Re: The UK report, press coverage, and the Green and Gold Roads to Open Access

From: Stevan Harnad <>
Date: Fri, 6 Aug 2004 00:43:23 +0100

On Thu, 5 Aug 2004, Dee Ann Divis wrote:


> Thank you for the information. I will take it into consideration as I
> put together the rest of the series.
> I do not agree with all of your points however and stand by the article
> as reported. For example, I have the actual House language and it is
> quite specific to PubMed, not to general self-archiving.

Dee, apart from the usual error (and it really is an error! but a
very wide-spread one, so it is not surprising that you fell into it
too!) of mixing up the House's OA Self-Archiving Mandate (green) with
an OA Journal-Publishing Mandate (gold), your article was accurate in
its reporting of the reactions of others (and it was mostly those other
reactions on which I was commenting).

But let me straighten out a second misunderstanding:

"Self-archiving" means the depositing of an article in an Open Access
(OA) Archive by its author. There are two forms of author self-archiving:
(1) central, subject-based self-archiving, and (2) distributed, institution-based

The fact that the recommended house mandate specifies PubMed Central simply
indicates that it is central self-archiving they have in mind. But it
is still *self*-archiving either way!

The *self* in self-archiving is the *author*. The US Government, in
making conditions for receiving research funding, is in a position to impose
conditions on the what the researcher, the recipient of the research
grant, must do. But it is in no position to make it part of the
condition for receiving research funding what it is that *publishers*
must do! It is, after all, not the publisher who is getting the money,
but the researcher!

So you are quite right that the US House recommendation was also specific
about *where* the author should self-archive (PubMed), but that is still
*self*-archiving, and my comment was simply referring to self-archiving.

I think there might have been some latent confusion too, in the article
-- over and above the confusion about whether it was OA Publishing
that was being mandated or OA Self-Archiving -- about *who* was being
mandated to provide the OA: It is the *author*, not the *publisher*
who must deposit the article in PubMed! When the author makes his own
article OA for himself, rather than the publisher doing it, that is OA
Self-Archiving (green). When the Publisher does it, it is OA Publishing
(gold). The House was recommending that *green* be mandated, not *gold.*

Could the House have recommended mandating gold? i.e., mandating that as
a condition of receiving research funding, all grant recipients must not
only publish their findings in a peer-reviewed journal (as is already
mandated now: "Publish or Perish"), but that that journal must be an
OA Journal?

The House *could* have recommended mandating that, but it wisely did
not, for the reason that there exist 24,000 peer-reviewed journals today,
but only about 1200 of them are OA Journals. That is about 5%. The
House recommendation would have had little chance of being passed by
the Senate if it had imposed as unrealistic a condition as that.

Such a condition would have been unrealistic not just because it would
try to squeeze 100% of research in 5% of the available journals, but
unrealistic also because whereas researchers don't mind being told to
publish, and don't mind being told to self-archive their published
articles (see below), they definitely *do* mind being told *where*
to publish them (and so they should!):

It is researchers themselves who must decide which journal, with what
level of peer-review standards and track-record for quality they choose
to submit their work to, not their granting agencies!

So the House was wise not to try to mandate OA Publishing (gold), but only
OA Provision, via self-archiving (green). We already have the indications
that grant recipients will willingly comply with such a mandate from
Swan and Brown's (2004) JISC/OSI survey, in which they asked authors:

    "how they would feel if their employer or funding body required
    them to deposit copies of their published articles in... [OA
    archives]. The vast majority... said they would do so willingly..."

    Swan, A. & Brown, S.N. (2004) JISC/OSI Journal Authors Survey

    Swan, A. & Brown, S.N. (2004) Authors and open access
    publishing. Learned Publishing 2004:17(3) 219-224.

One last point, about central self-archiving as opposed to institutional

I think the House Appropriations Committee was less wise in going on
to specify *where* grantees should self-archive their articles to make
them OA (in PubMed). Surely it is enough to mandate that they should
be made OA! For reasons discussed in an early posting in the American
Scientist Open Access Forum (reproduced below), it no longer makes any
difference where an article is self-archived, as long as the Archive is
OAI-compliant. In this regard, the recommendations of the UK Parliamentary
Science and Technology Committee
which were released within a fortnight of the US recommendations were
wiser (though otherwise very similar). The UK did not stipulate that funded
research must be self-archived in a central OA Archive, only that it
must be self-archived, hence OA. (In fact, they expressed a preference for
Institutional Self-Archiving.)

Below is the discussion of the pro's and con's of central and institutional
self-archiving, and how the US and UK recommendations differed on this.

For a discussion of the related Sabo Bill (which will fail for having recommended that
research articles be made *public domain* instead of recommending, as did the House and
the UK Parliament, that they merely be self-archived), see:

    "Public Access to Science Act (Sabo Bill, H.R. 2613)"

The remedy, of course, is to join the Sabo Bill with the Senate version of the House
Appropriations Bill, and this time do it right: Just mandate OA Self-Archiving!

Best wishes,
Stevan Harnad
    "Re: Mandating OA around the corner?"


Apart from the fact that the US recommendation only applies to biomedical
research and the UK recommendation applies to all research, there are
only two substantive differences between the US and UK recommendations:

    (US) The US House of Representatives Appropriations Committee
    recommends that the National Institutes of Health (NIH) provide
    free public access to research articles resulting from NIH-funded
    research :


    (UK) The UK Government Science and Technology Committee's recommends
    that all UK higher education institutions establish institutional
    repositories on which their published output can be stored and from
    which it can be read, free of charge, online

The two differences are:

    (1) The UK Select Committee is recommending immediate self-archiving
    of all funded research (within a month of publication), whereas the
    US House Appropriations Committee is recommending self-archiving
    after six months.

    (2) The UK is recommending self-archiving in the author's
    own institutional OA Archive, whereas the US is recommending
    self-archiving in in a central OA Archive (PubMed Central).

On the one hand, the differences are not important enough to bother
too much about. On the other hand, if something is worth doing, it's
worth doing optimally, and there are 2 simple ways to optimize the
US/NIH policy:

(1) SIX-MONTH-TIME-LIMIT: The wording of the NIH mandate should be that
all research must be self-archived "by 6 months after publication at
the latest." Nothing further should be specified explicitly about the
purpose of the 6-month delay. It is not a good idea at all to state
that the delay is needed in order to allow journals to make their
sales. The effect is the same either way. Allowing up to 6 months will
have the intended effect, and will calm publishers' fears, and their
motivation to resist the US recommendation. But stating explicitly that
the purpose of the 6-month delay is to allow journals to survive and
to make their sales is speculative, unnecessary, and establishes an
extremely bad precedent, one that would then be even harder to uproot

What NIH needs to bear clearly in mind is that 84% of journals already
give their official green light to *immediate* self-archiving (with no
6-month embargo)! See:

So if NIH were to make an explicit pre-emptive point that implied that
journals for some reason *need* to impose a 6-month embargo in order to
survive -- when in reality they do not, and 84% have already realized
this! -- then NIH would be needlessly inviting a big step backward for OA!

The up-to-6-months condition option is fine for now, and is just what
is needed to encourage the remaining 16% of journals that are not yet
green; but explicitly portraying this as a *necessary* delay, for the sake
of journal sales and survival, would be a big strategic mistake.

(2) INSTITUTIONAL VS. CENTRAL SELF-ARCHIVING: The self-archiving mandate
should be generic: All NIH-funded articles must be self-archived within 6
months of publication. That's all that's needed. Don't stipulate that the
self-archiving must be central, in PubMed Central (PMC)! NIH can mention
PMC as one of the options for self-archiving, but don't stipulate it; and
mention the option of self-archiving it in the author's own institution's
OAI-compliant OA Archive -- preferably as the preferred option.

The three reasons why preferential institutional self-archiving is the
optimal strategy are:

    (i) Given the existence since 1999 of the OAI archive-interoperability
    standard we are now in the age of distributed digital archives,
    all made interoperable by compliance with the OAI convention.

The reason the OAI standard was created was precisely this: So that
articles could be archived in any OAI-compliant OA archive on the web,
and then they would all be interoperable with one another, which means
they could all be harvested into one global "virtual" archive, they
could all be jointly searched, indexed, etc., by multiple harvesters and
searchers, including those specialized only for the OAI literature -- such
as OAIster and the citation-based
OAI engine, Citebase --
as well as all other search engines (including google).

PMC is itself OAI-compliant:
But PMC is only one of a growing number of OAI-compliant OA Archives:

There is no reason to constrain the NIH-mandated self-archiving to
PMC: What should be mandated is self-archiving in an OAI-compliant OA
Archive. PMC can be given as an example of a central OA archive, but the
default option should be the author's own institutional OAI-compliant
OA archive, with PMC recommended only as a back-up option, in case the
author's institution does not yet have an OAI-compliant OA Archive.

The reason institutional OA Archives should be given priority is
explained in (ii) below. Here let me just add that no functionality
whatsoever is lost by having the biomedical OA literature distributed
across institutional and central OAI-compliant OA Archives rather than all
being self-archived in one central archive (e.g., PMC). On the contrary,
much more functionality is gained. This is because of the distributed
searching, harvesting and indexing capabilities that are provided by
OAI-compliance. It would be a handicap, not a benefit, to have all of
the OA biomedical literature in PMC only. The idea that it would be either
necessary or preferable to have it all in one place is just a hold-over
from obsolete paper-based thinking! That is not the nature of the digital
medium or its many newfound powers.

On the contrary, the best policy is for authors to self-archive in their
own OAI-compliant OA archives and for PMC to harvest the metadata, so
PMC too indexes those articles! There is no reason whatever for their
full-texts to be physically deposited in PMC. Offering central full-text
self-archiving in PMC is a good back-up option for those authors who do
not yet have an institutional OA Archive, but it is a bad and restrictive
option if it is the only option, or the primary one, that is mandated.

    (ii) The reason institutional OA self-archiving should be made the
    preferred option for the NIH self-archiving mandate is that that is
    the way to generalize the effects of NIH policy on OA in general,
    across disciplines and institutions.

If the NIH self-archiving mandate is made specific to self-archiving in
PMC, then those authors reporting NIH-funded research will self-archive
in PMC, and that will be all. When they do non-NIH-funded research, or
if they are not in the biomedical sciences, the NIH mandate will have
minimal influence on them, and on whether OA grows beyond NIH-funded
biomedical research.

If, on the other hand, it is institutional self-archiving that is mandated
by NIH, then the NIH influence is most likely to propagate beyond just the
boundaries of NIH and PMC and biomedical research to all disciplines,
at all institutions. Much will be gained for OA to research in general,
and nothing will be lost -- and a good deal gained also for NIH-funded
biomedical research in particular.

    (iii) Another reason central self-archiving should not be specifically
    stipulated or preferred is that a number of the 84% of journals that have
    given their green light to self-archiving have specified that this
    must be institutional self-archiving, not 3rd-part central archiving.

In the OA and OAI world, once an article's full-text is OAI-compliant and
openly accessible (OA) anywhere on the web, it is in fact *irrelevant*
where it is actually located physically -- whether at the author's
own institution site or a central site -- but publishers are still
understandably leery about having their green light to *self*-archive be
interpretable as a legal green light for 3rd-party free-rider publishers,
re-publishing their content and perhaps even re-selling it.

As I say, this worry is wrong-headed, because once a paper is accessible
free for all, there is not much to be gained by anyone's trying to
republish or re-sell it. But here again, as with the 6-month delay,
it is better not to add needless specifications or restrictions that
will needlessly encourage publishers to want to fight, and perhaps even
to withhold their green light to self-archive.

Just mandate that all NIH-funded research must be OA self-archived within
6 months of publication, preferably in the author's own institutional OA
Archive, otherwise in PMC, but that's all! Don't give a rationale for
the 6 months and don't stipulate that the self-archiving must be central

The UK recommendations are spot-on already, hence require no further
optimization, just implementation!

Stevan Harnad

UNIVERSITIES: If you have adopted or plan to adopt an institutional
policy of providing Open Access to your own research article output,
please describe your policy at:

    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.

A complete Hypermail archive of the ongoing discussion of providing
open access to the peer-reviewed research literature online (1998-2004)
is available at:
        To join the Forum:
        Post discussion to:
Received on Fri Aug 06 2004 - 00:43:23 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:32 GMT