Re: Online Self-Archiving: Distinguishing the Optimal from the Optional

From: Stevan Harnad <>
Date: Tue, 11 May 1999 17:54:23 +0100

On Tue, 11 May 1999, Arthur Smith wrote:

> The distinction is really one of responsibility. Under system III
> the author is responsible for the freely distributed version.
> Under system II, the journal or other authoritative source takes
> responsibility. Who do you trust more to get accurate information
> about acceptance and publication status?

During the growth phase of self-archiving by authors, things will
only get better and better, with more and more of the literature
becoming available free online. And the more definitive S/L/P version
will continue to be there if/when anyone needs it (if they can afford it).

But as the culture evolves into this new niche, as the free archive
becomes the locus classicus for search and retrieval, authors (who are,
after all, the reader/users too) will learn that it is in their
interests to make their free versions as definitive as possible (and it's
possible to make them VERY definitive).

A metadata tagging system will evolve for labeling the author-based
definitive version, and a warning system could even be developed for
filing and linking any discrepancies that readers notice between
"definitive" versions and the S/P/L versions. (It will be in the author's
interest to minimize such discrepancies.)

I could go on prognosticating like this, second-guessing how ingeniously
the community will capitalize on the remarkable resources of the online
medium once their precious intellectual wares are in the same communal
basket, but it should already be abundantly clear that this free
archive will be remarkably useful, and that its usefulness will only
increase its use, which will again augment its usefulness...

In short, I don't think the limitations you cite will be an obstacle
in the growth phase, and it's too early to talk about the end-game.
The reason this proposal is subversive is that the author has only
to gain, and nothing to lose, by self-archiving.


> sh> (1) [Authors] will self-archive all their papers online
> They may or may not. Some authors will be diligent on this, others
> will be lazy - there are barriers to "self-archiving", though not large.
> I find it unlikely that you will ever see 100% self-archiving...

Let's wait and see. For now, I am concerned with removing the barriers
to self-archiving, which are (1) publisher (not APS) attempts to
prevent it via copyright agreement and author intimidation, (2) author
unawareness of the possibilities and the benefits, (3) the need for
much stronger archiving platforms, both local (servers for
self-archiving for all departments of all universities) and global
(pandisciplinary extensions of LANL such as E-bionet and Scholar's

Once authors fully realize the revolutionary advantages of
self-archiving, have the archives to self-archive in, and do not feel
prevented by their publishers, sit back and watch what happens.


> In any case, the empirical evidence from physics is actually not
> overwhelmingly one way or the other, but it actually
> seems to be heading more in the direction favoring the journals than
> the eprint archives, contrary to your prediction.

This is an interesting empirical question. I leave it to my
comrades-at-arms to reply. But to the extent that the S/L/P archive is
still preferred because of citation linking, relief is on the way in

> sh> System I, the traditional system, is one in which the journal
> sh> literature is available only via S/L/P tolls. Systems II and III
> sh> are identical.
> Well, I tried to argue above they are not. Under system II all
> articles are necessarily freely available with all bells and whistles.
> Under system III some or even most articles are freely available
> with whatever the author attaches to them, but journals still
> provide an S/L/P version that is complete, authoritative, and
> in many ways different.

Systems II and III start out the same. The Self-Archive does not have the
bells and whistles, just the final refereed, accepted draft and figures.
Whether the quality control continues to be supported by S/L/P or is
instead paid for by page charges depends on whether there continues to
be an S/L/P market for it. That's an empirical question.

> But even if every physicist did put every published paper in, does
> that still make it free of barriers? Free of monetary barriers perhaps,
> but there are likely to be other barriers, particularly in time to
> find the right paper, and trust that it really is what you wanted to
> find. Since under III the free system is controlled by authors,
> all sorts of pieces could be missing or incorrect. Even with version
> control, any difference from the "authoritative, published" version
> is likely to make readers uncomfortable. Suppose the author gets
> the published page number wrong - a volume-page based look-up would not
> find the paper the reader wants, even though it is there. Suppose
> the author doesn't even insert the journal publishing information - how
> would a reader find the article? Suppose the same author has several
> versions of the same paper of different lengths with slightly different
> titles, none of which corresponds exactly with the published version?
> These things happen, and are one of the things that makes the raw
> archive under system III (the current system in physics) less useful
> than the S/L/P journal literature.

You seem to underestimate the power of character-string matching. But
the short answer here is that none of this worries me. Let's facilitate
self-archiving and then see what happens.

> And I think this IS an important point - the current publishing
> system makes significant allowances for the outsider while ensuring
> that ridiculous work does not get prominent play - I think at least
> one side of that balance is likely to be lost under the Harnadian
> system.

I don't see this at all. Besides, until further notice, the entire
current publishing system is a subset of the "Harnadian system."
So there are no outsiders.

> sh> If journals can scale down S/L/P so that it can survive in co-existence
> sh> with a free archive, without the need to switch cost-recovery to
> sh> page-charges, I will still be at peace in my grave. The free literature
> sh> is fee enough for me.
> Well good. I think the next step is for other publishers to accept the
> system as it has evolved in physics. I think this is already happening
> in mathematics, for example, and may soon spread to the bio-medical
> fields. It'll be interesting to watch this new era of competition evolve!
> Arthur (


Stevan Harnad
Professor of Cognitive Science
Department of Electronics and phone: +44 1703 592-582
Computer Science fax: +44 1703 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 - 19:45:32 GMT