Comments on:

    "Dissemination of and access to UK research outputs"
    Response from the 
    Association of Learned and Professional Society Publishers (ALPSP)
    to the RCUK position paper
    http://www.alpsp.org/RCUKResponse.pdf

>  ALPSP encourages the widest possible dissemination of research
>  outputs; indeed, this furthers the mission of most learned societies
>  to advance and disseminate their subject and to advance public
>  education. We understand the benefits to research of maximum access
>  to prior work...

An excellent beginning!

>  ALPSP recognises that maximising access must be done in ways which
>  do not undermine the viability either of the peer-reviewed journals
>  in which the research is published

No one would disagree with this either.

>  Understandably, therefore, [publishers] may not wish their
>  "value-added" version to be made freely available in repositories
>  immediately on publication.

Quite understandable, and self-archiving is accordingly *not* about
the publisher's value-added version -- not the copy-editing, not the
XML markup, not the publisher's PDF -- but only about the own author's
preprint (unrefereed draft) and postprint (corrected final draft). That
is what is to be made freely available in repositories.

>  Even if the freely available version lacks some or all of the value
>  added by the publisher, it may be treated as an adequate substitute
>  by uninformed readers

The freely avialable version is intended for the use of those potential
researcher/users worldwide whose institutions cannot afford access
to the publisher's value-added version. It is accordingly a more than
adequate substitute for informed users who do not have acccess to any
other version!

>  (and, indeed, by cash-strapped libraries). And any new model
>  which has the potential to "siphon off" a significant percentage
>  of otherwise paying customers will, understandably, undermine the
>  financial viability of all these value-adding activities.

Surely the financial viability of the values-added is determined by their
market value. As long as the added values have a market value, they remain
viable. All evidence to date is that the self-archived free versions
co-exist peacefully with the publishers' value-added versions, serving as
supplements for those who cannot afford access to the value-added version
rather than substitutes for those who can:

    "[W]e asked the American Physical Society (APS) and the Institute of
    Physics Publishing Ltd (IOPP) what their experiences have been over
    the 14 years that arXiv has been in existence. How many subscriptions
    have been lost as a result of arXiv? Both societies said they could
    not identify any losses of subscriptions for this reason and that
    they do not view arXiv as a threat to their business (rather the
    opposite -- in fact the APS helped establish an arXiv mirror
    site at the Brookhaven National Laboratory)."

    Swan, A. and Brown, S. (2005) Open access self-archiving: An author study.
    JISC Technical Report, Key Perspectives Inc
    http://eprints.ecs.soton.ac.uk/11004/

>  The National Institutes of Health in the USA has attempted to address
>  this concern by delaying, for up to 12 months after publication,
>  the point at which deposited material becomes freely accessible. The
>  12-month period was arrived at after considerable discussion with
>  society and other publishers; it goes some way to addressing their
>  fears about the impact on subscription and licence sales. Even
>  the Wellcome Foundation, which has not consulted with publishers,
>  recognises the need for a 6-month embargo.

NIH and Wellcome embargoes concern the date of deposit in a central
NIH/Wellcome Archive, PubMed Central (PMC), in which the metadata and perhaps
also the full-text will appear in an enhanced ("value-added') form added
by PMC.

The RCUK mandate concerns the self-archiving of the author's own preprints
and postprints by the author in the author's one institutional repository,
for the sake of maximising *immediate* research progress and impact.

Research impact and progress are certainly not maximised by imposing 6-
or 12-month embargoes! The value-added publisher's version can wait,
but research itself certainly cannot, and should not.

>  Although in some areas of physics, journals have so far coexisted
>  with the ArXiv subject repository, some of our members in other
>  disciplines already have first-hand evidence that immediate free
>  access can cause significant damage to sales.

It would be helpful to see precisely what this "other" evidence is, and
precisely what it is evidence *of*. As physics and computer science are
the fields that have self-archived the most and the longest, and all of
their evidence is for peaceful co-existence between the author's drafts
and the publisher's value-added version, it would be very interesting
to see what evidence, if any, exists to the contrary. But please do make
sure that the putative evidence does address the issue:

   How much (if at all) does author self-archiving reduce subscriptions?

The evidence has to be specific to author self-archiving, anarchically,
article by article. It cannot be based on experiments in which journals 
systematically make all of their own value-added contents free for all
online, for that is not the proposition that is being tested, nor the
policy being recommended by RCUK!

>  We therefore recommend that the Research Councils should respect
>  the wish of some publishers to impose an embargo of up to a year
>  (or, in exceptional cases, even longer) before self-archived papers
>  should be made publicly accessible.

RCUK should require *immediate* self-archiving of the author's own
postprint drafts (and strongly encourage preprint self-archiving too)
for the sake of immediate research usage, progress and impact. Access
to the publisher's value-added version can be embargoed for as long as
the publisher judges necessary.

>  It should be stressed that any restrictions are intended simply to
>  ensure the continuing viability of the journals. which allow authors
>  (under either copyright model) all the rights which our research
>  indicates they require, including self-archiving;

The message is clear: Authors can and should self-archive their own
drafts ("inadequate" though these may be), immediately, for the sake
of research progress. The publisher's value-added version can be subject
to whatever restrictions publishers see fit to impose.

>  It seems to us both inappropriate and unnecessarily wasteful of
>  resources to create permanent archives of versions other than the
>  definitive published versions of articles.

It is not at all clear why publishers should be concerned with what
authors elect do with their own "inadequate" versions, in the interests
of research. Publishers' concern should surely be with their own
definitive, value-added versions, not whatever else the research community
elects to do to maximize research progress and impact.

>  [A] significant proportion (41%) of existing Open Access journals do
>  not, in fact, cover their costs

It is not clear why the topic has been changed here to Open Access Journals:
What the RCUK is requiring is self-archiving; it is not requiring publication
in Open Access Journals.

>  while ALPSP supports the principles which underlie the RCUK policy,
>  we believe that existing publishing arrangements go a long way towards
>  meeting the first three principles, and that publishers' concerns
>  about the potential negative impact of self-archiving must be
>  addressed.

Existing publishing arrangements go a long way, but the RCUK policy goes
the rest of the way, for the sake of all the potential researcher/users
worldwide whose institutions cannot afford the publisher's value-added
version, despite the existing publishing arrangements.

It is in order to put an end to the needless and costly loss of that
potential positive impact on research that the RCUK self-archiving
mandate has been formulated.

Stevan Harnad