whether it
counts for their CVs, as journal publication does. And the answer is:
No (unless what they are self-archiving IS a refereed journal article).

If the self-archived paper is brilliant and original, despite being
unrefereed, that may get the recognition and reward it deserves (if it
is not competing with too much unrefereed sludge to be detected)
without the help of refereed publication, but I would not recommend
taking that as the paradigmatic case.

So the summary is: refereed publication is still refereed publication,
and vanity publication is still vanity publication. Just as both forms
existed on paper, they will continue to exist online -- and the
distinction will continue to be noted and made. Caveat Emptor.

Stevan Harnad
Professor of Cognitive Science
Department of Electronics and phone: +44 23-80 592-582
Computer Science fax: +44 23-80 592-865
University of Southampton
Highfield, Southampton

NOTE: A complete archive of this ongoing discussion of "Freeing the
Refereed Journal Literature Through Online Self-Archiving" is available
at the American Scientist September Forum (98 & 99):
Received on Mon Jan 24 2000 - 19:17:43 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:45:40 GMT