Two members of STM have kindly, at my request, allowed me to see a
     copy of the STM Briefing on IRs and Deposit Mandates. I focused the
     commentary below on quoted excerpts, but before posting it I asked
     STM CEO Michael Mabe for permission to include the quotes. As I do
     not yet have an answer, I am posting the commentary with paraphrases
     of the passages I had hoped to quote. If I receive permission from
     Michael, I will re-post this with the verbatim quotes. As it stands,
     it is self-contained and self-explanatory.

     Full hyperlinked version of this posting:
     http://openaccess.eprints.org/index.php?/archives/507-guid.html

The International Association of Scientific, Technical and Medical
Publishers (STM) has circulated a fairly anodyne briefing to its member
publishers. Although it contains a few familiar items of misinformation
that need to be corrected (yet again), there is nothing alarming or
subversive in it, along the lines of the PRISM/pit-bull misadventure of
2007.

Below are some quote/comments along with the (gentle) corrections of the
persistent bits of misinformation: My responses are unavoidably --
almost ritually -- repetitive, because the errors and misinformation
themselves are so repetitive.

STM BRIEFING DOCUMENT (FOR PUBLISHING EXECUTIVES) ON INSTITUTIONAL
REPOSITORIES AND MANDATED DEPOSIT POLICIES

[Publisher policy on IRs is concerned with how IR deposit mandates might
affect publishing and publishing revenues, particularly in the case of
refereed final drafts.]

This is a fair statement: The issues for the research community are
research access, uptake, usage, applications, impact and progress. The
issue for the publisher community is their financial bottom line.

[Publishing and distribution today is successful and adequate. IRs
publish an inferior version.]

(1) IRs do not publish: peer-reviewed journal publishers publish. IRs
provide access to their own authors' (peer-reviewed, published) output
-- for all those would-be users webwide who cannot afford access to the
publisher's toll-based proprietary version -- so as to maximize the
access, uptake, usage, applications, impact and progress of their
research output.

(2) The version of the published article that the authors deposit in
their IRs is the final, revised, peer-reviewed draft (the "postprint"),
accepted for publication, but not the publisher's proprietary PDF. Hence
deposit does have the quality controls provided (for free) by the
peer-reviewers. (If the copy-editing should happen to detect any
substantive errors -- which is exceedingly rare! -- these too can be
corrected in the deposited postprint.)

[Publishing by IRs compromises quality, preservation and
discoverability.]

IRs are not substitutes for publishing but supplements to it, providing
access to research for access-denied would-be users, for the sake of
maximizing research progress. The deposited postprints have undergone
the essential quality-control for researchers: peer review.

The discoverability of postprints in IRs (via search engines like
google, google scholar, citeseerx, scirus and scopus) is excellent. No
problems, and no complaints from all the would-be users webwide who
would otherwise lack access to them.

(Preservation is a red herring: Preservation of what? As supplements,
rather than substitutes, authors' self-archived postprints are not the
versions with the primary preservation burden (although IR deposits are
of course being preserved). The primary preservation burden is on the
publisher's proprietary version, the official version of record, as it
always has been.)

[Exclusive copyright transfer is essential so the availability of
alternative versions does not prevent publishers from making ends meet.
Publishers add value in return for the exclusive rights.]

(a) In their IRs, authors deposit supplementary versions of their own
peer-reviewed publications in order to maximize their uptake, usage,
applications, and impact, by maximizing access to them.

(b) So far, all evidence is that this self-archiving has not undermined
the traditional toll-based (subscription/license) funding model for
peer-reviewed journal publishing: rather, they co-exist peacefully.

(c) But if and when IR deposit should ever make subscriptions
unsustainable for covering the remaining essential costs of
peer-reviewed journal publishing, there is an obvious alternative:
conversion to the Gold OA publishing funding model.

(d) What is definitely not an acceptable alternative for the research
community, however, is to refrain from maximizing research access,
uptake, usage, applications, impact and progress (by mandating IR
deposit) merely in order to insure publishers' current funding model
against any possibility that universal IR deposit might eventually lead
to a change in funding model.

(e) Unlike trade authors, researchers transfer to the publishers of
their peer-reviewed research all the rights to sell the published text,
without asking for any royalties or fees in return. They have always,
however, exercised the right to distribute free copies of their own
articles to all would-be users who requested them, for research
purposes. In the web era, OA IRs have become the natural way for
researchers to continue that practice, in order to maximize research
access, uptake, usage, applications, impact and progress.

[It is not just publishers that think IRs pose risks; librarian Dorothea
Salo has questioned IRs' costs and usefulness in “Innkeeper at the Roach
Motel”.]

(Publishers might do better to pay serious attention to the substantive
rationale and evidence concerning IR deposits and IR deposit mandates,
rather than to the opining of roach motel keepers.)

[It is inaccurate to speak of IR policies as “authors’ rights” policies
or “open access” policies.]

IR deposit mandates are accurately described as institutional open
access policy. (But IR deposit mandates are certainly not "authors'
rights" policies.)

[Talking points in responding to the media: subscription publishing does
require exclusive copyright transfer; perhaps OA publishing doesn't.]

This mixes up issues: The only relevant issue here for IRs and IR
deposit policies is whether or not the publisher has formally endorsed
providing open access to the peer-reviewed postprint immediately upon
acceptance for publication. (This is called a "Green" publisher policy
on OA self-archiving. It has nothing to do with author-pays/Gold OA
publishing models. And authors paying for the "right" to deposit would
be absurd and out of the question.)

[Should we endorse IR deposit? Under what conditions?]

If the publisher has formally endorsed providing open access to the
peer-reviewed postprint immediately upon acceptance for publication, the
publisher is Green. If there is no endorsement, or OA is embargoed, the
publisher is Gray.

[Should we make distinctions between preprint repositories, unmandated
IRs and mandated IRs?]

The only potential distinction is between authors' own institutional IRs
and institution-external 3rd-party central repositories. Although OA is
OA (and means free online accessibility webwide, irrespective of the
locus of deposit), some publishers only endorse deposit in the author's
own IR, in order not to endorse 3rd-party free-riding by rival
publishers: This limitation is innocuous, and no problem for OA. (In
fact, there are many reasons why it is preferable for both kinds of
Deposit mandates -- those from funders as well as from institutions --
to converge on institutional IR deposit, from which the metadata can
then be harvested centrally.)

What would be arbitrary (and absurd, and unenforceable) would be to
attempt to endorse only voluntary IR deposit and not mandatory IR
deposit by authors!

[Should we only endorse IR deposits that are open only to
institution-internal users?]

Let there be no ambiguity about this: Such a policy would be Gray, not
Green, on OA IR self-archiving.

[Should we endorse deposits that are open webwide only after an embargo
period?]

Without an embargo, this policy would be fully Green, and neither IRs
nor OA ask for anything more. With an embargo, it would be Gray.

[Should we only allow links from IRs to final versions on the
publisher's website?]

If the posting on the publisher's website is done immediately upon
acceptance for publication, and access to it is immediately open to all
users webwide, that would be fully Green too. (For such cases, IRs
could, for internal record-keeping purposes, mandate the deposit of the
author's postprint in the IR, but in Closed Access, with the OA link
going to the publisher's freely accessible version for the duration of
the publisher's embargo on making the IR version OA too: no problem.)

[Should we endorse deposits that are open webwide only for a fee?]

Paying to deposit in researchers' own IRs would be absurd, and roundly
rejected as such by the research community.

[Inform the media that publishers have made journal articles more
accessible today than ever before.]

True (though thanks also to the advent of the Web). But this literature
is not yet accessible to all those would-be users webwide whose
institutions cannot afford to subscribe to the journal in which it was
published -- and no institution can afford to subscribe to all or most
peer-reviewed journals. It is in order to maximize research access,
uptake, usage, applications, impact and progress by making all research
accessible to all of its would-be users webwide (not just those whose
institutions can afford to subscribe) that the OA movement was launched.
And that is why Green OA self-archiving, generated by funder and
institutional IR deposit mandates, is growing, to the great benefit of
research, researchers, their institutions, their funders, R&D
industries, and the tax-paying public that funds the researchers'
research and institutions.

(The publishing industry has to remind itself that the reason
peer-reviewed research is conducted, peer-reviewed and published is not
in order to fund the publishing industry, but in order to maximize
research access, uptake, usage, applications, impact and progress.)

[Online refereed journals are crucial for funders, universities,
authors, and authors' careers.]

Correct. And both the research itself, and the peer review, are provided
by the research community, free of charge, to the publishing community,
in exchange for the neutral 3rd-party management of the peer review, and
the certification of the outcome with the journal's name and
track-record. The publishing community is compensated for the value it
has added by receiving the exclusive right to sell the resultant joint
product (and no need to pay authors royalties from the sales of their
texts).

But that does not mean that researchers cannot and will not continue to
give away their own peer-reviewed research findings also to those
would-be users who cannot afford to buy the resultant joint product. Nor
does it mean that researchers' institutions and funders cannot and will
not mandate that they do so, in order to maximize research access,
uptake, usage, applications, impact and progress for the benefit of
research, researchers, their institutions, their funders, R&D
industries, and the tax-paying public that funds the researchers'
research and institutions.

[Depositing in an IR is not equivalent to journal publishing, with its
editing, peer review, and other added values.]

Correct. And individual authors depositing the final, peer-reviewed
drafts of their published articles in their IRs is not publication but
supplementary access provision, for those would-be users who cannot
afford paid access to the publisher's proprietary version.

[IRs might provide a lower quality option that makes publishers unable
or unwilling to perform their value-added services.]

This is merely the repetition of the same point made earlier:

No, IR deposits of peer-reviewed postprints of published articles are
not publishing, nor substitutes for publishing, they are author
supplements, provided for those would-be users who cannot afford paid
access to the publisher's proprietary version:

    (a) In their IRs, authors deposit supplementary versions of their
own peer-reviewed publications in order to maximize their uptake, usage,
applications, impact, by maximizing access to them.

    (b) So far, all evidence is that this self-archiving has not
undermined the traditional toll-based (subscription/license) funding
model for peer-reviewed journal publishing: rather, they co-exist
peacefully.

    (c) If and when IR deposit should ever make subscriptions
unsustainable for covering the remaining essential costs of
peer-reviewed journal publishing, there is an obvious alternative:
conversion to the Gold OA publishing funding model.

    (d) What is definitely not an acceptable alternative for the
research community, however, is to refrain from maximizing research
access, uptake, usage, applications, impact and progress (by mandating
IR deposit) in order to insure publishers' current funding model against
the possibility that universal IR deposit might eventually lead to a
change in funding model.

    (e) Unlike trade authors, researchers transfer to the publishers of
their peer-reviewed research all the rights to sell the published text,
without asking for any royalties or fees in return. They have always,
however, exercised the right to distribute free copies of their own
articles to all would-be users who requested them. In the web era, OA
IRs have become the natural way for researchers to continue that
practice, in order to maximize research access, uptake, usage,
applications, impact and progress.

[IRs cost money and should only be created if they have a distinct goal
rather than just parallel publishing and access-provision]

IRs are undertaken by universities and research institutions -- i.e.,
the research community. It is not at all clear why the publishing
community is providing this advice to the research community on its
undertaking...

[Researchers should be advised of the damage IRs could do to research
publication and dissemination.]

Researchers can and should be fully briefed about the already
demonstrated benefits to research, researchers, their institutions,
their funders, R&D industries, and the tax-paying public that funds the
researchers' research and the researcher's institutions -- the benefits
generated by maximizing research access, uptake, usage, applications,
impact and progress through Green OA self-archiving and IR deposit
mandates.

Researchers need this full briefing on research benefits, because it is
based on actual facts and experience.

But is the publishing community suggesting that -- in addition to these
empirical and practical facts -- researchers should also be briefed on
publishers' speculations about how Green OA self-archiving might
conceivably induce an eventual change in publishers' funding model?

Why?

    If and when IR deposit should ever make subscriptions unsustainable
for covering the remaining essential costs of peer-reviewed journal
publishing, there is an obvious alternative: conversion to the Gold OA
publishing funding model.

What is definitely not an acceptable alternative for the research
community, however, is to refrain from maximizing research access,
uptake, usage, applications, impact and progress (by mandating IR
deposit) in order to protect publishers' current funding model from the
possibility that universal IR deposit might eventually lead to a change
in funding model.

[Researchers should stay free "to choose how and where to publish."]

By all means. And they should continue to exercise their freedom to
supplement access to their published research by depositing their
postprints in their IRs for all would-be users webwide who cannot afford
access to the publisher's proprietary version.

[Institutions that want their employees to reserve certain rights for
their published journal articles should collaborate with journal
publishers so as not to damage their business.]

It would be excellent if all authors reserved OA self-archiving rights
in their copyright agreements with their publishers. Then all authors
could immediately deposit all their peer-reviewed research in their IRs,
and immediately make them OA without any further ado. But for at least
63% of journals, formally reserving that right is already unnecessary,
as those journals are already Green, so those articles can already be
made immediately OA today by self-archiving them in the author's IR.

For the remaining 37%, their authors can likewise already deposit the
postprints in their IRs immediately upon acceptance without the need of
either copyright reservation or any formal endorsement or permission
from the publisher: if they wish, they can set access to the deposit as
"Closed Access" -- meaning only the author can access it. Then the
authors can provide "Almost OA" to those deposits with the help of their
IR's "email eprint request" button: Individual would-be users who reach
a Closed Access deposit link (led there by the deposit's OA metadata)
need merely press the Button and insert their email address in order to
trigger an immediate automatic email to the author to request a single
copy for personal research purposes; the author receives the eprint
request, which contains a URL on which he can click to trigger an
immediate automatic email to the would-be user containing a single copy
of the requested postprint. This is not OA, but it is Almost-OA.

OA is indisputably better for research and researchers than Almost-OA.
But 63% OA + 37% Almost-OA will tide over the worldwide research
community's immediate usage needs for the time being, until the
inevitable transition to 100% OA that will follow from the worldwide
adoption of Immediate IR Deposit mandates by institutions and funders.

This is the information on which the research community needs to be
clearly briefed. The publishing community's conjectures about funding
models are important, and of undoubted interest to the publishing
community itself, but they should in no way constrain the research
community in maximizing access to its own refereed research output in
the Web era by mandating IR deposit universally.

To repeat:

    What is definitely not an acceptable alternative for the research
community is to refrain from maximizing research access, uptake, usage,
applications, impact and progress (by mandating IR deposit) in order to
insure publishers' current funding model against the possibility that
universal IR deposit might eventually lead to a change in funding model.

    The publishing industry has to remind itself that the reason
peer-reviewed research is conducted, peer-reviewed and published is not
in order to fund the publishing industry, but in order to maximize
research access, uptake, usage, applications, impact and progress.

[It is much harder, however, for institutions to successfully achieve
consensus on adopting an IR deposit mandate at all if the mandate in
question is a copyright-reservation mandate rather than an IR deposit
mandate. And because it is even harder to ensure compliance with a
copyright-reservation mandate (because of authors' worries that the
negotiations with their publishers to reserve immediate-OA
self-archiving rights might not succeed and might instead put at risk
their right to "choose how and where to publish"), the one prominent
institutional copyright reservation mandate (Harvard's) contains an
author opt-out clause that makes the mandate into a non-mandate. The
simple solution is to add an Immediate-Deposit requirement, without
opt-out. Even simpler still, adopt an Immediate-Deposit mandate as the
default mandate model suitable for all, worldwide, and strengthen the
mandate only if and when there is successful consensus and compliance in
favor of a stronger mandate.]

Stevan Harnad
American Scientist Open Access Forum