The Green and Gold Roads to Open Access

Stevan Harnad harnad at
Tue Mar 23 09:59:06 EST 2004

On Tue, 23 Mar 2004, Jan Velterop wrote, under the subject thread
"What is Open Access?":

> I fully agree with David Goodman that clarity of terminology is needed

Jan Velterop's terminology is welcome, but please not that the
subcategories he introduces below are all merely subdivisions among
articles that are OA because they have been published in OA Journals
(i.e., gold). The terms and categories do not apply at all to articles
(published in TA journals) that are OA because they have been
self-archived by their authors (i.e., green).

>   At BioMed Central we speak of two categories of articles:
> -'Research articles' ('primary' articles, 'fact-oriented') and
> -'Other articles' (which includes reviews, comments, and other
> 'opinion-oriented' articles as well as news and the like)

These are BMC categories, rather than OA categories. The only OA category
is whether the article has or has not appeared in a peer-reviewed
journal, and whether the OA version is the pre-refereeing preprint or
the peer-reviewed postprint.

> 'Open Access'
> This DOES apply to ALL the Research articles; it MAY apply to Other articles
> -- if they have somehow been paid for, e.g. by sponsorship or a subvention.

Again, these distinctions do not seem to be OA distinctions but BMC distinctions.

> Otherwise they are...
> 'Subscription-Paid'.
> Much of the 'added-value' for these articles either comes from, or is made
> possible, by the publisher, there is choice (no 'publish-or-perish pressure;
> publishing comments or review articles is rarely a 'must'), and subscription
> charges or sponsorship, or a combination of those, are sensible ways of
> recovering the costs).

This is not quite clear. Is this stating that review articles in BMC are not OA
but must be paid for by the subscriber?

Or is this not about BMC, but a general observation about review articles
in other journals? In such cases I would say that it is up to the author
whether he wishes to provide OA for his article -- unless, of course, the
author has been paid by the journal to write the article as a work-for-hire,
in which case I agree of course that it is *not* a candidate for being OA:
The OA is movement applies only to author give-away articles:

    "The literature that should be freely accessible online is that
    which scholars give to the world without expectation of payment."

> 'Input-Paid'
> We avoid the term 'author-paid', because that is rarely true, and it is
> certainly not expected that authors pay from their own pocket. We speak of
> 'input-paid' (suggestions for a better term are always welcome) and we
> regard...

This is again a subcategory of OA publishing (gold) not of OA itself.

> 'Article Processing Fees'
> (NOT 'author-fees') as an integral part of the research expenditure, as
> publishing the results is necessary and integral to a research project.

Again an OA publishing category. Call it whatever one will, these fees
are what must be paid to the OA journal to cover its costs. 

There is no agreement yet on how high those costs are, or what they
should cover. According to some OA publishing models, they should
cover all the current "values-added" of TA journals, with the possible
exception of the costs of printing and distributing the print-on-paper
edition. According to other OA publishing models (including the one I
would favour, if/when OA publishing should ever prevail), they
should cover only the cost of implementing peer-review. The task of
text-generation and mark-up can instead be offloaded on the author and
the cost of (online) archiving and access provision can be offloaded on
the author's institution.

    "Distinguishing the Essentials from the Optional Add-Ons"

> 'Deposit'
> is what we do with all the Open Access articles immediately, in several
> trusted Open Access Repositories, such as PubMed Central (and INIST in
> France, and Potsdam University in Germany). 'Other' articles are deposited
> and 'opened' after two years.

This is a confusing and even somewhat contentious matter, again as between the
meaning of the terms in OA publishing in particular (gold), OA self-archiving
in particular (green), and OA in general.

First, we must note that for an article to be OA at all, it must
necessarily "deposited" somewhere online, so as to fulfill the defining
criterion of OA: that it must be accessible to everyone toll-free online!

Now in the case of articles being made OA by being published in a TA journal and
then self-archived online by their own authors, the word "deposit" is synonymous
with the word "self-archive," and it is unambiguous that it is the author who
does the depositing (or his institution's digital librarian does it for him),
either on his own website or (preferably) in his institution's OAI compliant
Eprint Archives or in a central OA Eprint Archive.

With OA journals, the way OA is provided is superficially similar -- the
article may be deposited in the Journal's own OA Archive, or a central
OA Archive such as PubMed Central, or even (why not?) in the author's
own institutional OAI-compliant (hence interoperable) Eprint Archive.

Now comes the contentious part: The (necessary) archiving component of OA
Publishing is *not* what is meant by archiving in the case of OA self-archiving
(even though, as noted, the archives involved might even be the same ones!).

For OA publishing, archiving (not self-archiving: archiving/depositing *by the
journal*) is essential to making the OA Journal OA. Unless the journal sees
to it that the article is deposited in some OA Archive, the Journal cannot lay
claim to being an OA Journal at all!

(At least so it is in OA Journal Publishing today: If my own hunch proves
right, in future OA Journals will only be peer-review service-providers,
and they will offload all archiving and access-provision onto their authors
and their institutions. But that is the future. Right now, author
self-archiving is *not* what makes articles in OA Journals OA!)

But now let us consider the other (green) road to OA: Here it is the author and
his institution who provide OA for their *TA* articles by self-archiving them.

Considerable confusion and misunderstanding has been caused not only by
(1) equating OA with OA publishing alone (leaving out OA self-archiving,
which already provides considerably more OA yearly than OA publishing
does yearly at this time) but by (2) describing the role of archiving
in OA as merely being that of depositing an OA journal article in an OA
Archive -- leaving out the all-important function of archiving as the
OA self-archiving of TA articles by their authors.

This effectively gives the extremely misleading and incorrect impression
that the role of archiving in OA is merely to serve as the preservation
and access-provision wing of OA Publishing!

The scope and function of OA provision via self-archiving differs
profoundly from the scope and function of OA provision via OA publishing
-- and the depositing of those OA journal articles in OA Archives! From
the Bethesda Statement right through to the Berlin Declaration, OA has
been incorrectly equated with OA publishing alone, and "archiving" has
been incorrectly equated with the depositing of the OA Journal article
in an OA Archive. It is extremely important to correct both of these
errors, for the sake of OA.

As to "delayed deposits" (months or years after publication): though
of course better than nothing, those definitely fail to meet the BOAI
definition of OA, which is "immediate, permanent, toll-free, full-text
online access."

    "Shulenburger on open access: so NEAR and yet so far"

    "Needless Pruning of Research's Growth-Tip"

    Harnad, S. (2001) AAAS's Response: Too Little, Too Late.
    Science dEbates [online] 2 April 2001.

> 'Preserve'
> is what the Dutch Royal Library (Koninklijke Bibliotheek, KB) does for us.
> They have committed to 'preserve' the content in case format changes should
> be neccessary in the future. The KB archives and preserves material from
> other publishers, such as Elsevier and Kluwer, as well, but in contrast to
> that material, the BioMed Central articles and journals are not bound by any
> restrictive contracts and freely available from the KB.

Preservation is again largely an OA Publishing issue. It is not that OA
self-archiving is not also committed to keeping the OA version accessible
permanently (and the continued OA status of, for example, the physics
articles self-archived in ArXiv in 1991 -- still fully accessible today,
13 years later, and even successfully retro-fitted for OAI-compliance
in 1999 -- is compelling evidence of that). But otherwise, preservation
today, for both TA journal articles and OA Journal articles, is the
responsibility of the journals and the libraries that subscribe to
them. That is the *primary* preservation burden, and because OA Journals
are Journals, they too must share it.

OA self-archiving does not have this primary preservation burden, because,
being the self-archiving of TA journal articles, it is the TA journals
(and the libraries subscribing to them) that still bear that primary
responsibility. The self-archived OA versions of TA articles are merely
*supplements* to the primary corpus, not *substitutes* for it:

Of course (1) if/when OA ever prevailed, and (2) if/when OA Publishing (gold) also
prevailed, and then then (3) if/when my own prediction that archiving
and access-provision will be offloaded onto the distributed network of
Institutional OAI Eprint Archives should prove true, then -- and only then
-- will those archives also have to take over the primary preservation
burden too, and not just the secondary, supplementary one they bear now.

But OA has not yet prevailed. And there is a good deal more OA being
provided via OA self-archiving (green) than via OA publishing (gold)
today. So, again, the preservation terminology and considerations are
right now only for OA Publishers like BMC, not for OA in general.

> We host the material ourselves on the BioMed Central platform and, in
> addition, the Repositories mentioned above function as mirrors.

And that is something that OA Publishers need to do, if they are to be OA at all.
The meaning of "archiving" for the OA versions of TA articles self-archived
by their authors is very different. 

Hence none of these otherwise welcome terminological and functional details from
BMC answer the general question "What is Open Access?" but only the unilateral
question "What is Open Access Publishing?".

But there has already been (in my opinion) too much unilateral focus on OA
Publishing in the past year and a half. It is time to redress this balance
for the sake of OA itself, and to keep it balanced, between the bilateral
OA poles of OA Journal Publishing (gold) and OA Self-Archiving (green) --
*especially* given that by any objective measure, the actual weight of OA
is so heavily tilted toward green, and the potential weight even moreso.

Stevan Harnad

NOTE: A complete archive of the ongoing discussion of providing open
access to the peer-reviewed research literature online (1998-2004)
is available at the American Scientist Open Access Forum:
        To join the Forum:
        Post discussion to: 
    american-scientist-open-access-forum at
        Hypermail Archive:

Unified Dual Open-Access-Provision Policy:
    BOAI-2 ("gold"): Publish your article in a suitable open-access
            journal whenever one exists.
    BOAI-1 ("green"): Otherwise, publish your article in a suitable
            toll-access journal and also self-archive it.

More information about the Jrnlnote mailing list