Re: Free Access vs. Open Access

From: Stevan Harnad <>
Date: Mon, 16 May 2005 21:38:30 +0100

     Prior AmSci Topic Thread:
     "Free Access vs. Open Access" (began August, 2003)


        Stevan Harnad

For those without the time to work through the details, the punch-line is

    What research and researchers need, now, is toll-free, immediate,
    permanent, webwide, online access to the full-text of all 2.5 million
    articles published annually in the world's 24,000 peer-reviewed

That is is what is (or ought to be) meant by "Open Access" (OA) (and
it is certainly all I've ever meant by it -- and I've been at it a lot
longer than the day some of us coined that term for the "BOAI" after
Budapest in 2001!).

Librarians have been concerned about journal pricing, permissions and
preservation (including digital preservation) -- the 3 P's -- for a
long time too. But those P concerns had *nothing* whatsoever to do with OA
before, and they still have nothing to do with OA now -- except that, if
willfully *conflated* with OA, they can help embargo OA for yet another

(One of the many ways to willfully conflate the P's with OA is what
Charles has proposed: to co-opt the original meaning of a "Green"
journal -- a code coined to refer to journals that give their Green
Light to author self-archiving -- and use it instead to mean "Open
Access Journals," for which the already proposed and used color code
had been "Gold". This effectively washes out self-archiving, and instead
re-focusses everything on the three P's.)

As I said, one cannot legislate either tastes or colors, so I shall simply
try to show where and how both Charles Bailey and the library community
are missing the point and blocking progress on OA if they try to force
OA into their Procrustean bed of P's, thereby unwittingly helping to
delay and deny to the research user community the access to the content
they so urgently need today, to delay and deny to the research author
community the usage and impact that they are now needlessly losing, daily,
monthly, yearly, and to delay and deny to research itself the potential
productivity and progress that is all being still-born currently,
owing to the absence of toll-free, immediate, permanent, webwide, online
access to all article full-texts. And all this in the name of the library
community's all-important three P's: Pricing, Permissions, Preservation.

On Mon, 16 May 2005, Charles W. Bailey, Jr. wrote:

> Stevan Harnad has commented extensively on my "The Spectrum
> of E-Journal Access Policies: Open to Restricted Access"
> DigitalKoans posting.
> First, let me concede that if you look at this question from
> Stevan's particular open-access-centric point of view that,
> of course, the spectrum of publisher access policies is a
> complete and utter waste of time.

Translation: If you are concerned with OA instead of the 3 P's, the 3 P's
will look like a waste of time: No, the 3 P's are not a waste of time.
They are just *irrelevant to OA* and should cease to be conflated with
it, to OA's cost.

> I don't recall suggesting
> that this was a new open access model per se, even though it
> includes open access in it as a component and it makes some
> further distinctions between open access and free access
> journals.

Charles did not propose it as a new OA model; he simply chose to code
journals in terms of their "degree of OA" and to use a highly complicated,
particolored journal code that not only clashed with the simple two-color
journal code -- which had been designed specifically to claify things
and bring them into focus (OA Journals: Gold; Journals endorsing OA
self-archiving by authors: Green) -- but reassigned "Green" altogether, to
OA Journals, threw in a dizzying bunch of other color-coded distinctions
irrelevant to OA, and threw out the most fundamental distinction of all,
pertaining to self-archiving! And revived the bogus "*OA/FA" distinction
to boot.

Quick reminder: What distinguishes this notional "*OA" from OA (= FA)
is the following:

(1) Creative Commons License (irrelevant if the full-text is accessible
toll-free to all users, webwide, immediately and permanently).

(2) Republication/Re-Use Rights (irrelevant if the full-text is accessible
toll-free to all users, webwide, immediately and permanently). (OA
is about user access -- which includes reading online, downloading,
data-crunching, and print-off, and de facto also trawler-harvesting; it is
not about the right to re-publish, re-sell, or repackage in a database.)

(3) Redistribution Rights (irrelevant if the full-text is accessible
toll-free to all users, webwide, immediately and permanently). (OA is
about *online* access, not on-paper access: redistribute the URL instead
of the paper!)

"*OA" would be FA + (1) + (2) + (3),

whereas OA = FA. And that's all it was ever meant or needed to be.

> Rather, it is what it says it is: a model that
> presents a range of publisher access policies from the least
> restrictive to the most restrictive. The color codes merely
> enhance the model slightly, they are not central to it (and,
> of course, as Stevan says, he created this color coding
> Frankenstein to begin with). The model says nothing about
> e-prints.

(i) I must remind Charles that the name and theme of his blog is:
DigitalKoans: What Is the Sound of One E-Print Downloading?

(ii) When we have one color code calling a journal GREEN if it allows
self-archiving and GOLD if it is an OA journal, and another color code
calling a journal GREEN if it is an OA journal, we have a color clash
and a code war whatever "model" we may have in mind (especially with
the new code calling the old one "Frankenstein"!).

> That said, Steven's view that open access equals free access
> (period) is not, as he well knows, universal, and his green
> and gold models are based on this premise.

I have stated exactly why the OA/FA distinction is bogus. To move
forward now, we don't need a nose-count but reasons.

> Here is how Peter Suber defines OA in "Open Access Overview:
> Focusing on Open Access to Peer-Reviewed Research Articles
> and Their Preprints":
> * OA should be immediate, rather than delayed, and
> OA should apply to the full-text, not just to abstracts or
> summaries.

So far identical ("immediate," "full-text').

> * OA removes price barriers (subscriptions,
> licensing fees, pay-per-view fees) and permission barriers
> (most copyright and licensing restrictions).

Still identical ("toll-free," "webwide"). If there are no permissions
barriers blocking your access to commercial ads, porn, blogs, or
other blarney full-texts that are accessible to you toll-free,
webwide, why would you imagine that there are any special permission
barriers in the case of journal articles that are accessible to you
toll-free, webwide?

> * There is some flexibility about which permission
> barriers to remove. For example, some OA providers permit
> commercial re-use and some do not.


> Some permit derivative works and some do not.


> But all of the major public
> definitions of OA agree that merely removing price barriers,
> or limiting permissible uses to "fair use" ("fair dealing"
> in the UK), is not enough.

Correct. There must also be toll-free, immediate full-text access, toll-free,
webwide, permanently. With that, there is no more of substance left to disagree

> * Here's how the Budapest Open Access Initiative put
> it: "There are many degrees and kinds of wider and easier
> access to this literature. By "open access" to this
> literature, we mean its free availability on the public
> internet, permitting any users to read, download,

So far, identical.

> copy, distribute, print,

Meaning what? Print-off locally. Fine, go ahead. Print multiple copies and
distribute or sell: not fine. But just distribute the URL instead! Same
effect. No problem. And again, much ado about nothing.

> search, or link to the full texts of these articles,

No problem.

> crawl them for indexing,

No problem (or at least as moot as for all other web content that is
not password-protected or fire-walled).

> pass them as data to software, or use them for any other lawful purpose,

No problem.

> without financial, legal, or technical barriers other than
> those inseparable from gaining access to the internet
> itself.

Exactly, fully covered by toll-free, webwide access for all
(and the green journal's copyright agreement).

> The only constraint on reproduction and
> distribution, and the only role for copyright in this
> domain, should be to give authors control over the integrity
> of their work and the right to be properly acknowledged and
> cited."

For the online full-text -- which, to remind everyone, is the *only thing*
at issue with OA, which was a new possibility that arose with the online
age and does *not* change things for paper or paper-distribution -- this
is fine.

For the on-paper rights it is a bit misleading.

And of course authorship, attribution, and text-integrity are still protected,
as always, by copyright, for OA articles.

> * Here's how the Bethesda and Berlin statements put
> it: "For a work to be OA, the copyright holder must consent
> in advance to let users 'copy, use, distribute, transmit and
> display the work publicly and to make and distribute
> derivative works, in any digital medium for any responsible
> purpose, subject to proper attribution of authorship'"

This definition (Berlin simply quoted Bethesda verbatim) was formulated
during the "Gold Rush" when "OA" was being co-opted by OA journal
publishing (Gold) and ignoring OA self-archiving (Green) of non-OA
journal articles. It is therefore only a definition of an article published
in an OA journal, not a definition of OA.

The Gold Rush is now over, and we are now in the Green Era of Institutional
Repositories and Self-Archiving.

No "advance consent" to users for anything is needed for the author of a
published (Green) journal article to make his article OA by self-archiving
it (apart from the copyright agreement with the publisher, plus the act
of making the accessible toll-free webwide). "Public display" is probably
moot (display on-screen?); and derivative works (apart from the usual
quoting, using data, citing) is either irrelevant or downright incorrect
(and governed by whatever copyright agreement the author may have signed
with his Green publisher).

> * The Budapest (February 2002), Bethesda (June
> 2003), and Berlin (October 2003) definitions of "open
> access" are the most central and influential for the OA
> movement. Sometimes I call refer to them collectively, or to
> their common ground, as the BBB definition.

And in their substantive points they are all identical: OA = FA.

> So, by most OA definitions, a journal that "makes all of its
> articles immediately and permanently accessible to all
> would-be users webwide toll-free" is not OA unless it also
> uses a Creative Commons or similar license that permits use
> with minimal restrictions. It is FA (Free Access).

With all due respect, I completely disagree. Such a journal is 100%
OA. It may not be "*OA" (from the point of view of the three P's), and
this may be a matter of interest to some librarians, but it is totally
irrelevant to research and researchers. And after all, that is why these
articles are written and read.

> As I have said in an earlier dialog, we can count on no journal to be
> "permanently accessible" unless some trusted archive other
> than the publisher makes it so, an issue that Steven
> apparently disagrees with, believing that publishers never
> go out of business.

I have grown so weary with this theme that I cannot bring myself
to gather and link all the FAQs. Digital Preservation is a worthy issue;
it is being and will be taken care of. But it is not an OA issue. The
OA issue is getting from today's 15% OA to 100% OA, not to fuss about
the digital preservation of either OA or non-OA journals.

> I note that Steven has deviated from his "chrononomic
> parsimony" principle by having both "Green" and
> "Pale-Green," in his model and then lumping them both
> together in his discussions as "GREEN."

I haven't deviated in the least. There were always preprints and
postprints, and the two shades of green simply mark that pertinent fact.

> (In his Summary
> Statistics So Far site he also introduces the color Grey,
> for "neither yet.")

Every category needs a complement (i.e., the category of things that are
*not* in the first category). As mentioned before, Green and Gold are not
mutually exclusive: All Gold journals are also Green journals (but not
vice versa). But Green and Gold do not exhaust all the journals there
are. 8% are neither Gold nor Green. The original Romeo color code for
such journals was "white," but to avoid the white-washing connotations
of not being green, I substituted Gray: So what? The category is real,
non-empty, and so merited a name and color to tag it, and got one.

> If preprints and postprints are of equal
> value, why not just code them Green?

First, they are not of equal value: Preprints are unrefereed drafts,
postprints are peer-reviewed, accepted drafts. There has definitely been
"value added" by the peer review. Moreover, the specific target of the
OA movement is the postprints, not just, or primarily the preprints.

What is of just about equal value (but not equal convenience) is (1)
postprints and (2) preprints plus a list of the corrections that will
make them into postprints.

So journals that only give their green light to preprint self-archivng
are coded Green, but Pale-Green, to mark the (slight) inconvenience
that they cause their authors (and users) in trying to make them resort
to linking corrections instead of self-archiving their refereed final draft.
(In practice, they will turn out to be much the same thing.)

> If they are not of
> equal value (i.e., postprints that accurately incorporate
> the changes that occur during the peer-review process are
> the only real substitute for the published article), then,
> in reality, those 15.5% of "Pale-Green" journals are of
> limited value in terms of self-archiving, and the real GREEN
> journal number is 76.2%, not 92%.

Cut the cake any way you like. Today 15% are self-archived, and what
we need is 100%; but instead we are fussing here about the P's, and how
many are to be (1) postprints vs. (2) preprints plus corrections...

> I must admit to some confusion on his latest stand that all
> types of self-archiving are equal. In "Ten Years After," he
> seems to be expressing a different sentiment regarding
> author home pages:

No, I originally proposed anonymous ftp, superseded by personal
web-pages, superseded by OAI-compliant Institutional Archives.
All along, the idea was to self-archive all versions: preprint,
corrections, postprint, corrections.

> That said, there was a naive element to the Subversive
> Proposal, too, since Harnad's plan would have led to
> researchers posting their papers on thousands of isolated
> FTP sites. This would have meant that anyone wanting to
> access the papers would have needed prior knowledge of the
> papers's existence and the whereabouts of every relevant
> archive. They would then have had to search each archive
> separately. Today, Harnad concedes that "anonymous FTP sites
> and arbitrary Web sites are more like common graves, insofar
> as searching is concerned."

Correct, and I have prominently done my Mea Culpa's for that
technological shortfall (still infinitely better for would-be
users who would otherwise be access-denied) and updated to
OAI-compliance: but what is the point being made here, and
what has it to do with the preprint/correction/postprint issue?

> Perhaps I misunderstand what is meant by "arbitrary Web
> sites."

No you didn't misunderstand: It's non-OAI-compliant URLs -- the kind
of anonymous graves that Citeseer and Google Scholar have nevertheless
done a good job harvesting and sorting. But OAI-compliant Institutional
Repositories will help them, and OAIster will do that much better. What's
needed now is the OA content -- but instead we're here worrying about
the P's and the semiology of "OA" vs. "FA"!

> As the prior DigitalKoans dialog beginning with "How Green
> Is My Publisher?" shows, we clearly disagree on many points
> related to the importance of author copyright agreements
> (e.g., they have to permit deposit in disciplinary
> archives), the importance of deposit in OAI-PMH-compliant
> archives, and the mission and scope of institutional
> repositories.

More important, we disagree on their importance *for OA* (as opposed to their
importance for the P's; and we also disagree about the relevance of the P's
to OA, as well as on whether there is any substantive OA/FA distinction).

> A series of DigitalKoans postings that start with "The View
> from the IR Trenches, Part 1" provides numerous quotes from
> the literature that bolster my case.

The IR material that you review and quote from here is out of
date. (Things are moving fast.) Also, there are IRs and there are IRs. The
concern of the OA movement is OA IRs (self-archived journal articles,
their preprints and postprints) not IRs for all of the other digital
concerns an institution might have: institutional digital content
management, institutional digital content preservation, institutional
digital courseware, institutional digital publishing, institutional

None of these other kinds of IR uses (and their problems) should be
allowed to constrain or slow progress on OA-specific IRs.

> Second, while I admire Stevan's unflagging advocacy of open
> access (by which he really means free access), open access
> is not the only issue in the e-journal publishing world that
> is of concern to librarians to whom this missive was mainly
> addressed. This is because librarians, while hopefully
> working to build a better future, have to deal with the
> messy existing realities of the e-publishing environment to
> do their jobs and to make decisions about how to allocate
> scarce resources. Consequently, librarians have to scan the
> e-publishing environment, analyze it, categorize it, and
> make evaluative judgements about it. They have to make
> models of e-publishing reality to better understand it. They
> don't have the luxury of only dreaming about what that
> reality should be.

I could give a rather long list of concrete things (CogPrints, GNU
Eprints, Citebase, PsycPrints, OpCit, the OA impact studies, the journal
self-archiving policy registry, the institutional archives registry,
the institutional self-archiving policy registry, and all the policy
lobbying etc.) that couldn't be more remote from the oneiric, but what
Charles seems to be implying here is that to focus on OA instead of the
P's is to turn away from real work on everyday reality and dream: No, it is
real work on OA rather than on the P's. Perhaps it's not even library-related
work; perhaps the points of apparent commonality are illusory.

(I don't really think they are illusory, and some librarians have been
the greatest allies and pioneers in OA. But there always seems to be a
backwards tug, an undertow, from more traditional library concerns (viz,
the 3 P's and a few others), and when that tug gets the better of the
tug for OA, there is no doubt about where my own priorities lie.)

> Thus, while Steven is indifferent to many of those 894,302
> free full-text articles from 857 HighWire-hosted journals (a
> number which likely dwarfs all articles available from
> OA/free journals), librarians are not.

I am not indifferent to toll-free Back-Access, but it's clear *that's*
not OA. Research doesn't progress by being under a partial access embargo
for 6 months, or a year, or longer (any more than it progressed from
being restricted to those who can pay the access-tolls).

Moreover, whereas High-Wire journal Back-Access may dwarf OA (Gold)
journals, that certainly does not exhaust, let alone dwarf Green
(self-archived) OA. At 15% of the annual 2.5 million articles, that
means over 400,000 OA articles per year. High-Wire's, I take it,
is a total to date, not an annual figure?

And my concern is not with the 15% OA we have, but the 85% we haven't.
And the 92% it should already have been -- without any particular interest
in the 76.5%/15.2% breakdown for how much of the nonexistent OA content
will be postprints and how much preprints-plus-corrections: The point is
to get it into the IRs as soon as possible; and that's not facilitated by
over-painting the 92% journals that are Green with a spectrum of colors
that have absolutely nothing to do with their self-archiving policy! Nor
does it have anything to do with the curation of journals that eventually
provide free back-access.

> Paying attention to
> them is important. While many are not immediately free, they
> are free nonetheless after some embargo period. And EA
> (Embargoed Access) journals are better than RA (Restricted
> Access) journals in practical terms for users who have no
> other current access. And even limited access to more
> restrictive PA (Partial Access) journals is likely to be
> welcomed by users who today would have no access otherwise.
> I know that both kinds of access are welcomed by me as a
> user.

Yes, but you seem -- in this cornucopia of back-access options -- to
have forgotten OA entirely the fact that the cupboards are 85% bare,
whereas they could and should be at least 92% green! And your "users"
sound more like research historians and students (who don't use research
journals that much anyway) rather than the active researchers for whom
the annual 2.5 million articles are primarily written, and who need
them for their ongoing research *immediately* and not after an embargo
period that is (almost) as undesirable and every bit as unnecessary as
the toll-access that was the sole option in paper days.

> This is not to say that we shouldn't strive for journals to
> move up the spectrum from red to green,

By which you mean move from toll access journals (95%) up to open access
journals (5%) -- having entirely lost sight of the original green and
its vistas, with 92% of journals already green, hence 92% of articles
already potentially accessible online immediately, toll-free, if only
they are self-archived.

> but it is to say
> that: (1) some free access is better than no free access for
> journals that will never move further up the spectrum, and
> (2) it may be that some journals have to move step-by-step,
> not in one leap, for the change to take place, and, if they
> start higher, it may be easier to encourage them to move
> further and faster. (But we have to know which ones have
> this potential based on their current status.)

Because of your preoccupation with the 3 P's, you appear to have forgotten
about OA, and focussed exclusively on OA publishing (Gold!). But the
Gold Rush is over. The 5% Gold is not (and never was) going to rise to
100% within the foreseeable future on its own momentum. Whereas the 15%
Green is indeed poised to rise to at least 92% (if we don't obscure this
fact with a particolored brush that has nothing to do with OA). The
first awaits and depends on what what the publishing community elects
to do; the second awaits and depends only on what the research community
elects to do.

> Stevan's model has colors, but, in reality, each color is
> black and white: Gold and nothing, GREEN and grey. All or
> nothing.

I couldn't follow that! How did two more colors (black and white) get into
this, except in the sense of the two Artistotelean truth values T(rue)
and F(alse), which are indeed black or white: Something either is Gold
or it is not Gold. Something either is Green or it is not Green. (For
technical and logical reasons, as I said, Green and Gold are not mutually
exclusive, because all Gold journals are, a fortiori, Green. So that's
not B/W T/F: a journal can be both Green and Gold.) And Gray is just for
the rest, the complement: neither Green nor Gold.

That's the actual OA journal situation, and OA itself is
all-or-nothing (at the article level: clearly with "Open Choice"
there can be journals that are part Gold), and that is because of the
immediate/permanent/full-text clauses. If there were degrees of OA-ness,
then a journal that made its contents free after 100 years would be an
OA journal of a certain OA degree/color; so would a journal that made
its contents free for one day only; and a journal that made every fifth
page free. (We're not far from the slippery slope where a lower-priced
journal is more OA than a higher-priced journal, but by then we've
simply reduced OA to one of the P's! Maybe every subscription dollar
warrants a color -- perhaps different intensities of green!)

But this is all nonsense. This is not what researchers need; it's not what
authors want. And it is not what is really possible and within reach. It
is just a book-keeping taxonomy for those who are minding their P's rather
than taking their cues from research, researchers, and their real needs.

> And, as long as you accept his premises, it works,
> and it allows him to focus on his free-access goal with
> single minded determination, undistracted by the knotted
> complexities of the e-scholarly publishing environment. Long
> may he run.

Allows me to focus on research and researchers' needs, and the reachable goal
of 100% OA, rather than being distracted by the library community's worthy
but irrelevant (to OA) preoccupation with Pricing, Permissions and

Stevan Harnad
Received on Mon May 16 2005 - 21:38:30 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:53 GMT