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

From: Stevan Harnad <>
Date: Mon, 1 Sep 2003 17:59:26 +0100

On Wed, 13 Aug 2003, Samuel Trosow wrote:

> Stevan,
> I thought you might be interested in looking at a paper I'm writing on
> the Public Access to Science Act (Sabo Bill, H.R. 2613).
> I've posted a working draft on my website at:
> and will be updating it as changes are made. I would certainly value any
> feedback from you.
> Samuel E. Trosow
> Assistant Professor
> University of Western Ontario
> Faculty of Information & Media Studies
> Faculty of Law

Dear Sam,

Before I read your entire paper, I thought I would comment on your summary
(below). I think this will already anticipate the comments I would make
on your full-text. Then you can perhaps reply and also let me know which
parts of the full-text I would need to address more specifically:

My overall reaction (and this applies both to the Sabo Bill and to your
paper on it) is that the research access problem cannot be understood
purely on the basis of (1) whether or not the research is publicly funded
by tax-payer money, nor (2) whether or not the research publication is
freely accessible to the tax-paying public. It is about access to research
by and for researchers, for the sake of research progress, to the benefit
of all tax-payers. Nor does it concern all research-based writings (e.g.,
books, textbooks): only peer-reviewed journal articles. And it does
not require that the article's text be placed in the public domain: only
that its full-text should be publicly accessible for free on the web. That
is all the Bill needs to mandate. If it does that, it will provide
a great benefit for tax-payers and for the progress and productivity of
funded research.

(1) RESEARCHER ACCESS. The real issue is that *all* research (whether
funded or unfunded, whether health-related or not) that is published in
peer-reviewed journals is written to be read, used, applied, and cited
by *other researchers* in building further research on it. Anything that
blocks that access by any of its potential user-researchers worldwide is
blocking the impact and progress of the research itself, hence blocking
the intentions of its funders.

Free access to research for researcher-users should therefore be the basis
for the appeal for open access, rather than free access for the tax-paying
public. (Public access is a welcome thing too, but it is not the tax-paying
public that is interested in reading and using the research in the
planet's 24,000 peer-reviewed journals, but the research community
itself; it is the *benefits* of this open access to research, in terms
of research progress and applications, that the tax-paying public then
inherits from open access. Just an appeal for layman-access to specialized
research will not get very far once it is scrutinized closely and
thought through.)

in the Sabo bill is that it does not make any principled distinction
between the different ways that authors may report and use their research
findings. The specific target is (and should be) peer-reviewed journal
articles, because those are the writings that authors themselves have been
giving away all along: Researchers never sought royalty revenue from their
journal publications because those were written exclusively for research
impact: to be read, used, applied and cited by other researchers. It was
the impact from that usage and citation that was in turn rewarded (by
salary increases, promotions, tenure, further research funding, prizes,
prestige), because it constituted research productivity and progress.

But sometimes researchers wrote books, or textbooks, in addition to
journal articles, and from those writings they sought royalty revenues,
like any other author. (They never actually earned very much in the
way of royalties, but that is another matter.)

The point is that those books, unlike journal articles, were *not*
author give-aways, written only for research impact, and they were
hence in exactly the same category as other books written by authors,
whether or not based on funded research. Without making a *principled*
journal/book distinction (i.e., an author give-away vs. author
non-give-away distinction), the Sabo Bill is doomed to slide down the
slippery slope that makes it seem equally applicable to *everything*
a researcher ever writes that has been informed in *any way* by funded
research. (This slippery slope is definitely to be avoided, if the PASA
is to have any hope of success).

The principled distinction is clearly between (G) the category of publication
that, without exception, the author *chooses* to give away, seeking
only research impact -- namely, refereed journal (and conference)
articles -- and (for the time being) (N) every other form of scholarly
publication, whether or not it is based on or derived in some way from
funded research.

(3) RESEARCH TEXTS VS. RESEARCH CONTENTS: Related to the foregoing
point, it must not be forgotten that whereas the research funders can
lay some claim to the ownership of or a stake in the research *findings*
-- and hence can mandate that they are to be made public ("publish or
perish"), they cannot lay claim to the researcher's *writing* itself:
A researcher is paid to do research and to make his findings public, but
he is not a paid writer-for-hire. This distinction between the public
reporting of research *content* and the actual text of researchers'
writings is another weak point of the Sabo Bill as it now stands.

There is no need to run aground on this problem, however, for the
journal/book distinction salvages the spirit of the intended outcome.
Researchers *want* to give away their peer-reviewed journal writings,
even though they do not *have* to do so. And it is their *reason* for
wanting to do so -- namely, to maximize their research's impact by
maximizing other researchers' access to and usage of their writings --
that should be the core rationale of any policy or legislation.

PRIORITY, TEXT-INTEGRITY. I don't fully understand the notion of making
one's writing "public domain" instead of retaining copyright, but if
that puts either the text's authorship or the text's verbatim integrity
at any risk -- i.e., if someone else could then legally reproduce my text
without my name as author, or even attaching his own name, or could
reproduce my text in an altered form, with or without my name -- then it
is certain that researchers will not want that! It's one thing to give
away access to one's text for free online, for anyone and everyone to read
and to use (the *content* of the text, while quoting/citing/attributing
any actual *words* used from the text itself), and quite another thing
to renounce one's right to protect the integrity of one's text, or to
be fully credited with its authorship.

foregoing point, it is not at all clear that any great change in copyright
law or policy or in copyright transfer agreements is needed other than
to ensure free public full-text access online. This can be
accomplished within existing copyright or licensing agreements
with publishers, and requires only that the publisher formally
agree to allow the author to self-archive the full-text for free
access worldwide on the web (as 55% of journal publishers already do:
If the Bill instead requires authors to ask for *more* than is really
necessary to provide open access, it will merely provoke needless author
opposition to the very Bill that is meant to benefit them! Authors will
not want to be forced to publish only in the journals (if any) that
agree to renounce copyright transfer and to publish public-domain texts.

to the foregoing point, your own paper and the Sabo Bill both seem to be
focused mainly on only one of the two complementary means of gaining
open access: Publishing in an open-access journal, which provides
full-text access to all of its contents for free, is one way to gain
open access. But there exist only 500 such journals at the present time.
What about the articles in the other 23,500 peer-reviewed journals?
The only thing that the authors of articles in those journals require is
that those journals should support open-access self-archiving in the
authors' own institutional open-access archives, as 55% of journals
sampled already do officially (and most of the others will do, if asked).

worldwide research's and researchers' rationale for open access (to their
give-away, peer-reviewed journal research -- 24,000 journals' worth,
over 3 million articles annually) is research impact. It is through
impact (uptake, use in further research, citation) that research
progresses; and it is for impact that researchers are rewarded. Hence
it is in all researchers' (and all research's) interest to maximize
its impact by maximizing access to it. Although government is only in a
position to influence funded research, it is important to place their
rationale for mandating open access for this research in the broader
context of open access to all refereed research, whether funded or
unfunded, and whether scientific or scholarly.

And just as open access is not just about funded scientific research, it
it is not just about US research output, or about US access to research
output. There is a global reciprocity in research on the planet that
knows no borders (just as the Internet knows no borders). Research is
cumulative and collaborative, and all researchers need access to one
another's research output so they can keep building on it. The Golden
Rule prevails (and the alternative is the Prisoner's Dilemma!).

The Sabo Bill should accordingly also be placed clearly in this context
of global reciprocity for all research. (The core intuition, based on US
tax-payer access to US-funded health research is in this respect both
parochial and misleading, and will generate obvious contradictions that
will undo the whole spirit of the proposal. Free research access for
researchers, for the benefit of all tax-payers, should be the broader
rationale, with lay access to health information only an important
special case.)

I now quote/comment your paper's summary, on the basis of the above:

> Should works resulting from research that has been substantially
> subsidized by the United States Federal Government be protected by
> copyright or immediately enter the public domain

This mixes means and ends: The goal is not that the "works" (which?
articles? books?) should lose copyright protection and enter the public
domain, but that the full-texts of refereed-research articles should be
accessible for free to all would-be users (world-wide). And although the
US can only mandate this for its own funded research output, the target
is all of research, whether funded or unfunded, as reported in all texts
that the researcher elects to give away (as all refereed-journal authors

This does not require renouncing copyright protection or copyright
transfer to publishers; it only requires a guarantee of the author's
right to make the full-text immediately accessible online for free, for
all would-be users worldwide, forever.

> The Public Access
> to Science Act PASA would place these works in the public domain in
> the same manner as works prepared by government employees

Works by government employees are presumably works for hire. The works of
university researchers (whether US-funded or not) are not. Researchers
are (for example) free to publish books based on their research, and
free to seek royalty from the sale of those books. But, without
exception, researchers do not seek royalty revenue from their
refereed-journal articles. They seek only research impact -- from their
uptake and usage. They do, however, still seek credit for those
writings, and for that, their authorship, and the integrity of their
texts, must continue to be protected. It is not clear that putting them
in the public domain would provide that protection; hence anything that
threatened to force them to renounce that protection would only arouse
needless opposition from the very authors that it was meant to benefit:
needless, because all they really need is the right to make their
full-texts freely accessible online. No need either to renounce
copyright or to renounce the journal publishers that refuse to publish
public-domain texts.

> This paper
> evaluates the merits of the legislation by placing the question
> of the appropriate copyright treatment of federally subsidized
> works within an historical perspective taking into account the
> underlying purposes of copyright policy as well as the changes that
> have taken place in field of scholarly publishing since Congress
> last considered the issue in 1976

It is not at all clear that a change in copyright law is needed if open
access is the objective. What is needed is at most a minor change in
copyright transfer policy (one already made by 55% of journals sampled):
the guarantee of the author's right to make his own full-text publicly
accessible online for free.

> The regulatory environment and
> practices of the major federal funding agencies are reviewed showing
> that agencies have failed to utilize their broad discretion over
> the treatment of funded works resulting in an over protection of
> works as the default rule

True. But funding agencies cannot treat all writings by funded
researchers that are in any way based on funded research as if they were
works-for-hire paid for by the funding agency. The university researcher
is not a US government employee. The funding agency *can* however,
mandate, that the full text of all peer-reviewed journal publications
reporting the findings of funded research must be made freely accessible
online (either by publishing them in an open-access journal, when a
suitable one exists [500 journals to date], or [for articles in the
other 23,500 journals] by self-archiving them in an open-access archive).

The slippery slope of books and other derivative works should be
explicitly avoided. This is about refereed journal (and conference)
articles only.

> The paper considers how the balancing
> of interests that has historically informed copyright policy
> should be applied to works that have been federally supported

Federally supported research does not mean writing-works-for-hire. The
funders can mandate publication of research findings, can even mandate
that the full-text of the publications must be accessible for free *if the
author elects to publish royalty-free (as all refereed-journal authors
do)*, but they cannot mandate that authors who wish to sell their texts
should not be allowed to do so.

These fine distinctions all become moot and trivial, however, if the
target is clearly and explicitly stipulated in advance to be *refereed
journal articles only*, and if the only objective is that the full-text
must be made immediately (and permanently) accessible for free online
upon publication in the refereed journal.

> it will review and assess the initial reactions to PASA from the
> point of view of various stakeholders including the commercial
> publishers non commercial publishers universities authors and
> researchers and library associations

Refereed-journal publishers will accept (and most already do accept)
author retention of open-access self-archiving rights. If that is the
sole point of discussion, agreement will be reached quickly. But if the
waters are muddied with needless "public domain" constraints that are
unnecessary for the goal of open access, then only further confusion,
confrontation, and little progress will result.

> and it will consider whether
> the PASA's purposes might be accomplished through other mechanisms

They can be: via simple retention of their open-access self-archiving
right by refereed-journal authors, as reinforced by research-funders'
mandating the provision of free access to the full-text of
refereed-journal articles reporting the research. No need for public

> The conclusion is reached that works resulting from extramural
> research that has been substantially subsidized by the Federal
> Government should enter the public domain in the same manner as
> works resulting from intramural government research undertaken
> by federal employees and that PASA provides a straightforward
> mechanism for reaching this result.

The conclusion goes far beyond the call for free online full-text access
to refereed research, and I believe it would generate needless confusion
and resistance. Writings based on funded extramural research are not works
for hire, and cannot be treated as such. Publication in peer reviewed
journals can be mandated. And free full-text online access to those
publications can be mandated. But that does not require or constitute
treating those texts as works-for-hire, destined for the public domain.

It would also give this initiative more credibility and scope if it were
portrayed (as it indeed is) not merely as a US funded-science issue,
but as part of a initiative to maximize research impact by maximizing
research access, worldwide, for all refereed research, scientific or
scholarly, funded or unfunded. The PASA would then be seen as one of
several means of accessing this global goal.

Stevan Harnad

NOTE: A complete archive of the ongoing discussion of providing open
access to the peer-reviewed research literature online is available at
the American Scientist September Forum (98 & 99 & 00 & 01 & 02 & 03):

Discussion can be posted to:
Received on Mon Sep 01 2003 - 17:59:26 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:02 GMT